Is there more than one definition of natural selection?

Percy, although I don't want to screw things up similarly to what happened on the other thread on NS, I'm not sure this is an accurate description. Possibly I'm misunderstanding what you're trying to say.Almost every definition of natural selection I've read emphasizes that in order for selection to occur, there must be some difference for selection to act upon. Whether these differences are allelic (as Mod, Justin and the other Defenders of Evolutionary Purism prefer), phenotypical traits carried by individuals, or even characteristics posessed at population or species levels, there has to be some consistent difference. As Douglas Futuyma defines it (in the glossary to Evolutionary Biology),

quote:

---

The differential survival and/or reproduction of classes of entities that differ in one or more characteristics; the difference in survival and/or reproduction is not due to chance, and it must have the potential consequences of altering the proportions of the different entities, to constitute natural selection.

---

As you can see, the key focus here (whether at the genetic level or the species), is that there must be some kind of differences for natural selection to operate. What you described with the marble example above would therefore not be natural selection - unless there was another group of marbles within the overall population that didn't pass through the holes for some reason.The conditions you describe could, of course, exist at least theoretically. It would be the ideal case of HW equilibrium. I can imagine a population of invariant organisms (at least invariant in NS terms) exposed to absolutely equal environmental pressures simply continuing to exist and reproduce ad infinitum, until and unless the pressures change. Whether such a population could exist in the real world, with all its constant microvariations is arguable, IMO.

I think the basic difference between what I and Mod actually agreed with on the previous thread, and what you are trying to say here, is that natural selection per se doesn't include the kinds of chance events that you describe. Chance events obviously have an effect on individual survival, and may even change the proportion of alleles/genotypes/phenotypes existing in subsequent generations. However, NS is inherently "anti-chance". Let me try and explain this point using a hypothetical population.1. Chance: Assume a population containing 15 individuals. Ten of them have characteristic A, five have characteristic B. Neither A nor B provide any selective advantage in the current environment. The traits confer no difference in reproductive rate, survival, etc. In other words, all the marbles are the same size, but some are red and some are blue. A chance event occurs - lightning strikes the tree in which they all live. Simply by the luck of the draw, 3 A's and 4 B's are killed. Is this selection? I would argue "no". The proportion of A's and B's has changed - and assuming this breeds true - but it wasn't natural selection that changed them, simply because the differences between them had no bearing on either survival or reproduction. There may have arguably been selection of A's, since more of them survived, but there was no selection for A - which is the definition of NS.2. Natural Selection: Assume the same population. In this case, however, there is a climatic fluctuation - an ENSO event, for instance - where the characteristic B provides a slight survival advantage. More B's live to reproduce in subsequent generations. Slowly (depending on the severity of the pressures) B's will come to represent a larger proportion of the population over time. Both A's and B's are still reproducing "after their kind" , but B's are surviving better - and hence are providing a greater contribution to subsequent generations. If the pressure is maintained over sufficiently long time frames, it is possible that A's may become extinct, and the population contains only B's. This is natural selection, and at the end there it constituted evolution. If things go back to "normal" before this occurs, however, thenThis latter case is essentially what the Grants found in the Galapagos. You probably know the story. There was a severe, multi-year drought brought on by an extended ENSO. Long-beaked finches began to die off because they were able to forage less and less on the fewer, hard-shelled seeds. Short-beaked finches, better able to deal with harder seeds, had a slightly better chance of survival. Both types were hard-pressed, and died in droves (the population collapsed). It finally came down to a 4mm difference determining the potential for life or death. Natural selection for different beak sizes based on available food supply. However, unlike what someone mentioned on another thread, this wasn't evolution. Short- and long-beaked finches kept being produced in the normal proportions. The long-beaked birds simply died of starvation before reaching adulthood. This particular selection pressure simply didn't last long enough for the overall change in alleles to become fixed. When the weather went back to normal, so did the proportion of different sized beaks.Chance may promote one individual over another in a population, but it is random, and thus can not really be considered natural selection. Selection, on the other hand, promotes an individual based on its posession of a particular characteristic.

Hi Percy,I fully concur with your desire to explain things in the simplest possible way before jumping into details - the old walk before you run concept. That is, in fact, the point of the marble analogy - a simpler explanation of what happens in NS I have never encountered. However, I think you've actually introduced more - and unnecessary - complexity with your fitness 0 discussion, as well as lumping chance under NS. Finally, I'm not sure where you derived your graph, but the only thing I can recognize in what you're graphing is possibly a population-level response to cyclic ecological change. It certainly doesn't appear to have any bearing on fitness, or NS (beyond the cyclic nature of some types of selection pressures). It certainly doesn't represent what happens in "differential reproductive success over time" that I can see. I agree with your first sentence. So I have to ask: Why are you doing precisely that? Isn't that exactly what first your re-wording of the marble analogy and now this weird graph is intended to convey? That fitness can be 0? If I'm confused on this, I would be willing to bet others are as well. Obviously, I'm not understanding you at all.From your message 24 on this thread:

quote:

---

And so in my version of your marble example I referred only to selection, not natural selection, and it was intended to be analogous to only a single generation. Even though the marbles had identical phenotypes with respect to size, that is no guarantee that the next generation would all be the same size. The marble drop represents the selection part of the process, not the reproduction part.

---

NS operates on each generation independently, although the outcomes of NS - adaptive response being one example - can only be seen after multiple generations. You are vastly overcomplicating what happens by claiming that selection - and I'm unclear what the distinction you're making between "selection" and "natural selection" actually amounts to - doesn't affect changes from one generation to another. This would be an extraordinarily special case. Further:

quote:

---

In the next generation some of the marbles may be of different sizes. Since selection pressures are operating in each generation, sometimes choosing more, sometimes less, sometimes all, sometimes none (extinction), the process of natural selection is going on. The fact that in one generation every marble survived doesn't mean natural selection suddenly stopped or was absent.

---

No one has EVER said this. In fact, near as I can make out, you were the one to introduce this idea. Regardless of your perspective on the gene/individual/whatever debate, NS is constant although the exact composition of the selection pressures is highly variant - leading, ultimately, to changes in the relative contribution of organisms with different phenotypes/genotypes/alleles/whatevers to future generations. This is NS, and it don't get any simpler.With ref to NS vs AS, I'm not sure why you brought it up. It's never (to my knowledge) been a point of disagreement between us. You had some quibbling with Mod in the other thread, but you and I certainly didn't differ on the definitions.If you're interested in simplifying things, I'm all for it. However, please be careful how you go about doing that, as there is potential to lose meaning at the same time.

With your permission, I'll lump the two responses into one, here. I am not peeved. I am honestly confused over what you're saying. I don't want to get into a "who said what to whom" digression here, so I'll just address what I think are the main areas I don't "get".1. The graph is really throwing me. Beyond the mechanics of drawing a graph on EvCForum (an explanation for which would unfortunately be wasted on me given how technologically challenged I am at best), I literally can't comprehend what you are trying to show. Assuming we are using the same definition of fitness (i.e., {avg. fecundity} X {% progeny surviving to reproduce in the F1}), I don't see how the graph represents this. The only synonym I can envision for "zero fitness" under that definition is "extinct" - i.e., either the organism didn't reproduce (zero avg. fecundity), or none of the progeny survived to reproduce (zero % survival). Moreover, I am having a difficult time wrapping my brain around the idea of "negative fitness" in this context. How does a fitness function become negative? Apologies in advance for the reductio ad absurdum, but when I tried to figure out what you meant the first image that popped into my head was someone going back in a time machine and killing the organism's ancestors (I've probably read too much science fiction). It's fine if you are using a different metric for fitness, but you need to explain what you're using (or maybe explain what it is you're actually trying to illustrate if it isn't fitness), 'cause otherwise the graph doesn't make sense to me.2. Okay, I think I'm beginning to catch a glimmer of what is going on - and where my confusion seems to arise. You and I tend to regard NS as a process, whereas Hoot is thinking of it as an outcome. Differential reproductive success is indeed an outcome of NS, and from that viewpoint Hoot isn't wrong in what he wrote. I don’t agree with him that this is necessarily a good way of looking at NS in isolation, however. It is sort of like the genecentrist inability to separate “selection” from “evolutionary response to selection”. It doesn’t mean the viewpoint is invalid, it just means that they’re not making as fine a cut as others are.Perhaps I can contribute some simplified, short “definitions” that might clarify (or potentially confuse things further):1. environmental selection pressures: The biotic and abiotic factors that impact an individual organism or population of organisms during their lifetimes. Selection pressures operate continuously, but are variable in their effects, strength, and nature.2. natural selection: The process, based on environmental selection pressures, that filters genotypes/phenotypes in the wild. Depending on the relative strengths of these selection pressures, the outcome of natural selection may influence the relative contribution of these genotypes to future generations. It may increase, decrease or stabilize the proportional representation of a particular genotype in the population.3. adaptive response (evolution): The long-term permanent change in the relative proportion of alleles or genotypes in a population in response to natural selection.I don’t know if that does any good, or merely obfuscates things further. One possible suggestion - don’t try and directly respond to someone’s misconceptions (Hoot, in your case). Rather, try and develop a working and concensus definition of natural selection (if that’s your goal). And THEN address the misconceptions in that context.Edited by Quetzal, : edited out redundant sentenceEdited by Quetzal, : clarity

Hi Percy,Just got back from the field last night, so now have some time to post again. I'm not sure your characterizing my interactions with Hoot accurately. It's less that I think he's right about a lot of things, but rather that I discern some validity to the questions he brings up. Unlike Brad, who I've never understood, Hoot DOES seem to be on the right track fairly often. It's like trying to evaluate a boundary condition - Hoot seems to come right up to the edge, so I can't help but try and explain things in a different way, or think that if I answer the question clearly the little light will come on and he'll make the crossing. I may be wrong, of course. Or maybe it has something to do with puppies...Anyway, I think it's worth trying. I can understand yours, crash's (and other's) frustration. I may arrive at that point myself eventually. I'm just not there yet.

This is pretty much what I was trying to convey. Obviously the way I went about it was, erm, confusing. Apologies. Glad there's someone on the board that can explain things better than I can. Actually, the only person that seems to be having a problem with this concept is Hoot. It's not clear why, however. We've all been saying this consistently, albeit in different ways. It's one of those boundary issues I mentioned. He's absolutely close to the edge of understanding, but can't seem to step over the line.

Wait a second, Hoot. A bottleneck is a result, not a process. It can be the result of sampling error (as in the founder effect), the result of an extinction event, or even the result of genetic drift (in a very small population). It is not, however, a mechanism of genetic drift or natural selection. It simply represents a severe restriction of a population's gene pool. Careful with your terminology.

Why? It's just a model for large-scale extinction that tries to explain the apparent "selectivity" (i.e., why this one and not that one) of mass extinctions. As near as we can figure, the survival or not of taxa in a mass extinction is simply random, although some taxa may have a slightly better chance of recovering than others after the event (e.g., lazarus taxa) due to natural selection. It really has nothing to do with natural selection per se, as NS is seen to operate after rather than during. At this level (species or higher) I'm not even sure you can talk reasonably about natural selection in the first place, although that may be an argument for another thread.

Okay Hoot, I give up - I can't guess the answer. What is an "evolutionary rest state"? "Evolution" implies "change", after all - it's what the word means. If there's no change, how can there be evolution. Did you find this term somewhere, or did you just make it up to try and explain some concept I'm not grasping?

The founder effect (where a fragment of a larger population gets isolated for one reason or another), is one way a bottleneck can occur. Sampling error is inherent in a founder population, as it is highly improbable that the founder population will have a full representation of all the available alleles in the parent population. Perhaps I worded things improperly. So while it is true that sampling error itself does not reduce a population (i.e., create a bottleneck), it is a description of what occurs at the genetic level in this type of bottleneck. No. Sampling error represents that fragment of a population's gene pool that gets segregated in a founder or otherwise severely reduced population. If you have 20 pairs of brown and black socks, and one bright orange pair with duckies on it (that your maiden aunt gave you for your birthday because she still thinks you're five), and you randomly select 3 pairs in the dark for a trip without looking at them, the odds are you will have pretty much all brown and black socks. In this case, your traveling (founder) socks would NOT include the full range of possibilities. But you could end up with the orange pair. This is sampling error.

Because it isn't. As Allopatrik pointed out, a bottleneck simply refers to any severe reduction in population size. The process that causes the reduction may be many things - your example was over-hunting, for instance. The result is a genetic bottleneck. I try not to invent terms, especially when they can lead to confusion. I might suggest you do the same. I see NS as a process. I thought I'd made that clear. It's a mechanism, not an outcome. We measure the outcome of NS in different ways depending on what kind of question we're trying to answer - reproductive success, change in allele frequency, change in fitness, etc. Each of these measures tells a different tale, but the process that leads to them is NS in each case. Drift is also a process. It's just a non-selective (non-discriminative, if you prefer) process, whereas NS is discriminative by definition.

Actually, we're both correct. In the most generic sense, sampling error refers to any random variation of alleles between an adult population and its ancestral population. If a population, for instance, carries two or more alleles in any frequency, it is statistically likely that the frequencies will change from one generation to the next purely due to random variations (how many unite to form zygotes, variation in numbers of progeny produced by carriers of the different alleles, variation in the numbers that survive to reproduce, etc). In short, the definition you provided.However, if you'll look back, I was specifically linking sampling error to the founder effect - because the founder effect relates to bottlenecks. I did not intend to imply that this was the ONLY type of sampling error. It was, however, the type referring to the question to which I was responding. Obviously that got confused somehow.

Damn. Just when I think you're starting to get a glimmer, you come up with something like this. What part of "severe reduction in population size" do you think relates to genetic drift? It doesn't matter what caused the reduction (although it's an interesting question in its own right) for the purposes of understanding genetic bottlenecks. In the elephant seal example you used, over-hunting by humans was the cause of the resultant bottleneck because it reduced the population (in this case the entire species) so severely as to edge it towards extinction.Now drift can happen (more readily) in a reduced population like this (in fact, it happens readily in any small population, whether bottlenecked or not), but drift is not a requirement for bottlenecks as you imply here. Basically, in a small population, allele frequencies are simultaneously affected by natural selection AND drift. A bottleneck, representing a population of extremely low size, can possibly provide a sort of temporary advantage to the mechanism of drift OVER ns (for instance, by allowing a deleterious allele to increase in frequency among other things). However, they are two separate mechanisms.

And the conflation is what, exactly? I referred to a process that caused an outcome. This is pretty standard English. Substitute "results in", or "creates" or any other term in place of "causes". Alternatively, replace "process" with "mechanism", "event", "phenomenon", whatever trips your trigger. Somehow you're the only one who didn't understand the really simple English "process yields result" in that sentence. What the heck are you talking about, Hoot? No one said anything about over-hunting being an "evolutionary process" (except you, right above). The process we're talking about is the process that caused the bottleneck. In other words, the event or sequence of events or whatever that reduced the population to the point of bottleneck.Secondly, what the heck are you talking about with "allelic equilibrium"? Are you still fixated on H-W equilibrium? Haven't at least three different people, including myself, told you that H-W is a theoretical construct that has no actual counterpart in the "real world"? It's a valuable construct in its place, but hasn't got much if anything to do with the topic of conversation. Get over it. So what? All you've done is shown a graph of genetic drift. We KNOW genetic drift occurs. It occurs in any bounded population, but its effects are most noticeable (and important) in small populations. Read this very carefully: GENETIC DRIFT DOESN'T LEAD TO BOTTLENECKS. It can lead to increased homozygosity (as rare alleles get eliminated by chance). It can work in concert with or counter to natural selection. Usually it just gets buried. Your graph doesn't show any correlation with a bottleneck (except that the population was pretty small). It certainly doesn't support your contention that drift causes bottlenecks. What part of my explanation didn't you understand, Hoot? Bottlenecks don't "act" at all. They are a RESULT, for the hundredth time. "Bottleneck" is a descriptive term for severe reduction in population size. Obviously, the reduced population is going to have only a sample of the alleles present in the source population. In fact, that's one of the major pieces of evidence that a population has historically been through a bottleneck - increased homozygosity caused by inbreeding depression, drift, and a number of other things. Bottlenecks don't "affect drift". They don't "affect" anything, because they are not a mechanism - they're (all together now) a result. Ya know, I was going to take a stab at responding to this paragraph. But the more I looked at it, the less sense it made. Sorry Hoot, but this looks like pure jibberish to me. Maybe it's my lack of understanding of biology. After all, you're talking to a guy whose idea of biology is spending his days playing frisbee on a tropical beach while waiting for the nighttime arrivada of several thousand olive ridleys so he can collect more data for a nesting behavior study. Not a REAL biologist by any stretch.Edited by Quetzal, : No reason given.

You're apparently still confused, somehow. Drift is NOT a "process stemming from a sampling problem caused by a bottleneck", etc. Drift is the random shifting of alleles in a population from one generation to the next due to chance. Frequency can shift "up", it can shift "down", and it randomly do the opposite in subsequent generations, etc. It can eliminate an allele over generations, or it can increase the frequency of a rare allele to fixation. Drift has been called a "drunkard's walk" because there is no predicting which way things are going to go: there's no selection filter to "push" things in one direction or another.Populations reduced to the point of genetic bottleneck MAY be MORE susceptible to drift than the source or original population. It's statistically more likely that drift will have observable effects in this situation. However, there is no actual requirement as you claim for drift to have any effect all on allele frequencies.As for anyone seeing either bottlenecks or the founder effect as "conditional types of random genetic drift", sorry but you're going to have to provide a literature citation on that one. This is simply inconsistent with the way the terms are used and understood. Increased susceptibility to drift may be one of the results of a severe reduction in population size (i.e., bottleneck), but doesn't cause the reduction. As to the founder effect, it is actually an extreme form of bottleneck (sometimes termed "individual speciation", although that's getting a bit esoteric). No, the KT impact was an "event" that "resulted in" a mass extinction (like that formulation better?). In some species, this "event" may also have resulted in a genetic bottleneck. Of course, we have no way of knowing since none of the species alive at the time are alive today (although there is a on-going disagreement over whether dinosaurs are in fact only pseudoextinct, since at least some of their genetics "lives on" in birds). My personal opinion is that the KT event undoubtedly caused bottlenecks in the few species that both survived the event and its aftermath. Possibly one of the geneticists or paleontologists on the forum can clear this point up. I need to get it straight? Too funny. One of us certainly does... Right. What's your point? How does what you wrote here rebut or even call into question anything I've written on this subject to date? Hell, I told you what the graph represented (to correct your earlier mis-representation). Here it is again, in case you missed it the first time: Repeating the graph doesn't help your case in any way.Moving on... You did, Hoot. Consistently. Here's an example: Don't get no clearer than that Hoot. Unless, of course, you're going to claim that's not what you meant, even tho' that's what you said. And the difference is what, exactly? The (or rather one possible) result of a severe reduction in population size is a genetic bottleneck. The "bottleneck effects" if I understand what you're referring to are things like inbreeding depression, increases in homozygosity/loss of genetic variability, increased susceptibility to drift, etc. that occur in severely reduced populations. None of these are "causes" in the standard way we talk about causes. They are further phenomena that occur in populations which survive extreme reductions in size (i.e., that have passed through a genetic bottleneck). They are not "caused" by "bottlenecking" (using the term you invented) - except in the most trivial sense that each of these phenomena occur after the population reduction we refer to as a bottleneck. I can't make it any simpler for you. I strongly recommend you go back to the texts or articles you originally got these strange ideas from, and re-read them in light of what you've been told on this thread. I think you'll find the texts are more consistent with what we're telling you that what you think you read the first time around. Naw, I'm fine with it. There are LOTS of things in biology I don't get. Anything to do with differential equations, for instance.However, I might suggest that rather than being a pissant, you might want to consider that I'm being straight up with you: as I said, I didn't understand what you wrote. It really did look like gobbledygook to me. So maybe if you try and make it less jargon-intensive and explain what it was you were trying to convey in simpler terms, I might even agree with you (who knows?). As it stands, your apparent inability to explain what you wrote leaves me with my original thought: it don't mean nothin'. Try again?