Can Evolution Be Measured?

Hello,This may be a tall order, but I am wondering if it is possible to somehow measure or quantify the evolutionary change that occurs in a population of animals whose life spans are too long for us to directly observe significant change? By significant, I mean an alteration of function, appearance, etc. In other words, given that we are able to observe change amongst organisms that have shorter life spans, and therefore have "high turnover", how can we measure change, at any level, amongst longer-living animals such as humans?I realise that most mutations will not propogate themselves, that much time can be required for changes to manifest themselves within a population (way longer than our lifetimes), and that words like "significant" and "longer" are subjective, but I'm curious to know how such a measuring process would occur, and if it's even possible!ThanksThis message has been edited by Gofor45, 07-19-2005 03:27 PMThis message has been edited by Gofor45, 07-19-2005 03:29 PM

Thread moved here from the Proposed New Topics forum.

Sure. From the study of short-lived organisms we're able to learn about the genetic precursors to the sort of change you're talking about.For instance, we can quantify:1) How many mutations per generation per individual per billion base pairs;
2) The way competing alleles wax and wane in frequency among the gene pool of a species experiencing no selection (the Hardy-Weinburg equation)and a few other things that I can't think of right now.Basically, we refine these models on sexual organisms with rapid generational times, and they do basically apply to everything else, including us. The specific genes that influence those traits are fairly well-understood, but now we get into selection. I'm not sure there's a way to quantify environment, except in the simplest biochemical examples, so predicting how selection will act on a certain trait is difficult. The broader our information about the human gene pool, the more we can take advantage of a shallow (in terms of time-depth) sample and spot trends in genetic change sooner.

quote:

---

This may be a tall order, but I am wondering if it is possible to somehow measure or quantify the evolutionary change that occurs in a population of animals whose life spans are too long for us to directly observe significant change?

---

Yes. As crashfrog pointed out, this is done through the analysis of allele frequencies. Pretend you have a population of organisms that is mating at random with respect to a given locus with alleles B and b and also that this population is large (so we can ignore the effects of inbreeding). That means that at fertilization,Genotypes---------Frequencies
BB-------------------p^2
Bb-------------------2pq
bb-------------------q^2Now, further suppose that each has a probability of survival into adulthood given byX_BB, X_Bb, and X_bbOnce they’re adults then, the frequencies will beGenotypes----------Frequencies
BB--------------------(p^2)X_BB
Bb--------------------2pqX_Bb
bb--------------------(q^2)X_bbNow, we need to make these frequencies add up to 1 or else any further calculation makes no sense, so we divide by a common factor--the mean fitness of the population (here a bolded letter denotes an average), which is:X = (p^2)X_BB + 2pqX_Bb + (q^2)X_bb = pX_B + qX_bGenotypes---------Frequencies
BB-------------------(p^2)(X_BB /X)
Bb-------------------2pq(X_Bb /X)
bb-------------------(q^2)(X_bb /X)Now, if all of the organisms that survive into adulthood reproduce, the frequency of the B allele in the next generation (p’) is given by:p’ = BB + Bb = (p^2)(X_BB /X) + ()2pq(X_Bb /X) = (p^2)(X_BB /X) + pq(X_Bb /X) = p ((pX_BB +qX_Bb)/X)However, we also know that pX_BB +qX_Bb is just the average fitness of the B alleles, so X_B = pX_BB +qX_Bb.So we can rewrite the equation as simply: p’ = p X_B / X This is an important result. It states that the changes in allele frequencies due to natural selection are only dependent on the relative fitness of some allele B to the mean fitness of the entire population. Thus, if the average fitness of the B allele is greater than that of the mean fitness of the population as a whole, the frequency of the B allele will increase in the next generation. If the fitness of the B allele is less than that of the mean fitness of the population as a whole, then the frequency of the B allele will decrease in the next generation (and the frequency of allele b will consequently increase, since q = 1 - p). The implications of this are important: it means that the fact that natural selection acts on a population does not mean that it will produce a population that is somehow invulnerable to extinction, or one that is perfectly adapted to its environment. The reason for this is that, despite the fact that natural selection is constantly moving a population upward on the adaptive landscape toward an adaptive peak, genetic drift may play a role in deciding which mountain on the adaptive landscape a population ascends. This equation is the reason that natural selection cannot make a population descend an adaptive peak into a valley between neighboring peaks and then up another mountain, even if the new adaptive peak is higher than the previous one.Now, suppose we want to analyze the change in allele frequencies:Dp = p’ - p = p (X_B / X) - p = p (X_B / X) - p (X/X) = p (X_B -X)/XNow, remember thatX = pX_B + qX_bNow, substituting this into the old equation, we get:Dp = p (X_B -X)/X = p (X_B - pX_B - qX_b)/X = p ((1-p)X_B -qX_b)/X = p(qX_B - qX_b)/X Dp = pq (X_B -X_b)/X This is another important result. It states that the change in allele frequencies is dependent on the difference in average fitness between the two alternative alleles, and that it also depends on neither of the two alleles being fixed. That is, neither q nor p can be equal to 1, or close to one. The closer each is to 0.5, the more effective natural selection is. Indeed, if q were 1, then p must necessarily be 0, and thus Dp = 0, and there is no change in allele frequencies and thus no evolution. On the other hand, if q and p are both close to 0.5 then pq is large, and thus Dp is large. This equation illustrates an important point about natural selection: it is directly dependent on the availability of variation to act on. The more variation there is to act on, the larger the change in allele frequencies.

---

"Chance is a minor ingredient in the Darwinian recipe, but the most important ingredient is cumulative selection which is quintessentially nonrandom."
--Richard Dawkins, The Blind Watchmaker

Just to note that evolution can be measured, and its unit is called a 'darwin'.