Estimating the Speed of Evolution: Theoretical Approaches & Limitations

[PIT2]

Is there any way to estimate the speed of evolution as it should* work by theory (which means: looking at natural selection and random mutations, as opposed to looking at the fossil record and dating them, reconstructing evolution as it has worked)?

Is this completely impossible as of yet?
Or has it been done and produced ridiculous results (like 300 billion years to get a fish)

* For instance, it would be permitted to look at the fossil record, but not to put any timestamps on them by carbon or any other dating methods. The timestamp should be put on the fossils by using the theory of evolution to predict how long such a organism would need to develop from previous stages.

 

[selfAdjoint]

Is there any way to estimate the speed of evolution as it should* work by theory (which means: looking at natural selection and random mutations, as opposed to looking at the fossil record and dating them, reconstructing evolution as it has worked)?

Is this completely impossible as of yet?
Or has it been done and produced ridiculous results (like 300 billion years to get a fish)

* For instance, it would be permitted to look at the fossil record, but not to put any timestamps on them by carbon or any other dating methods. The timestamp should be put on the fossils by using the theory of evolution to predict how long such a organism would need to develop from previous stages.

I don't think this is a well posed question. Evolution depends on many factors; the scale of the organisms from bacteria to whales, the presence or absence of rapid environmental changes such as comet strikes and climate changes, and the ability of the genome to meet environmental challenges by different expression of existing genes rather than waiting for some random gene change to meet the problem. The rate of evolution would depend on all these and many others, known and unknown, and evidently varies from multimillions of years for the coelecanth to an afternoon for bacteria exposed to artificial challenges in the laboratory.

 

[arildno]

Well, evolutionary fluctuations can be seen for every new generation of the species, but when averaging these fluctuations over some reasonable time period, there will usually be no discernible evolutionary trend over time.
Often, it is only when either the environment undergoes some permanent change (like,say, the introduction of a new predator or parasite), or, by random mutation, some particularly beneficial gene is introduced in the species' gene pool, that a discernible evolutionary trend is initiated.

This, at least, is how I understand the concepts behind the "punctuated equilibrium" idea.

Of the above, and in accord with SA's post, it should be seen that it is very difficult to give an answer to your question with some single time-scale.

 

[muadib2k]

You could probably estimate the rate of random mutations an organism undergoes, but when it comes to "measuring" natural selection you would hit a dead end (I would think). There are simply too many factors that would influence the manner by which natural selection would work. And while global-scale changes in the enviornment will no doubt influence evolution (and leave some sort of geological timestamp), little changes (i.e. a really dry climate for couple of years) may affect local populations just as much and not leave any long-lasting geological marks for you consider.

 

[PIT2]

Too bad, it would have been useful to see if the theory is consistent with the recostructed timeline.

 

[selfAdjoint]

One thing they do try to express as a clock rate is "neutral evolution", that is change without adaptation. If the part of the genome that is changing is indifferent to the change, then the environmental components of the evolution can be neglected. This is the basis for the "mitochondrial clock", for example, that is used to estimate the age of "mitochondrial eve", the apparent bottleneck individual early in the history of our species. But that depends on more factors than the originators of the idea thought of, so it has become a little contentious. Nothing is really simple in molecular biology.

 

[Danniel]

nope... and looking at the fossil record as an estimate of rates of evolution is very misleading. Even if a lineage appears to be pretty much the same along a great extent of time, there were more subtle oscilations, which could, but did not deviate much more mainly due to selection.

But maybe with populational genetics, specially in microbiology, something in this sense could be done (having in mind things like mutation rates, how many generations would usually, pessimistically and optimistically take to a mutation to fixate in a population of a certain size and things like that, but always with a certain margin to error, and with respect only to gene frequencies, disregarding morphology. Nothing like "well, there's this wing/eye/whatever adatation or trait, and it would only evolve in about 2 M.y., and 6 M.y after life began"

 

[DaveC426913]

Considering evolution is an emergent effect, not a cause - and that environmental changes (or other kinds) are the cause, your question is premature.

You first need to ask how often the environment *should* change so as to render the organism at a disadvantage such that its species will perish if it doesn't adapt. (This is just one example of evolution.)

How fast *should* a disease wipe out a species' food supply?
How fast *should* the climate evaporate its breeding ponds?
And most importantly: how many species *should* not be able to adapt fast enough, so that they go extinct?
etc. etc.

It becomes apparent that, not only is the question premature, but it is non-sensical.

 

[Rade]

Is there any way to estimate the speed of evolution as it should* work by theory (which means: looking at natural selection and random mutations, as opposed to looking at the fossil record and dating them, reconstructing evolution as it has worked)?[/i]

Good question. One can study chromosomes:
http://www.pubmedcentral.gov/articlerender.fcgi?artid=431793
One can study molecules:
http://abc.zoo.ox.ac.uk/Papers/tig06_clocks.pdf
Some alternatives to study of fossils that you seek--but be sure to study fossils also:
http://www.stanford.edu/~degusta/African%20Mammals/Vrba%20and%20DeGusta%202004.pdf