1

Fastest possible evolution rate?

What is the theoretically quickest evolution rate?

For example, I could calculate this as the number of offspring an E. coli colony can generate at once * (time frame/time per generation) * mutations per generation, and then assume every organism on earth had this same evolution rate.

Is there a generally accepted value, perhaps based on mass of organism?

Remi.b
  • 68,088
  • 11
  • 141
  • 234
yters
  • 285
  • 1
  • 7
  • 1
    Without an explanation for how you are defining "rate" I think this question is fairly meaningless and I suspect it might be rooted in a misunderstanding of the term "evolution". – Bryan Krause May 09 '17 at 21:14
  • @BryanKrause Fair point. I can describe what I mean in terms of a simplistic model. If we use a binary string to represent DNA, there is a space of finite length binary strings that we can explore through mutation, crossover and selection. From an initial starting population, and using biologically plausible parameters, what proportion of the space can be explored within a certain time delta? Is this rephrasing clear enough? If so, I'll replace my question. – yters May 11 '17 at 14:07
  • 1
    I sometimes like to share this comic in circumstances like this - I don't mean any offense, but I think it sums up the problem nicely if you approach it with an open mind. My point is that a simplistic model will never capture "evolution" because evolution is an incredibly broad theory encompassing many separate parts. For one, although it's reasonable to consider finite length binary strings, biology won't produce genomes of the same length, so you need to consider strings of varying lengths. You very quickly get to a combinatorial explosion. – Bryan Krause May 11 '17 at 16:08
  • And although people often talk about "random mutation" the mutations in biology aren't purely random. You often have duplication of entire segments creating multiple copies of the same protein which can then diverge from each other. Further, mutation is only one tiny part of evolution, you also have selection, which is incredibly dependent on environment, and non-selective changes like genetic drift. – Bryan Krause May 11 '17 at 16:10
  • @BryanKrause I'm not saying the model is comprehensive, but there surely must be a way to mathematically model evolution rate according to some definition or another in a field over 150 years old. All I was doing is offering one simple model to define rate, I'm sure someone has invented a better one. – yters May 12 '17 at 00:06
  • @BryanKrause I've updated my question with an explanation of how "rate" is defined. Can it be removed from hold? – yters May 12 '17 at 00:18
  • 1
    As Remi.b already pointed out, if you allow fitness and phenotype to vary sufficiently, you could state that the rate of evolution is effectively infinite: complete in one generation. So no, I think it is still a poorly defined question. – Bryan Krause May 12 '17 at 01:53
  • 1
    I happened to notice this previous question in the sidebar of a question unrelated to yours, but I thought it might be of some use to you. The question is on selection but the answers get a lot more broad. You won't find a precise answer to your question but it might help you as a starting point to how people have attempted to model relatively abstract evolutionary processes. Like I was alluding to with my cartoon link, though, this is an entire field of work. – Bryan Krause May 12 '17 at 21:55
  • @BryanKrause, is an infinite evolution rate scientifically plausible? I'm doubtful. Anyways, I've simplified my question per Remi.b's suggestion. Can it be taken off hold? – yters May 13 '17 at 15:52
  • 1
    See How fast is fast? Eco-evolutionary dynamics and rates of change in populations and phenotypes I think this what you were asking about – the gods from engineering Oct 14 '17 at 23:44

1 Answers1

3

Theoretically, it really depends how crazy you want your model to be! Also it depends what you mean by "evolution rate".

Let's consider a very simple and quite crazy model. Let's assume that by evolution rate you mean the difference between mean phenotype between one generation and the next. Let consider a trait (which heritability is $h_N = 1$) which takes only two values $x_1$ and $x_2$. In a given generation, half the individual carry the trait value $x_1$ and half the trait value $x_2$. If the relative fitness of $x_1$ and $x_2$ are 1 and 0 respectively, then the evolution rate is simply $| \frac{x_2 + x_1}{2} - x_1 |$. Allow $x_2$ and $x_1$ to be very big values and the rate of evolution will be very fast. In the most extreme case, $x_1 = \infty$, then the evolution rate is simply infinite.

Remi.b
  • 68,088
  • 11
  • 141
  • 234
  • The model needs to incorporate mutation so we don't need to assume the traits already exists in the population. It also helps if there is biological plausibility to the model. – yters May 09 '17 at 18:19
  • 1
    @yters How is there no biological plausibility to the model? For mutation, simply do not assume x1 and x2 are held by half of the population. – Bryan Krause May 09 '17 at 21:17
  • 2
    @yters The amount of biological plausibility is up to the one that answers. My point was to highlight the question made little sense for this reason. If you ask "In theory, ...." you have to get ready to get an answer that is not practical. Instead you could ask "What is the fastest fastest evolution ever observed?" But then and again, you will really need to define what you mean by "evolution speed". – Remi.b May 10 '17 at 01:42
  • @Remi.b, what is a good definition of "evolution speed"? I have no preference in the matter, I just want a good working definition that has evidence behind it. Not being a biologist, I do not know the correct technical terms that will get an answer, so asking me for the correct technical terms is not something I can provide. – yters May 13 '17 at 15:53
  • Talking about genetics, some may like to use the rate of neutral substitutions as a measure of the speed of evolution. Talking about phenotypes, some may like to use the Darwin (unit) to measure speed of evolution. Some may like to talk about speciation rate. In all of these cases, you cannot really ask for a theoretical limit without specifying exactly the basis for the model you wish to consider. You can ask though for observed evolution speed. Coming to phenotype, you'll have to specify whether plasticity counts as evolution or not. – Remi.b May 13 '17 at 16:29
  • I note that the question was edited/changed substantially since this answer was given, making it look off-topic now, although it was addressing the original question "What is the theoretically quickest evolution rate, that maintains biological plausibility?", not the current one, which asks about observed evolution. – the gods from engineering Oct 14 '17 at 23:54
  • 1
    Sure, we typically ask OP not to change their question after it has been answered. So, I edited the question back to what it was when I answered. If the OP has another question he/she must ask it on a different post. Thanks @Fizz for highlighting this issue. – Remi.b Oct 15 '17 at 00:15