Natural selection is a simple theory because it can be understood by anybody; to misunderstand it requires special training.
Graham Bell, The Masterpiece of Nature
Interest has been expressed in a thread on selection and drift, so I thought I’d start one, and offer my own 2-cent summary of the concepts.
Evolution, as commonly understood in biology, simply involves change in a lineage. Through mechanisms of change – principally, mutation and the insertion of ‘foreign’ DNA sequence – offspring frequently contain DNA sequences that do not derive by simple copying from their parent(s). This change is inevitable and, iterated, inexorable. There is no memory, no externally-stored blueprint for organisms; the specification of a species ‘floats on the breath of the population’, as Doctor Johnson wrote of the unwritten Gaelic language. Unless there is some kind of boundary blocking all possible avenues, a continuing source of variation is sufficient to keep a lineage exploring never-before-seen genetic space ad infinitum.
Among close ecological competitors (among, for example, the genetically similar members of a species), at a given locus in a finite world one individual ancestor’s genetic sequence is headed towards being the ancestor of every instance of that locus in a future population, and all the others with which it shared a population are heading for extinction. This derives from two facts: samples are more likely to deviate from the frequencies in the wider group than match them, and the probability of fixation of an allele is equal to its current frequency. The distortions on generational sampling tend to reinforce, through to extinction of all but one variant. This same tendency underlies the ecological principle of competitive exclusion between non-interbreeding competing species.
If a particular locus is invariant in a population, fixation has already happened. An original mutation, occurring in a single ancestor, has been passed to every member of the current population. Looking forwards, the mechanism of this concentration continues to operate, and so one particular individual from the present population will become the ancestor at that locus of all members of a future population. From any given starting point, a population of N diploid individuals will take a mean 4N generations to achieve fixation of one ancestor’s copy, and the probability for any diploid locus of being that copy is 1 in 2N. This doesn’t mean that large populations cannot fix neutral alleles, however – the number of mutations occurring scales with population size, so mutations will be fixed at the same rate they are generated, completely irrespective of population size. Doubling the population gives twice as many mutations taking twice as long to fix – the result is the same number of mutations being fixed per generation.
At the point of fixation, all instances of that locus descend from the same ancestor – they coalesce upon that ancestor. The case described – where there is no variation at all at the locus, ie there is just one allele – is the baseline process, the neutral case. If there is no variation, there is nothing for Natural Selection to ‘see’. The only process in operation is random genetic Drift – even though in this instance, it effects no evolutionary change because there are no variant alleles. The change occurred with the original mutation. This latter fact leads me to prefer the view of ‘descent with modification’ over the population geneticist’s ‘change in allele frequency’. It is true that allele frequency change is also evolution, the only part over which selection and drift have a role, but as far as each lineage is concerned, the change occured at the moment of mutation. The lineage changed at that point; the population changed somewhat later, when this mutation became the norm.
Suppose we could uniquely label the locus for every member of the population, in a heritable manner. Now, we have essentially created 2N alleles. If we allow them to operate neutrally, just as when there was no variation, evolution will now occur because allele frequencies must change in the population. Because our labelling has had no effect on the neutral ancestry-fixation process, the label itself will surf to fixation on this process, while all others become extinct.
If, instead of labelling every instance, we simply labelled one, we would find that it still had the same 1 in 2N chance of becoming fixed. And this is the situation for any neutral mutation: 1 in 2N neutral mutations will become fixed; the neutral mutation simply functions as a label.
So now, having laboured the neutral case, where all is Drift, we can look to introduce a differential between alleles. If a new allele consistently performs better or worse than the existing one – meaning that it enhances or hinders the survival and/or reproduction of its bearers – then Natural Selection has come into play. It is a simple and obvious and non-tautological!) truth that a consistent increase in survival/reproduction – in fitness – will tend to favour such alleles over the purely neutral case, and render fixation more likely and speedy, while a reduction will increase the likelihood and speed of elimination.
Unlike the purely neutral case, in which population size is cancelled out, the behaviour of selectable alleles is affected by population size. In smaller populations, random factors have a greater influence than in larger ones, and hence alleles may behave as effectively neutral despite possessing an advantage which would see them selected in a larger population.
Drift does not simply disappear when you start to turn up the selective ‘heat’. Drift essentially derives from random sampling, the tendency of subsets to deviate from the distribution of the complete set, and such sampling is in effect almost all the way along the continuum of selective advantage (apart from alleles that are so strongly detrimental that they never gain a foothold). Even a favourable allele can disappear through Drift, likewise a deleterious allele can become fixed through the same mechanism. But more often, progress will go with the expectation, not against it. The large-number tendency is for genomes to become enriched in advantageous alleles and impoverished in detrimental ones. Because this process is environmentally conditioned, it allows populations to adapt to their circumstances, by purging the traits that do worst in the recent environment.
There continues to be a debate about the relative importance of Selection and Drift in evolution generally, and in driving speciation among sexual forms. Only selection can be adaptive, because it is the only component that is responsive to the environment. But they both have significant contributions to make, and cannot readily be teased apart. Both tend to reduce the variation in a population, which variation is only restored by mutation, recombination or immigration.
I’ll try to serve as a resident expert here. For the population genetic theory underlying all this, my course notes “Theoretical Evolutionary Genetics” can be consulted. They are available free here. There is also a nice recent book by Nielsen and Slatkin.
A few quibbles with Alan’s very clear presentation.
* When Alan gives the time 4N for the population to complete genetic drift, that is already for the case of neutral alleles. Any selection changes that.
* When he says that “the number of mutations occurring scales with population size, so mutations will be fixed at the same rate they are generated” he means the rate at which mutations are generated in a single copy. Of course, with a bigger population, more new mutations occur overall. But only one lineage of genes, chosen at random, has a long-term future, so that it is the mutations in the lineage that count.
* With more elaborate forms of natural selection, such as overdominance, natural selection can act to keep variation in the population. It doesn’t always act to eliminate it.
Want to listen to a whole course on this? Just search in, say, Google for the phrase “Genome 562 UW” where you will quickly find my theoretical population genetics course, complete with audio files of my lectures.
Very helpful. Thanks to both of you.
I requested it and I thank you. I have family visitors tonight, but will read it tomorrow.
A Lot said.
I say natural selection is not a theory of evolution and its not simple.
Biological life is more complicated in its nature that physics.
There is nothing simple about it.
In fact evolutionary biology is a simplistic hypothesis for a complicated thing.
It shouldn’t be called natural selection but instead natural mutationism.
Its the mutations that are claimed the origin for the glory of biology and not mere selection.
Not NS and not simple.
I do think biology is one big memory machine and so diversity in biology came from triggering this memory after the fall and later for needs of biology.
A intro criticism here.
Natural mutationism.
Hmmmmmm.
If you meant my notes, they are almost 500 pages long, so you may need more than one evening! The material on genetic drift versus selection is in Chapters 6 and 7, particularly the latter. However you may need material from chapters 2 and 5 to understand that.
I’m not suggesting people need to read my notes to comment here.
Joe Felsenstein,
A few quibbles with Alan’s very clear presentation.
Yes, I hinted at this, by indicating that fixation of non-neutral alleles is more ‘speedy’, but could have made it more explicit.
Yes, I jumped from the fate of a single locus to that of multiple loci evolving in parallel (in populations with crossover). A common error among Creationists is to argue that neutral Drift can have no power to change large populations because 4N is a lot of generations; they forget the contribution of mutation rate, acting across the genome.
True, also frequency-dependent selection.
Robert Byers,
How can you get so much wrong into one sentence?
Indeed there is nothing simple about living organisms, and there are complex dynamics at play. Nonetheless, the principle of Natural Selection is itself simple, and unquestionably a theory of evolution.
Certainly, without a source of variation, lineages would gradually ‘freeze’, by what amounts to wholesale inbreeding (even if mating individuals were only very distant relatives). Selection is only part of the picture.
I don’t want to get bogged down in apologetics, but where in the Bible does it say diversity arose after Genesis?
But evolution is not just a trait appearing in one organism. Mutation does not explain change in a population.
A couple more quibbles:
This is true in the long run, once the population has returned to equilibrium. More immediately, fewer mutations will fix. This can be an important practical distinction. For example, the recent large expansion of the human population means that equilibrium won’t be reached until well after the expanding sun has killed all life on Earth.
This is a little misleading. For new mutations, which start with a single copy, the great majority of beneficial alleles are lost through drift. This makes sense: when there is a single copy of an allele, its probability of being passed on is similar, regardless of whether it yields an average of 1.00 copies or 1.01 copies in the next generation. Either way, it’s a crapshoot.
Steve Schaffner,
1) Agreed – I meant ‘taking a population in equilibrium that is twice the size’, rather than doubling that population.
2) Agreed again, and again badly expressed. I was thinking of an equilibrium starting position, where the alleles are present at 50% apiece, and should have said so.
(ETA: Rethinking, I’d stand by what I wrote in the single-copy case too. Seed a succession of populations with individual beneficial, neutral and detrimental alleles in single copy and track progress, and the beneficial ones will fix more often than the neutral or detrimental ones, over large numbers of trials).
I was just referring to the thread. What I am interested in is “the controversy”.
We can make this more dramatic using Kimura’s 1962 formula for fixation probabilities, which is an approximation but a very excellent one.
The formula is, for population size N and selection coefficient s in favor of the allele, (1-exp(-2s)) / (1-exp(-4Ns)).
If the population size N = 10,000, a single copy of a neutral allele has probability of 1/20,000, as it is just as likely as each of the other 19,999 copies to drift to fixation.
If it is deleterious, with fitness 0.99, its probability of fixation is 3.8689 x 10^(-176), an awfully small chance.
If it is advantageous, with fitness 1.01, the probability is 0.0198013.
So the advantageous mutant is 396.026 times as likely to ultimately get fixed as a neutral mutant would be.
Since this speaks directly to my confusion, why is selection spoken of as an obsolete concept by some biologists. Not as being wrong, but as being over-emphasized?
I think they are reacting to past eras when many biologists assumed that if they saw a pattern, it was due to natural selection on that character. This ignored not only genetic drift but the possibility that natural selection was on a different character that happened to share some of the same genes as the observed character. Thus selection on (say) height might change weight as well.
Genetic drift can have major effects but I think too often people underestimate what natural selection can do.
If anyone wants to play around with that in computer simulations, there is a one-locus population genetics teaching program called PopG which I have used for years. The current version is in Java, so it will run on all platforms (for Windows you may also need to download a Java engine from
java.org. PopG is available for download here. You can try out different intitial gene frequencies, different fitnesses, and different population sizes and see what happens in multiple populations.Joe, is there any was to isolate the relative contribution of the various engines of creation and selection that make up evolution for a family of organisms?
Obviously selection selection has certain anatomical requirements, etc.
Richardthughes,
Creation and selection?
Some things create novelty and some things select for them (natural selection / sexual selection)?
Mostly mutation.
You could try PopG where you can play with mutation rates and fitnesses, etc.
Sweeping generalities are not easy to make, but for a single locus the relative strength of evolutionary forces can be judged by comparing s, u, m, and 1/N,
where
* s is the selection coefficient (fractional difference in fitness)
* u is the mutation rate
* m is the migration rate
and
* N is the effective population size.
petrushka,
I think there is something of a tendency towards wanting to be seen as revolutionary. There is also resentment of some of the popularisers and their focus upon selection. Dawkins’s ‘Selfish Gene’ relies upon selection (and, incidentally, crossover) for its power. And many people find that an incomplete presentation of the subject to the lay public. But … it’s about what it’s about.
I might add that I find Dawkins’s approach tremendously persuasive and stimulating. He drew, of course, upon the work of Maynard Smith, Hamilton and Williams. But as a biochemistry undergraduate in 1976, the centrality of the nucleic acids and their efffects on ther own persistence, made a great deal of sense. I’m still a bit of an adaptationist at heart, though I don’t think it explains everything.
I don’t see how this can be right, entailing, as it seems to me to do, that the claim that “the fit are more likely to survive” is in fact NOT tautologous. I’m therefore going to just assume that it must be WRONG, because that claim simply must be a tautology, because several bigshots in the ID world have said so (or maybe it’s in the Bible or something).
Either way, it’s wrong.
W
Joe,
What does advantageous mean? Isn’t the definition of advantageous-that it is more likely to spread through the population?
So what you have said is that if it is more likely to spread through the population, then it is more likely to spread through the population.
phoodoo,
It is more likely to spread through the population if it leaves more offspring on the average.
Allan Miller,
Oh my goodness, the way you guys use words. You mean it is more likely to leave more offspring if it leaves more offspring? More offspring than what, than others leave offspring?
I mean, you might as well just say, it is more likely to leave more offspring, if it has more babies. Or is it more likely to leave more offspring, if the odds are better that it will leave more offspring. Or it is more likely to leave more offspring , if the likelihood is HIGHER that it will leave more offspring.
Or perhaps you prefer, “It is more likely to spread through the population, if its genes become more distributed through the population, by (lets see, what can we replace “spreading” with, that means exactly the same thing, but sounds like we are saying two different concepts, oh wait, got it..) propagating throughout the population in increased numbers.”
“Or best still, It is more likely to spread, if it spreads!” Perfect! Spreads is different than spread see?
Have you ever played Risk, Phoodoo?
phoodoo,
I don’t know what you’re on, phoodoo. If a trait causes its bearers to have, on the average, more offspring than its locus alternative, then it will spread, it is advantageous, it is beneficial, it is … what am I not getting, regarding the profundity of your objection?
phoodoo,
Of course not. If it happens to leave more offspring, it is more likely to spread.
Look on it as compound interest on investment. The investment with the greater return will accrue more capital. Is there something problematic for the world of finance with such a statement?
Allan,
In order for you to completely miss the entire circle of Joes post, you would have had to have not read Joes post I think.
He gave the liklihood of a trait spreading through the population one would expect to see if a trait is neutral, deleterious and advantageous. And what is the meaning of a trait being neutral, deleterious and advantageous? Well, the meaning is, if it fits those percentage of likelihood, that it was just described as having. If it fits one of the other likelihoods, then it must be that meaning. In other words there is no other meaning to the terms Neutral, deleterious and advantageous, other than fitting those likelihoods of spreading. A trait can’t be advantageous and spread less than that likelihood on average, and a trait can’t be deleterious and spread more than that likelihood on average. The definition is the result.
So, if you don’t see the issue here, I am sorry, there is no easier way to explain it to you I am afraid.
Maybe let me try another way Allan. If a trait is thought to be deleterious, but spreads on average through the population much higher than the expected result-then where does the discrepancy lie-is that trait in fact NOT deleterious as they thought it was, or is the trait an exception to the rule?
How many exceptions to the rule can there be, before the rule has to change? None. Because if it is an exception to the rule, then it doesn’t fit the definition of deleterious, so it must be advantageous.
Just like a peacocks feathers.
We’ve been over this in another thread, ad nauseam, and phoodoo just doesn’t get it. Not about batting averages, not about fitness, not about compound interest.
Now if we have a diploid population, and the three genotypes AA, Aa, and aa have fitnesses 1.0 : 0.90 : 0.95, and we start out with a gene frequency of 0.4 for allele A in a population of 10,000 individuals, what is the probability of fixation? Kimura’s 1962 formulas allow us to compute it very closely, though with some need for numerical integration.
But I won’t bother doing the calculation because phoodoo thinks the answers are vacuous because they’re circular. We’ll just ask phoodoo. What’s the obvious, trivial, circular answer?
Dance, phoodoo, dance!
petrushka,
I think a lot of it has to do with whether the emphasis is on the genotype or the phenotype.
When you’re focusing on the genotype, you see a lot of variation that is caused by drift. When you focus on the phenotype, where the effect of neutral and near-neutral mutations is invisible or nearly so, then selection takes on more prominence.
phoodoo:
Allan:
A big part of phoodoo’s problem is that he can’t see the difference between a fait accompli and a propensity.
Joe Felsenstein,
Joe,
It seems very hard for you to answer a very obvious and easy question. Can there be an exception to the rule that says this is how often a deleterious mutation will spread through the population on (average!!)?
Clearly, if a mutation (according to your own words) that is said to be deleterious spreads though-out a population at a rate higher than the definition for a deleterious mutation, it is no longer considered a deleterious mutation! So its not an exception to the rule!
There can not be exceptions. THAT is what makes the rule a tautology!
Stick to numbers Joe, logic isn’t your strength.
May I suggest if phoodoo has had questions answered but is unsatisfied, the simple thing is to ignore repetitions. Maybe a link to previous responses at most.
You can lead a horse to water…
All rules are tautologies you nincompoop.
The definition of acceleration is an increase in the rate of change of velocity over time. Omg, if a vehicle has increased the rate of change of velocity over time, it has accelerated. There can be no exceptions. In so far as the rate of change in velocity over time has been negative, it has no longer accelerated. ITS A TAUTOLOGY OMG.
So what? It can still be used to predict future outcomes. For example if we measure the rate of change of velocity, we can predict how fast the vehicle will be going, or how far it will have got, at any future point in time.
Same for beneficial mutations. Yes, the definition is a tautology(definitions always are), but we can still use it to predict outcomes because we can measure how well carriers of certain alleles do on average, that way we can predict how likely it is that, when such an allele arises in the population, it will have some number of decendants X generations later.
Get it?
I look forward to seeing *your* book.
phoodoo,
Because allele spread is probabilistic, there is a frequency distribution of outcomes for each possible selective advantage. Individual runs must visit all parts of the distribution, so one run is not informative as to the underlying advantage/disadvantage. The ‘rule’ is actually that some runs MUST fail and some succeed, if there is a non-zero probability of either outcome. Fixation or extinction on any one run is not an exception to anything, but is expected. One or other is the ultimate fate of (nearly) all alleles, regardless of benefit. Advantage conditions the relative proportions of these fates over multiple trials.
Multiple trials, of course, are frequently not possible. In nature, each allele typically only gets one shot. But it’s not enough to say it got fixed therefore it can’t have been deleterious. It can’t have been particularly deleterious, because mathematical analysis argues against that. But whether mildly deleterious, neutral or beneficial, requires more information. And if a deleterious or neutral allele gets fixed … well, that’s still evolution, phoodoo.
phoodoo, I think the moral here, is that pretty much everybody has heard your “it’s tautologous!” plaint, everybody understands it, and nobody agrees with you or your buddies that you’ve actually found some kind of insurmountable problem for either science or baseball. Both the science and the sport progress, in spite of the fact that we learn what sort of properties provide propensities for certain results from aggregating the results and studying them.
YES, people can use “fitness” and “survival capacity” as synonyms if they choose to. NO, it’s not a problem.
Ich habe genug.
Allan Miller,
Allan,
I have a theory. No its a rule actually. I made the rule. It goes like this:
Good baseball hitters have a spread of where they hit the ball:
Good hitters ALWAYS hit on average 60% to right field, 30% to left field and 10% up the middle.
Mediocre Hitters ALWAYS hit (on average once again) 60% up the middle, 20% to left field and 20% to right field.
Bad Hitters hit percentages which are not in either of these two ranges.
Now, can you prove this wrong? You can’t . Its undeniable. Its totally unfalsifiable because the definition of someone who is a Good hitter, is someone who hits it 60% to right field, because I decided this is what the definition of a good hitter is. Its not batting average, or home runs, or anything else you might THINK it is, it is what I defined.
This is EXACTLY what you are doing when you make a definition of beneficial, neutral or deleterious, based on the predictions for how often they will spread though a population. There is no other criteria for measuring beneficial, deleterious or neutral, other than how it spreads through the population. The theory, and the results are one in the same.
For anyone who can’t understand this, you should lose all rights to comment about anything in the future. A third grader might have trouble getting this. A seventh grader never would. If Joe Felsenstein can’t get this, he has no business teaching a university level class. Is a very very simple concept. You need a different criteria for confirming your theory, then the definition of your theory. They can’t be the same thing.
Keiths, Allan, Joe, Alan, Omagain, Guillermoe, …you all lose the sweepstakes for being able to judge the validity of any theory. Its way over your heads. A simple point like this, you can’t understand. Its shocking really.
Although I have to say Alan Fox may still be smart enough to evaluate theories, because I suspect that he has realized that I am right, and that is why he suggest you don’t respond to me, because he is starting to see that I am showing how wrong you are.
Nice catch Alan.
“helpful or favorable”
You have been presented with an example in another thread. Resistance to a toxic compound, like chloroquine, is adventageous for a population of parasites exposed to that substance.
It’s not adventageous BECAUSE those parasites have increased survival rate. Those parasites have increased survival rate BECAUSE they are resistant to chloroquine and that is adventageous.
A runner is not fast because he wins the race. He wins the race because he is fast.
Ok, resistance to chloroquine and “survival of the fittest”. Where do you put resistance to chloroquine ins “survival of the fittest”? Is it survival? Is it fittest?
You are fucking up with the same idiocy over and over again. A fit runner tends to win more races. He wins more races because he is FAST. That’s what we call being fit. Being fast. He wins the races beacuse he runs fast. A fit boxer wins the fight because he hits hard and swift.
So, in general, the fittest wins. But fit is not defined, in general terms, as winning. The fit boxer is not fit because he wins. He’s fit because he hits hard and swift, and he wins because he is fit, i. e. he wins because he hits hard and swift.
The fit runner is fit because he runs fast. And he wins the race because he is fit, i.e. he wins the race because he runs fast.
No tautology. You are just getting lost in a phrase instead of consider the real concept and it’s application.
phoodoo,
It’s shocking that you should presume to school university professors in their own field. But … meh. I have before me a 763-page textbook, with 44 pages of references, 42 to a page, all dedicated to this supposed empty tautology. Kinda makes you think, don’t it? OK, I guess not.
Let me stop you right there. You are talking bollocks. The definition of advantage is not based on how it spreads through the population in a single trial. As I patiently explained, a single trial to fixation or loss is silent on the underlying advantage. But you can measure the advantage by counting the mean offspring numbers accruing to carriers and non-carriers. You don’t need to track progress through the population. If carriers produce consistently more offspring, then that is an advantageous trait. If fewer, it is disadvantageous. If they are about the same, neutral. Having thus determined the advantage, it is but a small intellectual step from this to understanding that the type with the greater mean offspring numbers will tend to contribute more copies to future generations than the type with lesser. But since evolution is not deterministic, and offspring number is a random variable, some alleles will go against that expectation. Most won’t.
You seem be spending an awful lot of time attempting to knock holes in a theory that you claim does not exist.
Cognitive dissonance much?
By the way, phoodoo, you didn’t answer some questions.
Back your idea that survival of the fittest explains anything: Show us a concrete situation where, for example, two traits are present for a characteristic and “survival of the fittest” explains both traits.
And, try to explain this without survival of the fittest:
” On the Hawaiian Island of Kauai, more than 90% of male field crickets (Teleogryllus oceanicus) shifted in less than 20 generations from a normal-wing morphology to a mutated wing that renders males unable to call (flatwing).”
You can use this example to explain how singing male crickets are as fit as silent ones.
I have an idea. I’ll open a couple of threads. One on the theory of evolution with a brief summary and an invitation for others to add good resources, examples, papers etc. that they have found useful, interesting and convincing. Ditto for ID. Unless you’d like to author the ID OP.
I’ll have chance later, I hope.
Allan Miller,
Allan, If (on average, on average , on average!! , let’s not forget I already included your caveat) a deleterious mutation spreads through the population more than is predicted, is it still a deleterious mutation?
If you still consider it to be a deleterious mutation, by what basis is it considered deleterious?