5
$\begingroup$

I have a basic question about evolution, for which I never found an answer. I understand how evolution works if we focus on one specific organ or trait. With each generation, some organism is more likely to reproduce, so the trait that leads to success gets more frequent. My problem is understanding how all the traits can evolve simultaneously.

Looking back to our ancestors, we evolved many different kinds of adaptations (in unrelated areas like eyesight, kidney efficiency or a healthy fear of predators, digestion of certain nutrients, balanced walking etc.). Natural selection can only work with what's there, so mutations are important. But if mutations are rare, it seems unlikely that many (thousands of) properties of organisms get improved in a single generation. You'd need to get very lucky to randomly improve all the genes responsible for it.

So, if we look back to our ancestors again, did they just improve on a single or a handful of traits in each generation? In that case we would need hundreds of generations before we have improved somewhat on each "front". By then, the other properties could "drift" back to a not-so-advantageous version. (This was supposing that each generation improves on a random trait, as opposed to long sequences of generations each improving on the same). Or were there always some super lucky organisms that randomly got an improvement on almost all different traits? But just having some of those super lucky ones is not enough. The good traits don't guarantee success, just improve the odds. So we'd need many very lucky ones in each generation.

Has anyone calculated or simulated how the adaptation for many different traits can happen simultaneously?

Improve this question
$\endgroup$

3 Answers 3

Reset to default
6
$\begingroup$

Has anyone calculated or simulated how the adaptation for many different traits can happen simultaneously?

There are a lots of studies on the subject but I don't fully understand what is your issue. So I'll try to give some words hoping that helps a bit but it is possible that I'll totally miss the point you want to make.

The mutation rate in humans is around $1.25 \cdot 10^{-8}$ per base pair per individual per generation. The human genome contains about 3400 Mpb, meaning that you probably carry $1.25 \cdot 10^{-8} * 3.4*10^9 = 42.5$ mutations that none of your parents had in their genome. Now if you consider the population size, you'll see that a fairly big amount of mutations occurs each generation in the human population. So mutations are not that rare! It might unfair to take human as an example though as human's population size is very big. Anyway, the great majority of mutations are deleterious. In consequence, you are right, probably not many beneficial mutations will occur in a given population at one given generation and especially not in the same individual (except if you think of viruses!).

Now, depending on how beneficial is the mutation you are considering (selection coefficient), depending on its dominance, depending on the (effective) population size, your new mutations might take quite a bit of generations before reaching fixation in the population. In this time other mutations might occur. And then, the question is very interesting. Imagine a population of asexual individuals where two beneficial mutations exist (at two different loci [=position on a chromosome]). Because there is no recombination (by definition of what is an asexual population) never these two mutations will be found in the same individuals and in consequence if one mutation reaches fixation, it necessarily mean that the other mutation would have disappear. In sexual population it is different. Recombination allow that the two mutations, once they reaches a sufficient frequency will probably be found in the same organism and therefore, the two mutations can reach fixation.

Hope that helped a bit

Improve this answer
$\endgroup$
5
  • 2
    $\begingroup$ Thanks. I think sexual recombination is the key to the answer to my question. It would allow the population to mutate in many different directions in parallel and then "merge" the successful mutations together later on. Very interesting! I never realized that recombination is so crucial in this regard. $\endgroup$
    – isarandi
    Apr 12, 2014 at 14:35
  • $\begingroup$ @quorilla You might want to go for few lectures concerning the concept of "Muller's ratchet". The absence/presence of recombination is a very interesting subject. For example, it is the absence of recombination between the chromosomes X and Y that causes the Y chromosome to shrink. $\endgroup$
    – Remi.b
    Apr 12, 2014 at 14:42
  • $\begingroup$ @Remi.b By "mutation rate" do you refer to any kind of mutation, or only nucleotide substitution? So far, I haven't encountered the rate for genomic rearrangements (duplications, deletions...). $\endgroup$
    – user38945
    Feb 21, 2019 at 18:42
  • $\begingroup$ Well, it depends which study you look at (consider finding some numbers on bionumbers). I would personally tend to think that the standard error for this kind of study is large enough so that whether or not we consider genomic rearrangements is negligible in the estimate as those mutations are relatively rare. That is just my opinion though.You might find some estimates of the relative frequencies of the different types of mutations in the work of Sharp $\endgroup$
    – Remi.b
    Feb 21, 2019 at 18:53
  • $\begingroup$ The question about the "improbability of all traits" has a hidden assumption: that traits have had a given function since they appeared. This is not true however. Hemocyanin is an oxygen carrier today, but in the past protected anaerobic life forms. Feathers allow flight today, but they probably began as "tendrils" that allowed insulation, only to later "branch" and improve this function. $\endgroup$
    – user38945
    Apr 12, 2019 at 10:45
2
$\begingroup$

I think the main issue you have is in this paragraph:

But if mutations are rare, it seems unlikely that many (thousands of) properties of organisms get improved in a single generation. You'd need to get very lucky to randomly improve all the genes responsible for it.

There are two basic misconceptions there:

  1. Mutations are not rare. They're actually very common. One of the cells in your body probably just developed one while you were reading this answer.

  2. It is indeed unlikely that thousands of properties get "improved" in a single generation. That is not how evolution works. You need to remember that species evolve, not individuals. So, say I have been selected for because I run faster and my mate has been selected for because she's smarter, our child might incorporate both characteristics. Now extrapolate that to all individuals of a species and add a few million years and you get both intelligence and speed selected for.

Finally, luck has very little to do with it. We're talking about a very slow process that works over millions of years. Evolution does not occur over single generations, the mutations driving it might but not the evolutionary process by which these mutations become incorporated into the genome of the population at large.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Thanks. The key sentence is "So, say I have been selected for because I run faster and my mate has been selected for because she's smarter, our child might incorporate both characteristics." This way we receive the good properties from an exponential number of successful ancestors in the past, each of which might have mutated in different advantageous ways. $\endgroup$
    – isarandi
    Apr 12, 2014 at 17:14
1
$\begingroup$

The basic idea is that mutations improving different traits can occur in different individuals, and then be brought together by recombination. So the frequency of recombination roughly sets the maximum rate at which the population can acquire good mutations. Nick Barton and I calculated it, and found that a population can gain about one good mutation per chromosome per generation: http://www.plosgenetics.org/article/info%3Adoi%2F10.1371%2Fjournal.pgen.1002740#pgen-1002740-g008

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .