What exactly is natural selection and precisely where does it occur?

I'm not so much going to disagree and elaborate and get to a slightly different conclusion.It's certainly true that the individual events that add up to selection happen on an individual level. But selection needs to work in aggregate over the lifetimes of many individuals. For asexually reproducing creatures where the entire package of genes is passed on it may make sense to look at the whole package of genes. But in secually reproducing creatures where genes get mixed up and reshuffled in every reproductive event such an approach can't work.Since whatever is being selected must be somethng that can be inherited and must be something that we can find in many individuals, it makes more sense to look at genes. The view of a single gene working in isolation is certainly oversimplified but - for sexually producing organisms - it comes closer to what is actually going on.On the grand scale your selfish genes are in a temporary alliance. Your success contributes to the spread of most, maybe all, of them. But it isn't enough in itself - you're a transitory event, unlikely to be repeated. Your genes must prove themselves successful in many different alliances to spread through the entire population.The individual events that make up selection happen on the scale of individuals. But it is the sum of those events that really matters on an evolutionary scale.

And to be more strictly accurate, at the individual level selection works on phenotypes, not genotypes. Maternal effect genes affect the fitness of offspring whether they carry the gene or not. Developmental abnormalities can be lethal even if they are not genetically based.And it's not just selection. Before a gene spreads widely it is vulnerable to misfortune. An accident can easily remove a single individual. In some species the death of a parent can doom the young. A local disaster could eliminate a family. A beneficial trait typically only loads the dice slightly in favour of the individuals carrying it.

I think that the issue is that the question of where natural selection happens is ambiguous.If we want the actual events then those happen at the individual level where phenotypic advantages tilt the odds towards reproductive success.But trying to see the evolutionary effects of speciation by doing that is lik looking for a forest by examining each individual tree. A higher level, aggregated view is more appropriate for that case. And that view will be closer to a gene-oriented view.So either approach leads to a valid answer to the question. I'd tend to use the second approach because it is more directly relevant to evolution. But a purist might well object that it is an idealised, simplified view that ignores the details.

quote:

---

crashfrog wrote:
Preferential mate choice causes differential reproductive success; therefore, clearly, mate preference constitutes a selective force.

No, it's not. Natural selection operates on the changes of allele frequencies resulting from preferential mating. Preferential mating, in and of itself, is not what is “being selected for.” It is the result of it that opens the door to NS.

---

I think that your reply is poorly worded and could do with a rewrite - it isn't clear what you mean. Natural selection is the process that produces changes in allele frequencies. Preferential mating is an example of that as Crashfrog said, and as you seem to agree, despite the "No, it's not".However you are also wrong to state that preferential mating cannot be subject to selection. Mate choice is clearly linked to reproductive success and obviously could have an impact on the quantity and quality of offspring. Where it has an overall positive impact it certainly could be selected for.

To start with the most obvious case, luck is surely not an example of natural selection. Yes, it contributes to evolution (in the form of genetic drift) but it isn't an aspect of the phenotype, it isn't heritable - it's just stuff that happens. So I don't see any selective element in luck at all.

You may find Message 38 clears up some of your questions.In the case of The Handicap Principle there are competing selection pressures. There is negative selection from the handicap itself balanced against a positive sexual selection. Any species is likely to find a balance point between these two pressures.