Because it was not based on schemata that were widespread in the population before the mutation was artificially selected.
The OP addresses the difference between natural and artificial selection. Obviously the machine being modified is the same in both cases. But, the result is different. Evolution can only occur in populations not in a few individuals. Artificially introduced variations will quickly be diffused.
Is there an example of stable artificial selection?
What does this mean?
The reason artificial selection can make rapid changes and then hits a brick wall is beause it exhausts the preexisting genetic variation that was available in the starting population at the relevant loci. If there’s no longer any genetic variation in the subject population, the strength of selection is irrelevant - there’s nothing for selection to act on until new mutations arise.
What do you mean by “stable”?
As @Riemann has stated, this is not a correct usage of the scientific definition of evolution. @The_Other_Waldo_Pepper posted the definition, “the change in the inherited traits of a population of organisms through successive generations”. Evolution can occur within a species without the formation of a new species. Some call that microevolution.
In response to the OP, he is confusing a mechanism for evolution (selective breeding) with evolution itself. Mechanisms for evolution include natural selection, artificial selection, genetic drift (no selective pressure, just an isolated population where random mutations get passed along), or gene flow (migration of traits into or out of a population). There may be more mechanisms than that.
Of course what you say is correct. However, it is not without controversy.
I proposed that the path from Wolf to Pekinese does not represent evolution because the phenotype is not stable. By stable I mean that it cannot maintain itself without continued human intervention. So the scope of human intervention was more than just selection. Hence the difference between natural selection and artificial selection.
Now, I foresee a counter argument that natural selection is driven and supported by the natural environment. And in artificial selection the human becomes the environment for both selection and support. Which wins the argument by definition.
The “mean” for a Saint Bernard is a Saint Bernard. The “mean” for a Chihuahua is a Chihuahua. As long as like only breeds with like, each is selecting from only a limited subset of all the alleles available for dogs as a whole. There would only be “progression towards the mean” of “dog” if there were intermediate sized-breeds that could allow gene exchange between Saint Bernards and Chihuahuas in the long term. An option that my hypothetical was explicitly designed to rule out. In my hypothetical of only Saint Bernards and Chihuahuas, the Saint Bernard would probably still be able to interbreed with wolves, but the Chihuahua probably would not interbreed or would do so very rarely because of mechanical isolation due to extreme size difference.
If that were true, good Darren, then it would be easy to breed championship St. Bernards. It is not, The standard is maintained only by rigorous selection and inbreeding. There is ample evidence that the mean is closer to the Pariah dog.
This may be true of ridiculous extremes such as the Chihuahua, which are obviously a strained result of human whimsy. But I believe the modern consensus (or at least a credible possibility) is that the primary process by which wolves first became associated with humans was one of self-domestication. In other words, it wasn’t that Thag the caveman found some orphaned wolf pups and thought they looked cute. It was ancestral wolves initiating the process, moving into a novel niche of association with human populations. And the specific wolves within the population that were best suited to this new niche were those that possessed certain traits that humans find acceptable and desirable - in other words, the wolves may have primarily self-selected, rather than being subject to any conscious artificial selection by humans.
Indeed true. Have you seen the movie Grizzly Man - the Fox domesticates itself.
Also the Russian Fox experiment indicates that the self domesticating animals have a different appearance.
Okay, if every Saint Bernard looks more or less like a Saint Bernard (even if some aren’t “champions”) and they only breed with other Saint Bernards, where do you think those “regression to the mean” alleles come from? The alleles have to be in the population before they can spread and become dominant in the population.
Oh, your hypothetical has only St Bernards and Chihuahuas. I doubt it would take them long to hybridize. A mature Chihuahua would readily mate with a juvenile or runt St. Bernard.
Dogs have a very high degree of natural variation.
A few thousand years ago human decided that since the meat and milk of Bos primigenius were tasty, they would try to pin some up for easy access. Today there are billions of Bos primigenius descendants spread around the world. If humans had not thought the were tasty, then the species would be extinct. Being beneficial to humans has been highly beneficial to Bos primigenius (and all of our other domesticated species with billions of individuals).
Okay, how about only these two?
Where beneficial means reproductive success. Is the tail of a peacock really beneficial to him? The reproductive success of the praying mantis is not too helpful to the male.
In selective breeding reproductive success is tied to what the breeder wants, as opposed to being simply success in breeding.
Not all natural evolution (as opposed to artificial or human driven evolution like selective breeding) is stable either. Evolution does local optimization, so unstable situations can easily occur leading to extinction. Why dogs require us (which has been a successful strategy for them) so do the various mites and bacteria that live inside us and on us. We might not have selected them, but if we go, they go.
There’s also the famous example of the “samurai crabs”: Heikegani - Wikipedia
Yes, it attracts mates and intimidates predators.
I thought that one was a little dubious, but I had heard of it (Sagan, of course)
Another wrinkle - some non-human creatures do things that would be considered selective breeding if people did them - like ants farming aphids.
All genes are at risk for mutation. In genes for traits that matter, damaging mutations get selected against. But if the trait no longer matters, broken genes are no longer weeded out by natural selection. Therefore all traits are stable only as long as they are useful. Think of loss of vision and pigmentation in cave animals, and loss of Vitamin C production in several animal groups.