The End of Evolution By Means of Natural Selection

In the real world, alleles are usually only in competition with one other allele. For instance, an allele for fur colour that provides better camouflage than the existing predominant fur colour of the species is only in competition with the existing predominant allele. Thus most often, one allele replaces another. This also works in reverse for deleterious alleles. Add to this consideration of the fact that the vast majority of allele variations confer neither a beneficial or a deleterious effect on their recipients, and what you have is an extremely variable species. Natural selection does not eliminate variation.Next consider the environment - it is not static. Also, many species are migratory. The survival value of an allele is not an absolute - it is relative to the changing environment(s). Furthermore, the genetic environment is not static - all species around our hypothetical focus species are also evolving, changing the survival value of alleles in our focus species. Natural selection does not eliminate variation.

What you say is technically correct, but I was specifically talking about alleles and not traits, which as you say can have more complex ontogenies. However, if we shift the focus to traits the same basic principle holds. It will almost always be a case of the emergent trait versus the predominant trait when one is more beneficial than the other.Traits that have several predominant varieties (such as eye colour) are almost always neutral in terms of natural selection (note that I am not introducing sexual selection into this topic). When a trait is subject to natural selection it's variability is necessarily highly restricted. You can't have massive leaps in genetic space and expect to land in a favourable position, the odds are far too high.Traits that are the product of complex gene interractions are often found in behaviour. Imagine the number of possible (however apparently absurd) behaviours there are for a situation such as being confronted by a predator. Compare that with the number of behaviours that would be as or more beneficial than the predominant trait. This rough calculation gives you an idea of the extent to which such complex traits are constrained.Whether talking about alleles or complex traits the situation remains one versus one in most cases, as long as you only compare like with like (ie traits with traits, alleles with alleles). I will search for exceptions to this to falsify my point. If I find I'm utterly wrong I will reply and tell you.

In all concrete observed instances of speciation by natural selection that I can find (though there aren't that many), the speciation event has centred around a single trait or class of traits (eg bird song). I cannot find an example of a single allele being naturally selected directly resulting in a speciation event. That's not to say it doesn't happen, but direct observation of this phenomenon is clearly problematic.In examples such as ring species, natural selection is often inferred to be the cause of the initial reproductive isolation of two previously interbreeding populations, but not directly observed.There are of course many examples where natural selection acts on a single gene (eg the sickel cell anaemia gene that protects against malaria), but speciation has not yet been properly observed in such cases.So, from what I can find it seems as though natural selection does usually select one trait or allele in favour of one competing trait or allele (or lack thereof).If you can find an example that bucks that trend i would be very happy to hear it. Unlike many people in these forums I have no problem with being proven wrong - as that is exactly how science advances (not that I have over inflated ideas about the importance of these forums of course).Best regards,richdix.