To me, I question where "normal variance" (variance implies a difference, not sameness or normality) comes from. In my opinion all variance is a result of a mutation somewhere down the line.
Quite possibly, yeah. Might be more of a philosophical/definition argument than a practical one though. And the creationist argument seems to be more that for population A to evolve into population B, it requires population A to have a new mutation arise, and for that mutation to be beneficial. Which is clearly not the case, all it needs is for population A to not be identical. The other argument is that if population A changes into population B, that's because population B was contained within population A, and information has been stripped away from A, until we were left with B. That's the gist of their rubbish about kinds.
Which does remind me of a question I've asked a couple of times, but never had answered: As a creationist biologist, how do I work out which kind an animal is from? Easy enough to say for spiders, or snakes, but which kind did koalas come from? Bear kind? Or something else?
The gene that made melanin express itself less so we have lighter skin is a trait that's a variance from what it had been prior, a mutation that happened and was then selected for. But even still, our skin colors are not exactly the same, and I do not believe that it only takes a single mutation to turn "african"-style skin into "northern european"-style skin. I think there are gradual changes there.
Yeah, of course. That's the point. Our skin colours aren't exactly the same. Regardless of how that originally occured, we can now have a population of people in which there's a slight variance in melanin levels. Say 20-50 on the just made up melanin scale. So if you're somewhere where there's some selection pressure, where having more or less melanin means it's easier to survive and have kids that survive to breeding age themselves, then that 20-50 range is going to change in successive generations. Say those in the 20-30 range are less likely to have kids, those in the 40-50 range are more likely. And in a separate population, that's reversed. And that each kid will have a melanin level the same as one parent, + or - 5. Easy to see that if you run it for enough generations, you could end up with one population having a 0-25 melanin range, and the other having 70-100, and both of them stable because going below zero is impossible, and going above 100 is possible, but causes other problems, like using too much energy to produce. Two obviously different populations, all without needing any mutations from the original population, just a gradual slide along a scale that already existed. The disruptive selection & sympatric speciation I mentioned above is basically the same idea, but without the original populations being physically separate.
And so those gradual changes are also expressed, and mixed together, and that's why we have variance in our skin color, just as an example, but it's all related to changes in our genetic code or "mutations", somewhere down the line.
Am I mistaken on how this works? Isn't "variance" always a direct result of mutation in the first place?
Again, I think it's more an issue with definition than anything else. I don't think there's anything factually wrong with what you're saying, but I do think it helps contribute to misunderstandings. I honestly don't know if you could say that at some point, long before we were human, melanin level was identical, but then slight mutations led to there being variance. Same with other qualities. I suspect that sort of thing would go right back to the first examples of meiosis & sexual reproduction. But whether it's a case of various mutations along the way, or simply a case of variation being a direct consequence of getting genetic material from two parents, and that actually discovering the trick of sexual reproduction is all the mutation that was required, I don't know. Either way it does require some sort of mutation somewhere in the population's ancestry.
Anyway, found the experiment I was thinking of, it's being run by Richard Lenski:
http://myxo.css.msu.edu/ecoli/ is the project website, I just found that and haven't actually looked at it yet.
http://en.wikipedia.org/wiki/E._coli_long-term_evolution_experiment is the wiki page on it, and nowhere near as interesting as the book I read. I suspect it was something by Richard Dawkins, I'll search the bookcase and see what I can find.