Evolution Requires Reduction in Genetic Diversity

I think this describes frequency dependent selection, not heterozygote advantage. I have been trying to figure out how that would work, but I'm not sure what you had in mind.For heterozygote advantage the allele frequency is dependent on the fitness of the homozygotes. If the relative fitness of both homozygotes is equal, the frequency of each allele will be 0.50. If the fitness of one homozygote is higher than the other, the allele frequency will be skewed toward the allele in the more favored homozygote.Another point regarding heterozygote advantage is that the selection pressure against each of the homozygotes (and therefore the relative fitness) comes from different environmental factors. So in the case of sickel cell, the wild-type homozygote is only disfavored in the presence of malaria pressure, so in the absence of malaria there is no fitness disadvantage and there will be selection against the sickel cell allele. Another example is warfarin resistance in rats. Susceptible rats (SS) will be killed by warfarin but homozygote resistant (RR) are vitamin K deficient. So, heterzygotes (SR) have an advantage over either homozygotes. However, in the absence of warfarin or if a sufficient supply of vitamin K can be obtained, the advantage is negated.Anyway, negative frequency dependent selection (when fitness increases as a genotype gets rarer) is another way, in addition to heterozygote advantage) that diversity can be maintained or even increased.HBD

Perhaps I have misunderstood the terminology, but it seems to me that heterozygote advantage must describe a case where the heterozygote has an advantage over both homozygotes, which is exactly the situation with sickle-cell in malarial regions. If that is not correct I think the term needs to be carefully explained when introduced because it is not obvious. As I have said, an advantage to the heterozygote over both homozygotes will have the consequence of what you called frequency-dependent selection although I would think that a more general term.I have thought a little more on the matter and it does seem that in this case diversity measured by the proportion of heterozygotes would increase, although diversity measured by the number of different alleles would not. Although this could also be said for selection in general, it is more marked in cases like this.Edited by PaulK, : No reason given.

Sorry, you got the description of heterozygote advantage right and sickel-cell is an example. I was commenting only on the part I quoted... which doesn't fit heterozygote advantage It will push allele frequency to an equilibrium point that is dependent on the relative fitness of the homozygotes rather than Hardy-Weinberg, so it could increase allelic diversity until that equilibrium is reached.I am sure this could be modeled mathematically but I would have to look it up (I will if anyone is interested)HBD

quote:

---

which doesn't fit heterozygote advantage

---

For cases like sickle-cell where selection maintains a balance between alleles it must be the case. Maybe there are cases where it isn't true, and it may depend on relative fitnesses but I'm pretty sure that it is for sickle-cell.

Perhaps I am looking at a wording problem that you find insignificant, but selection does not replace anything. Selection is operates on traits already present and favors animals with trait over animals with variations not including the trait. It does tend to subtract diversity other than in those special cases you discuss.But replacing traits? No, that requires selection in combination with mutation or other forces. Maybe 'displace' would be a better verb.

HBD got it right: alleles for two traits already available, one replaces the other. Trust you to misread me as usual.

As you say, it may be a wording issue. I think that when HBD says "Selection replaces one trait with another..." that he means changing things like beak size or fur color or melanic coloration by drawing upon variation already present in the genome.

Yea, I struggled a little with the wording but decided not to make an issue of it at this time. I didn't think she had meant replace as in "substitutes" but rather as in "becomes more prevalent," so I just ignored it. The important thing at this time was just making the chart clear. Right. In reality, selection acts on variation within a population and shifts the mean of that variation to a different optimum. Strong selection (especially when coupled with drift) can push a trait or an allele to fixation, but it is typically beneficial to the population to maintain some variation. Since environmental conditions and therefore selection pressures are rarely static, maintaining variation allows the population to respond to changing selection pressures. In addition, traits that are controlled by a single locus are more likely to go to fixation because of selection than polygenic (influenced by multiple genes) traits. Since the majority of traits are plolygenic, fixation of a trait is difficult.If you are familiar with the concept of an adaptive landscape (or fitness landscape), I think they are great for helping to visualize the concept of selection. The thing that an be confusing is that it can be easy to look at them as representing spacial relationships, but they don't. These "maps" have nothing to do with how close individuals are to one another but how combinations of alleles or traits are favored by selection and how those combinations change over time.Below is a dynamic landscape where the optimum of a trait is shifting back and forth. The X and Y axis can be the combination of alleles or some measure of a trait. The Z axis is fitness with up being increased fitness. This could represent the peppered moths whose color morphs shift back and forth depending on environmental conditions. Notice how the variation never goes away, it just shifts to a new mean when there is a new optimum.The adaptive landscape below shows a situation where there are multiple fitness peaks and the population evolves up only one of those peaks. Notice a smaller population begins to climb the peak on the left but disappears? This could be due to genetic drift but gene flow is probably involved as well. As one subset of the population climbs higher up on the fitness peak, it's combination of alleles are more fit and would be "replacing" or more appropriately, displacing alleles at those lower fitness levels by interbreeding.I wish I could stop the .gif image at a specific spot so you can see exactly the point I am talking about, but watch the image as the population splits into two groups - one on the left and one on the right. (Again, this is not a spatial representation - it can be hard to keep that in mind). Now if an individual high up on the right hand peak mates with an individual in the group on the left, a portion of the offspring will have this more fit combination of alleles and so will become part of the group on the right. In this way, the combination of alleles or traits in the group on the left is being displaced.Theoretically, the groups could climb the different peaks and so would be differentiated (sympatric speciation), but as long as there is gene flow between those groups it is unlikely because of the situation I described above.These adaptive landscape .gifs also visually show selection acting to reduce variation WITHIN the population. You can especially see it in the dynamic model when selection is released, the variation in the population increases until it begins moving up the adaptive peak.HBD--------------
Author of .gif fitness landscape maps: "Visualization of a population evolving in a dynamic fitness landscape" by Randy Olson and Bjrn ‘stman - Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons -

As I wrote, 'replace' seems to be a bad way to express the selection because it does not indicate why the result of selection can be to reduce diversity. Literally replacing one trait with another would simply preserve diversity. In the population undergoing selection, we may have multiple traits existing and selection tends to increase the appearance of one of the traits at the expense of one or more others. Replace, in my view does not get anywhere close to expressing that. Displace seems to work better or would any wording expressing a trade off.More to the point, what is your problem, Faith? I thought your wording was off, so I asked a question. Here is the beginning of HBD's post: I understand that you have a problem in particular with my making corrections to what you are saying, but you probably should just get over that.Edited by NoNukes, : No reason given.

Drift shows up as random fluctuations in allele frequency. In order to see the effects of drift, you would need to look at several generations. Since most traits are polygenic (that is, controlled by more than one gene and having a continuous distribution), it would be difficult to see changing allele frequencies over a few generations without doing genetic samples. You could easily see the effect with an annual flower in which flower color is controlled by a single gene. What you would see is a fluctuating number of colored flowers - one year there would be more blue flowers than white; the next year white would be more prevalent, ect. Well, in theory a subpopulation could form within a larger parental population, but how often that actually happens is debatable. Moreover, drift would probably not be the factor involved; rather it would be selection and mutation. The easiest way to picture sampling error is flipping a coin. A coin could land either heads or tails so there is a 50% chance that it will land on heads. If we flip it 10 times in a row we would expect that it would land on heads 5 times and tails 5 times. However, there is in fact only a 25% chance that if you flip a coin 10 times you will get 5 heads. There is actually a 21% chance that you will get either 4 or 6 heads. This would be a 10% error from the expected value. If you flip the coin 100 times, there is only an 8% chance that you will get 50 heads and only a 1% chance that you will get only 40 or 60 heads (10% error).The effect of this sampling error on organisms is that the offspring of a generation vary in allele frequency from the expected frequencies. The effect is more pronounced in small populations than in large populations.Below are binominal distributions for n=10, n=100 and n=1000. Notice that as sample size increases the probability of having exactly 50% success rate (heads) decreases from ~25% (n=10) to ~8% (n=100) to ~2.5% (n=1000). Also the range of error decreases from 20% (n=10) to 10% (n=100) to 5% (n=1000). Another source of drift can be when individuals are removed from a population by serendipitous events such as natural disasters. Imagine a population of birds on an island when a hurricane strikes and kills half of them. Chances are that allele frequency after the event will be different from the frequencies before. This is a type of drift, yes, and often what the common perception of drift is. However, it is not really where the majority of drift comes from. How could there not be this sampling error considering how the reproductive system works?Here is an image of experimental populations of Tribolium (flour beetles) that shows the result of drift. If the beetle is homozygous for the wild-type allele (b⁺), it is black. If it homozygous for the mutant allele (b), it is red. The heterozygotes are brown and can be easily identified.Populations were founded with 10, 20, 50 and 100 heterozygote individuals (5, 10, 25 and 50 breeding pairs) and each generation was maintained at that population level by randomly selecting the next generation. Each population size had 12 replicate populations and 240 individuals from each generation were rated as bb, bb⁺ or b⁺b⁺. The frequency of the wild-type allele (b⁺) for each population and each generation are plotted on the chart.Notice for the population with n=10, that 6 populations fixed the b⁺ allele (went to 100%) and 1 population fixed for the b allele (went to 0%). But for the n=100 no populations fix for either allele. Notice also that allele frequencies bounce around, especially in the smaller populations. Also notice the trend in increase of the wild-type allele. This appears to be due to selection or a disadvantage to the mutant allele.I hope that helps to better explain how drift works.Also, if you haven't already, check out Message 983 about selection.HBD

You've put up a lot of information to digest but for a while I'm not going to have the time for it. And most of it looks to be peripheral at best to the argument I've been pursuing. The Fitness Map is just going to drive me crazy though.

Understandable. As you can see, I am not able to post consistently either. It is peripheral in that it is background information. The argument you have been pursuing looks nothing like what we see when we study the genetic diversity of natural populations. But there is just no simple answer to your argument without explaining how we actually study diversity and what the results of those studies imply. One of the creationist claims is that we are all working with the same data, just interpreting it differently. But we are NOT working with the same data here - not at all. I am trying to explain what the data is that we work with; which tells a completely different story than the one you are arguing.So, do we all work with the same data and just interpret it different... or are creationists happy to only consider a tiny piece of the information available and draw conclusions from just that? Why? Because it looks like it represents spacial distribution? That's understandable. Think of it as a genetic diversity map. Each individual on the map represents a particular genotype, a particular combination of alleles or traits. The landscape is how that genotype (or phenotype) interacts with the environment. The individuals higher up on the landscape will bear more offspring than those lower on the landscape because of this interaction between genotype (phenotype) and environment.Also remember that these models are computer simulations. They are meant to be illustrations of a concept, not representations of some specific data.I hope to be able to discuss some ways we actually measure diversity and how we analyze and use that data. To analyze total genetic diversity of a population (that is the total number of alleles at all loci) is impractical at best and essentially impossible with our current capabilities. So understanding how we actually measure and analyze diversity is important to this discussion. This background information is important to understanding how and why we make certain predictions about diversity and the signals we are looking for when we analyze it.Taking our time is not a problem.HBD

I realized what the wording was that I took issue with. It was that selection was dependent on whether the allele was "rare" or "common" which isn't quite right. Here is my description of heterozygote advantage again for clarity: So in the case where the fitness of both homozygotes is equal and the equilibrium frequency is 0.50, if the frequency of one allele falls to 0.49 (the other would be 0.51), selection would work to push the frequency back to 0.50. Which I guess you could describe as "rare" or "rarer," but I don't think it accurately describes the situation.Maybe Faith is right and I am a "pedantic nitpicker," but hey, that's what we scientists do .But I didn't comment so much as to say you were wrong, but to give a better and more complete description of how heterozygotic advantage works and why.HBD

I'm reviving this thread because a lot of what Faith is saying on the Evolution. We Have The Fossils. We Win. thread is just a duplication of what is here, and I'm going to say a bunch of things here to refer to rather than clog that thread up. And obviously we are nowhere near the end of the road when species keep appearing. Sorry, but it is necessary to continually point out when terms are misused or misunderstood. You still sound confused. Or is it just poorly worded: all the "genetic stuff" dies? Random distribution. A different distribution of alleles is sufficient to form a different variety but not a new species that is genetically incompatible -- the evidence is that they were compatible in the parent population.You have to have new alleles not in the parent population for genetic incompatibility to occur. Except that it isn't. You can't get blue eyes in a population that does not have blue eye alleles by dividing that population into smaller populations. But I can get blue eyes in a population that does not have blue eye alleles by mutation. If they were built-in then the distribution analysis would not work. Show me the evidence! WHAT you HAVE evidence? Anyone could come up with THAT, but it's ONLY ONE!I mean really, Faith. One piece of evidence is enough to show that it has happened. Biologists don't expect to find evidence like this often, but they do expect mutations to occur because we know they occur. More often than is necessary. What occasion? Mice living beside volcanic rock for thousands of years before a mutation occurs that allows some mice to survive within the lava fields is an "occasion"??? Really? No, it is a random mutation that occurs and the mice with the mutation were able to take advantage of an ecology the parent population could not use.On the other hand, if the allele were "built-in" to the genome then there is no reason to wait for it to occur, mice would move into the lava fields and presto chango the trait would "emerge" ...Which doesn't explain why two different populations have two different mutations.Yes your "hypothesis" has been refuted.

---

..... enough for now. As anyone can see it is the same argument being pushed on the other thread, and it would be better to discuss it here.Enjoy

Message 154Essentially you have framed this request so that it cannot be answered satisfactorily. First of all, what is "normal" genetics? I am pretty sure our understandings of "normal" genetics are quite different. By "normal" genetics do you mean the kind that leads to "genetic depletion?" Cause I don't think that is normal. And I am pretty sure the kinds of things we look at genetically when trying to unravel issues like this will not seem "normal" to you. Heck, we could barely get past the basics of selection, drift, mutation and migration; how could we move into the deeper water of developmental genetics?Secondly, I am pretty sure that what you have in mind for "structural change" is beyond the scope of what we observe in the course of "normal" genetic studies. Most of the examples we have from experimental studies are relatively minor compared to what we see in the fossil record, and I imagine to you it will be a matter of "it's still a fish; a bacterium; a plant, etc..." I fell like you are expecting the proverbial dog turning into a cat right before our very eyes; and that just doesn't happen. However, we have documented flies that have developed legs where their eyes should be, which I would say is a major structural change, but somehow I don't think that is what you have in mind either.What the examples we do have show is how these kind of changes can arise. They show how turning on and off different switches can affect development and how those effects can lead to significant structural changes. But by all means we have not unraveled but a small, small portion of the secrets of developmental biology. There is much work to be done, but we do have enough evidence to suggest that the kind of changes we see in the fossil record are not only possible but are also probable.I will try to post a couple examples this weekend, but I doubt you will be all that impressed.HBD