Is evolution deterministic or chaotic?

[madness]

I don't know much about biology, my background is in maths and physics. Hopefully someone else here knows it better than me. Anyway, I was wondering if evolution would go the same way if it were repeated again with the same (or very nearly the same) initial conditions. I'm not sure how much evolution has been studied in the context of dynamical systems, but I can imagine many points of (quasi) stable equilibrium which life could have settled into, meaning it would be different every time. On the other hand, I've heard that the human type of eye is known to have evolved independently several times, which might imply that we would get basically the same life arising given the same initial conditions. Does anyone know about this? Is it an open question or has it been answered? Thanks.

 

[Borek]

I don't think HUMAN type eye evolved several times, I seem to remember EYE evolved on several occasions.

 

[madness]

Well I've heard that several different types of eye have evolved independently (e.g octopus' eye is different from a human eye), but I've also heard that at least one type of eye evolved the same way independently several times.

 

[Borek]

OK, but that's not the same as "human eye" evolving independently several times.

Also note that same type of eye doesn't necesarilly mean identical eye. There are situations when similar structures evolved in a very different way. Lame example (but I can't think of a better one) - rhino has a horn and elk has a horn, but one is made of bone, the other is mostly keratin.

 

[Kracatoan]

Evolution would never follow exactly the same course if repeated again due to the uncertainty of every single event, so even if you started with the same conditions, what you would end up with would be different. However, it is not unlikely that certain favourable traits such as the example already given, the eye, would turn up again because they are one of the best tools for the job.

 

[russ_watters]

Rather than look at one feature of one species, just consider multiple species with the same ancestor and the answer is plain: with different starting conditions, different species will evolve differently. Really, that's most of the point of evolution.

 

[madness]

Rather than look at one feature of one species, just consider multiple species with the same ancestor and the answer is plain: with different starting conditions, different species will evolve differently. Really, that's most of the point of evolution.

But the individual species do not evolve independently, they evolve together and in response to each other. Given the same initial conditions, it's still conceivable that the same set of species would evolve together.

Evolution would never follow exactly the same course if repeated again due to the uncertainty of every single event, so even if you started with the same conditions, what you would end up with would be different. However, it is not unlikely that certain favourable traits such as the example already given, the eye, would turn up again because they are one of the best tools for the job.

When dealing with large numbers, for example in statistical mechanics, uncertainties often cancel out on average. This is why I asked if evolution is deterministic (ie stable with the initial conditions) or chaotic. You seem to start from the assumption that it's chaotic.

 

[madness]

I just did a quick google search for "evolution deterministic" and found two relevant papers: http://www.sciencedaily.com/releases/2007/11/071119123929.htm, http://mmbr.asm.org/cgi/content/full/65/1/151. The first seems to be basically another example like the eye, whereas the second is more technical and talks about modelling evolution as stochastic or deterministic. A quote from the second - "In deterministic dynamics, which applies only in very large populations of infected cells, if one knows the initial mutant frequency and has the appropriate equations, one can, in principle, predict the mutant frequency at later times with arbitrary precision."

 

[DanP]

I don't know much about biology, my background is in maths and physics. Hopefully someone else here knows it better than me. Anyway, I was wondering if evolution would go the same way if it were repeated again with the same (or very nearly the same) initial conditions.

No. The responses are both to changes in environment, and genetic drift. Even if you would by magic create a linear evolutionary medium, genetic drift will play a important role. This process is random.

 

[DanP]

A quote from the second - "In deterministic dynamics, which applies only in very large populations of infected cells, if one knows the initial mutant frequency and has the appropriate equations, one can, in principle, predict the mutant frequency at later times with arbitrary precision."

Would the prediction of mutation frequency allow you any prediction of affected genes and hence phenotype ? No. Knowing the frequency will only give a image of how often genetic drift will occure, but will provide no information on what genes will be affected, and hence have a very poor value of predicting the phenotype.

 

[madness]

Would the prediction of mutation frequency allow you any prediction of affected genes and hence phenotype ? No. Knowing the frequency will only give a image of how often genetic drift will occure, but will provide no information on what genes will be affected, and hence have a very poor value of predicting the phenotype.

But with a large enough population the randomness of the genes affected will average out and the result may be predictable. This is a common concept in physics and is what allows us to model things like gases.

 

[Monique]

Historically this is not true, many factors have affected population dynamics and genetics and this is reinforced be the relative isolation of the different populations. Most notably are the differences in the skin color that clearly have not averaged out. With the current rate of globalization this, however, may change.

As for the question whether evolution would go the same way, clearly that would depend on the conditions during the time that a species evolves. Current life has adapted to the conditions it encountered and life has proven to be very versatile, there are halophiles, chemophiles, thermophiles, acidophiles, photophiles. We are still encountering new, strange species. The species you encounter depend on the conditions within they thrive.

 

[DanP]

But with a large enough population the randomness of the genes affected will average out and the result may be predictable. This is a common concept in physics and is what allows us to model things like gases.

I see your point, but it does only return us to the original question: is the system linear or chaotic ? This leaving aside any external conditions, i.e , evolution of toher species and environment.

 

[madness]

I see your point, but it does only return us to the original question: is the system linear or chaotic ? This leaving aside any external conditions, i.e , evolution of toher species and environment.

Certainly, I wasn't arguing either way, just explaining why I wasn't convinced by some of the answers given.
And to Monique - by average out I didn't mean become homogeneous, I meant that with large numbers random behaviour can become predictable. And yes, it would depend on the conditions but I'm assuming here that they are the same.

 

[Borek]

Bitozoa always evolve different, even if the conditions are identical...

 

[JerryClower]

Evolution is fueled mainly by natural selection. Natural selection states that organisms who are more fit for the environment around them will survive and produce favorable traits, while organisms who are less fit will die. When I say fit, I mean the probability of an organism to live long so that it can produce offspring to pass on its genes. Natural selection is dependent on random mutations, or changes in an organism's DNA. Some mutations are randomized. Natural selection is itself is not random, because 10 out of 10 times only the fittest organisms will survive and thrive. So, natural selection itself is not random, but the force (mutations) that drives natural selection is random, putting evolution itself in a category that isn't random. Evolution gives organisms what is beneficial for their survival, and if it by chance gives it something that is harmful to its survival, it will probably die and the 'bad trait' will not be passed on and won't thrive through generations.

So technically, evolution in long terms is not random. If you could see a time line history of evolution, you would see that it stayed on a steady progression of beneficial mutations. But short term, evolution is random, because mutations are what causes evolution to happen. You got what I'm saying?

 

[Borek]

You got what I'm saying?

No.*

 

[madness]

Evolution is fueled mainly by natural selection. Natural selection states that organisms who are more fit for the environment around them will survive and produce favorable traits, while organisms who are less fit will die. When I say fit, I mean the probability of an organism to live long so that it can produce offspring to pass on its genes. Natural selection is dependent on random mutations, or changes in an organism's DNA. Some mutations are randomized. Natural selection is itself is not random, because 10 out of 10 times only the fittest organisms will survive and thrive. So, natural selection itself is not random, but the force (mutations) that drives natural selection is random, putting evolution itself in a category that isn't random. Evolution gives organisms what is beneficial for their survival, and if it by chance gives it something that is harmful to its survival, it will probably die and the 'bad trait' will not be passed on and won't thrive through generations.

So technically, evolution in long terms is not random. If you could see a time line history of evolution, you would see that it stayed on a steady progression of beneficial mutations. But short term, evolution is random, because mutations are what causes evolution to happen. You got what I'm saying?

Here you seem to be assuming that evolution is deterministic (linear), you haven't shown it. Perhaps a small random fluctuation in one species causes a response in another which sets off a chain of events which cause a completely different path of evolution. What "fittest" is fittest depends sensitively on all other species in the ecosystem.

 

[JerryClower]

Well I've heard that several different types of eye have evolved independently (e.g octopus' eye is different from a human eye),

Just because they are different eyes doesn't mean they evolved from each other or independently. And yes, you did HEAR that.

 

[madness]

I didn't say that just because they were different they evolved independently. In fact my whole point was that they might not have evolved independently but if they had then it suggested evolution was deterministic. And the link I gave earlier in the thread showed that another example (vulva in roundworms) had evolved to the same final result independently.

 

[Borek]

Question is how you define "same final result". Because of convergent evolution shape of the body of tuna and ichtiosaurus is very similar - would you call it "same final result", when structure underlying fin shapes is different? I believe there are many cases where similar organs/tentacles/whatever evolved in different animals and in different ways for a particular task - shape is very similar, working is very similar, but internal structure and tissues are different.