When Darwin wrote On the Origin of Species in 1859 he kick started modern understanding of how things work but, oddly, he didn’t explain how species originate. He didn't even really distinguish between species and varieties -- he thought all variation was on a continuium and the lines drawn were arbitrary. This is not a popular position today, but I'm not convinced he was wrong. Today, species are typically defined as having the ability to produce fertile offspring, in nature. But what is "nature"? And how fertile is "fertile"? The complete gradations in each of these does pose an awkward problem.
But let's accept that there are species, as there clearly are organisms in nature that can't have offspring with one another. Others do mate, produce strong offspring, but the offspring are infertile.
It's easy to understand the incompatibility of a starfish and a crow. Probably it physically can’t reach. If so, it wouldn't fertilize, and if it did the fetus would not develop and if was born it would not be viable but imagine a starfish-crow cross anyway. Where would it live? What would it eat? There is no half-way between the sea and the land. Clearly, utterly, and permenantly separate species.
But sheep and goats can’t breed, either, and they run in similar circles. Neither can robins and sparrows, rats and opossums, orb spider and garden spider. When wolves and coyotes mate, their offspring is fertile, but strange. Neither parent species want to breed with them, and they don't much care for their own kind, either. For some good reason there are lots and lots of different species. Two million species have been named, there are probably 10 million more. Why so many species?
One situation is understandable -- genetic isolation. Allopatric speciation is when a new boundary separates populations -- say a new river, contintental drift, a desert or glacier. Paripatric speciation is very similar. Say, a crew of critters floats to another island and morphs over time into something altogether different. In either case it could be natural selection or genetic drift that does it -- drift will actually occur quickly when the founding population is small. And if a population doesn't travel much groups on the periphery (expecially) can also develop local variation that eventually crosses the threshold of speciesdom. This is a perapatric process. All of these are fundamentally geographic.
But geographic isolation is not necessary for species to develop. Someone fogged 19 trees of one type in Panama 1,000 species of beetle fell out. No, something is definitely going on besides geographic isolation. When speciation occurs in the same area, with no geographic or temperal separation, it's called sympatric. This is much much harder to understand.
First there are the infertile hybrids, like mules, which sometimes have excellent characteristics of both parents; Mules have "hybrid vigor" like the strength of a horse but the steadfast temperment of a donkey. Ligers, the cross between lion and tiger, are huge. In a sense hybrid vigor is the genetic opposite of inbreeding – the vigorous hybrid offspring are fast growing, large, and healthy -- better, you might say, than the originals. They are not uncommon. Modern corn crops are hybrid vigorous. Black Angus crossed with Hereford, and Hampshire swine crossed with Yorkshire make vigorous hybrid livestock. The big eggs we eat come from hens with parents of different stock. Mules come from high-strung horses and little donkeys. Not all hybrids are robust of course, in fact most are inviable or sickly but there does seem to be a sweet spot between species.
Now think of what this means. If there was no hybrid infertility near species which drift apart but then chance to mate would fill the gap between them with a possibly vigorous intermediate. The result would be no distinct species at all, just a grand smear of life forms from bacteria to fruit fly to crow to giraffe to human. Imagine putting all dog breeds in a gymnasium for 5 years, but at a global scale, with everything.
But this does not happen. We have unique species because those healthy hybrids have a fatal flaw: they are incapable of having offspring. That big mule, in essence, is the thick end of a wedge that is being driven between species; it’s the end we can see. The sharp end is its sterility. The males are sterile (technically it's the XY which is infertile, so it birds it's the females).
There may be advantages to there being different species. Take wolves and deer. Each fits into a niche environment, allowing the other to survive, keeping each other in check. Wolves are programmed to kill deer, and deer are programmed to avoid wolves. Imagine how much difficult it would be to program these creatures if there were all manner of deer-wolves in between.
But it's not species that are driving evolution. I am with Hamilton and Dawkins on this point; not only species or groups, even organisms don’t generally drive evolutoionary change; it happens at the level of the gene. Genes, in effect, succeed or fail by the number of times they can duplicate themselves. Most mutations are bad -- a monkey wrench thrown in the works -- but a good genetic mutation can succeed self-replicating but work against a higher level of organization (organism, group, a species, etc.). Genes will naturally "cheat" the higher orders if it helps them.
So is there a genetic advantage of hybrid infertility? No. There can't be. Genes specifically making a hybrid infertile wouldn't duplicate and would be removed from the gene pool.
So maybe the infertile hybrid offspring – mules – are servants for their parents? Maybe the mother benefits from the protection a mule may provide so she can have more regular babies? This can't be either. The amount of investment a mare would have to put into making a sterile mule would have to be less than the protection she receives from it. Very doubtful.
What’s more, mule infertility is gender-specific. In mammals, males are infertile (in birds, females). What would a male horse gain by making his son infertile -- just so it can protect a mare donkey whom the stud is unlikely to impregnate again? There is just nothing in it for the stud – for his genes, that is. And therefore, they just won't last.
I've read that maybe speciation is a byproduct of aggressive genetic variation designed to foil bacteria and viruses that go for the dominant form. The rarer variants simply slip away from the attack. But it doesn't seem to address the issue of vigorous but infertile hybrids.
And there is another possibility I found intreuging. If infertility comes from a combination of obstacles: 1) temperal incompatibility, 2) geographic separation, 3) sexual preference, 4) gamete inviability, 5) hybrid mortality, or 6) hybrid infertility, any one of these would make a variant less attractive, because the effort spent on procreation would be less often productive. Therefore, once one of these is in place some others will follow. For example, once red-winged flies become even slightly less fertile with yellow-winged flies, individuals of either color will tend to avoid the other type so to waste less of their precious efforts. One group may then morph to mating in the morning, the other in the evening to increase the chance of a successful match, and so on until they become so different that they can't or won't mate at all. As I understand it, this effect is called Reinforcement.
But the way through to hybrid infertility is still not clear to me; even this explanation is riddled with big questions, the biggest among them is why the vigorous but infertile hybrids? I’m at a loss. Darwin didn't know, but by now I believe someone does. But not me.
Can someone please enlighten me. Why oh why are the vigorous hybrids sterile? In other words, and this is the point: Why are there species.