One of the assumptions of the Hardy-Weinberg model is that we have a large population, and the reason is that in a large population, chance events are not going to have much of an impact on allele frequencies. If the population is small, say 10 to 20 individuals, or even 50 to 100, or even a few hundred individuals, it's... much more likely that just by chance the allele frequencies are going to change over time just based on which alleles happen to appear in the few individuals that we have in the population. Some of you may have taken the biodiversity lab and have participated in a population genetic simulation.
in which you tossed a coin to simulate the outcome of meiosis and then simulated random mating without selection and then did it again with selection and in that simulation the population is small you know, maybe a couple dozen students at most And so sometimes, just by chance, you know, when you toss the coin, more coins come up heads, sometimes more coins come up tails. And you're not going to always get 50-50, you know, sometimes those coins are going to be skewed quite far one way or the other. And so, in a given generation, because there's so few individuals in the population, The frequency of one allele or the other can fluctuate up or down quite significantly. But if you were to do that simulation with, say, a thousand students, you would be tossing hundreds of coins each generation, and it would be unlikely that the ratio of those tosses would deviate significantly from the expected 50-50 ratio.
given the large sample of coin tosses. So in a large population, allele frequencies are not going to fluctuate based on what the alleles do in meiosis or what sperm fertilizes what egg. We've got a big enough sample in a large population that chance is not going to impact the frequency over time. Now, sometimes real populations are small.
And in that case, chance has an effect. And when chance has an effect like that, that process is genetic drift. And we can refer to this also as what's sometimes called a random walk.
And if the allele is neutral in its effect, if there's no selection acting on it, if it doesn't affect fertility, It's going to fluctuate randomly. And so given enough time, the allele will eventually be either eliminated or fixed. It's going to go one way or the other.
And you may have seen this in your simulation in the biodiversity lab. You may not have gone to fixation or elimination, but you may have seen it going in that direction. kind of an analogy of a drunk walking down a railroad platform.
He's so far gone that he can't correct the direction he's walking. His next step is basically random, other than that he's more or less going forward, but he's going side to side, more or less randomly. And you know, eventually... He's going to fall off onto the tracks either to the left or to the right.
You see, they're going to be fixed or the allele is going to be lost. Now, you know, it may depend on how wide the platform is or how far he has to walk. If the platform is narrow, he's not going to get very far. If the platform is wide, he might have to walk a ways.
But eventually, given enough time... that a leo if the population is small is going to go one way or the other and so the put the population is really small that would be like a very narrow platform that he's walking down and it's not gonna take long before it goes off one way or the other. If the population is a little bit larger, he might have some leeway.
He can sway back and forth one way or the other, but eventually he may, 100 yards down the platform, he's going to stumble off one direction or the other. So it'll occur quickly in a small population, narrow platform, or more slowly in a large population. Also, the probability of elimination or fixation is dependent on the current allele frequency.
We might, as a starting point, think of an allele frequency of 50%. We have 50 of one allele and 50 of another allele, but that's probably not usually the case. We might have 95 of one allele and 5% of another allele. And if that's the case, then obviously the allele that is at 95% is more likely to be fixed, and the allele that is at 5% is more likely to be eliminated. Although, if the population, again, is small, the drunk is starting out on the left side of the platform, he's more likely to go off the left if it's narrow.
But, if it's a small population, if it's a very narrow platform, just by chance he may stumble a few times to the right, the coins may come up tails instead of heads, and he may end up very quickly on the right side, and genetic drift will cause the population to go the other direction. But the probability is, if you start out with that frequency, that's the direction you're going to go. Deem. I think we have not talked about deems, but a deem...
is a small sub-population. We've talked about populations a lot, but a deem is a subdivision of a population, a smaller group of organisms that are very likely to interbreed with each other, and more likely to breed with each other than with the larger population. population outside.
So this deem is a smaller group that has some of the properties of a population, and there's more likelihood of breeding within the deem than outside of the deem, and consequently that within the deem, within that small subpopulation, There are, there's a possibility of allele frequencies being affected differently within the deem than in the population at large.