@gardenhead What I'm describing isn't natural selection but filtering. I.e. butterflies with all manner of wings are born, but statistically speaking the ones with smaller wings are killed earlier and therefore your observation is skewed since the "currently alive" population has a different wingspan distribution as the "when born" population. This filtering through death can occur without any adjustment to the genetics, if the deaths happen after procreation.

@gardenhead: Think of it this way: People will evolve to combat a deadly disease you get at 9 years old, because the ones that can't fight it die and never have children of their own, thereby eventually making the new population born only of children that survived this disease, which impacts the gene pool. That's natural selection.

@gardenhead: Comparatively, a disease you get at age 40 isn't going to be evolutionarily weeded out through genetics, because even the ones who die of this disease will probably already have had children and passed on their weak-to-the-disease genes.

So if you go into an old folks' home, and you notice that almost all people living there are strongly immune against the 40-year-old disease, that's because the weak ones died off. But it does not indicate that almost all people are born with immunity to that 40 year old disease - i.e. there is no proof of genetic selection, just filtering through death.

Also, to correct myself, when I said "What I'm describing isn't natural selection but filtering", I meant to say "What I'm describing, i.e. filtering of the population, doesn't always lead to genetic selection."

@Flater Yeah, I understand what you're saying. But why are you assuming that the butterflies who go splat have already reproduced as much as they are going to? I'm no expert in butterfly reproductive habits, but I don't think there's any reason to believe that only old, infertile butterflies get splattered. Eventually this should lead to natural selection.

@gardenhead: I didn't claim that only post-reproduction butterflies get splatted. I questioned the ability to deduce evolutionary change by going out and catching butterflies, as this observation is biased towards non-splatted living butterflies without necessarily proving an evolutionary change.

I only mentioned post/pre reproduction to explain why some natural selection entails gene pool changes and some natural selection doesn't.