Um... evolution proceeds through the accumulation of DNA changes, aka mutation. You simply cannot talk about mutation and pretend it has nothing to do with evolution, any more than you can talk about architecture and pretend it has nothing to do with building construction. And there are plenty of scientists who *do* wonder what will happen to a population if they put selective pressure on it. Oh, yeah--since literal creationists place so much emphasis on the word "species", I'll just point out that it *is* pretty much a human concept, and one that is not easy to define in a scientific manner. If the offspring of A+B is fertile, and the offspring of B+C is fertile, but the offspring of A+C is sterile, then where do you draw the species line? It gets even trickier with bacteria--they give DNA to each other whether they belong to the same (human defined) species or not.
Trait selection works on already-present genes. No matter how many dogs I cross I am never going to get something that is not a dog.
Sure. And every one of these dog breeds looks exactly like the original wild dogs < /sarcasm >. Dog breeds look so different from each other because they have genes that never existed in wild dogs, because their genes mutated and humans made sure they survived. From a scientific point of view, I would call teacup poodles and great danes different species--they certainly can't breed together, so they fit the definition.
You know corn, that yummy yellow grain that we use for so many things? The corn species does not exist in the wild. It exists because humans 10,000 or so years ago started selectively breeding a kind of wild grass, and they ended up with a whole new species, corn.
Mutation does not prove evolution. I can witness a child playing with legos, but that doesn't mean he is building a skyscraper.
Dog breeds look so different from each other because they have genes that never existed in wild dogs, because their genes mutated and humans made sure they survived.
Are you sure about that? How do you know selective breeding is not just expressing different genes? You know the function of what, two percent of the genetic code? You really can't rule out much with that little knowledge, yet you are making grand leaps of faith and putting on ever bigger hats to frighten the natives.
It exists because humans 10,000 or so years ago started selectively breeding a kind of wild grass, and they ended up with a whole new species, corn.
That is hardly an example of random mutation and natural selection. That is an example of crossing two plants with desirable qualities to increase that trait. The genes already exist, they just aren't expressed. The random crossing in nature is preserving lots of genetic information, but not strongly expressing much of it.