Check out studies of island ecologies. Evolution can proceed quite rapidly in an isolated community.
Yes, but it is difficult or perhaps impossible for a biologist to "simulate" that kind of selection process with an artificial environment, or to monitor events on a real island. Perhaps if biology continues for the next few thousand years we will observe species diverge in places like the Galapagos Islands. To produce this artificially, it is not enough to simply breed animals such as show dogs. You would have to provide a full range of predators, diseases, food supplies, and selective advantages which are often very difficult to understand -- and I think impossible to control for. It is often astounding what turns out to be an advantage. As the neo-Lamarkians point out, cutting off mouse tails for generations proves nothing -- there is no selective advantage to having no tail.
I haven't got references at my fingertips and haven't got the time to go digging but I'm pretty sure I've read of longitudinal studies of islands which showed substantial changes in a quite short period (less than a century, IIRC). Not sure if the island residents diverged so far that they couldn't cross-breed, tho.
The "crossbreed" definition of species is a can of worms, that is growing wormier with each passing year. It turns out all kinds of different species can crossbreed, albeit usually with sterile offspring. Zebras can breed with horses, and donkeys with horses to produce mules. Mules are usually sterile, but not always, it turns out. I do not think anyone would assert that donkeys and horses are the same species.
As I see it, the fact that such widely divergent species can crossbreed is good evidence that evolution did occur. It turns out species have not really diverged very far after all, and the distinction between species is somewhat artificial. (I mean it is man-made.) As far as nature is concerned zebras and horses are still pretty much the same animal.
Also, by the way, the distinction between sexes is somewhat artificial. People are very hung about this distinction and about sex roles, and most people are horrified when confronted with hermaphrodites and other sexually ambiguous people, but in fact there are, as I recall, six classes of human hermaphrodites and they are surprisingly common. Nowadays surgery is usually performed to hide the results, but nothing can hide the genes -- or prevent their _expression_ from time to time. Nature never completely finished the job of separating male and female.
- Jed
