Because I work on speciation, I’ve been inundated with emails this week (okay, really just one email) asking about Sean B. Carroll’s New York Times piece on hybrid speciation, “Hybrids may thrive where parent species fear to tread.” I won’t summarize it here because it’s short and you can read it yourself, but the upshot is that “hybrid speciation” may be much more important than evolutionists believe. I think Carroll’s probably wrong, and wanted to say why.
The topic of hybrid speciation comes up surprisingly often. Indeed, when I talked to a group of high school kids in Houston via Skype this Monday, one of them asked if a new species could arise from mating between two other species. There are two ways this could happen. The first is simple fusion: members of two different species hybridize, and their gene pools fuse back into a “new” species, supplanting the parents. This probably happens pretty often in evolution, since geographically isolated populations that are at least on their way to becoming different species might lose their geographic isolation in the fullness of time. This may be what is happening now with polar bears and grizzly bears. Their ranges are changing, probably due to global warming, and they’re coming into contact. Although they’re considered different species (they were previously limited to different areas because of their differing ecologies), hybrids—called “pizzly bears“—have been found in the wild. Here’s one:

Fig. 1. A slaughtered pizzly bear with patches of brown on its fur. Its hybrid status was confirmed by DNA testing (from National Geographic.)
We’re not sure whether polars and grizzlys will fuse into one species in the next few centuries, or whether hybrids will remain only a sporadic occurrence. But certainly fusions like this have occurred over evolutionary history.
But there’s another way a new species can form by hybridization. Two species can hybridize, and a few individual hybrids can give rise to a third species that is distinct from the two parents and can coexist with it. This depends on hybridization being more than a once-in-a-lifetime occurrence, and the hybrids sorting out the two parents’ genomes into a new, genetically different population that can’t reproduce well with its two parental ancestors.
This seems to occur fairly often in plants through a process called polyploidy, where an inter-species hybrid doubles its chromosome number. I describe this in WEIT, and won’t repeat it all here. This form of hybrid speciation may be responsible for as many as 4% of new plant species.
But the form of speciation Carroll discusses is called diploid hybrid speciation, which doesn’t involve doubling of chromosome number. Two species simply hybridize, and the semisterile hybrid then sorts out the parental genes and chromosomes into a new genome that is reproductively isolated—and ecologically different—from those of its two parents. Carroll describes a few cases, mostly in plants, and concludes:
The discovery of hybrid species and the detection of past hybridizations are forcing biologists to reshape their picture of species as independent units. The barriers between species are not necessarily vast, unbridgeable chasms; sometimes they get crossed with marvelous results.
I work on speciation, and although we’ve known about these cases for a while, I haven’t reshaped my picture of species—nor have many others. The problem is that Carroll’s few examples—three sunflowers and a fly—pretty much exhaust the known cases of diploid hybrid species (there’s a North American butterfly, too, but he didn’t cite that one). And it’s not that biologist haven’t looked for hybrid species: in many groups we have, and simply haven’t found them. There are several thousand species of the fruit fly Drosophila, for example—it’s probably the most heavily-studied group of animals on Earth, at least from a genetic point of view. And not a single species is a hybrid between two others. Indeed, we have only a half-dozen or so cases of any interspecific hybrids at all being seen in the wild, and almost all of those are not only one-offs, but the animals are completely sterile (and hence evolutionary dead ends that cannot produce new species).
Ditto for birds. It’s been estimated that 10% of bird species are known to hybridize, but this should not be taken as showing that hybrid speciation is important. For one thing, most of those hybrids are one-offs, very rare occurrences that don’t lead to anything. And not only are they rare, but they are probably either physiologically sterile or unable to find mates. Finally, we know of not a single bird species that is a diploid hybrid between two others. Diploid hybrid speciation in animals is likely to be quite rare, and certainly not something that’s going to “reshape our picture of species.” That’s not to say it’s not interesting, but just that it’s probably not very common—and hence not world-shaking.
My friend Loren Rieseberg and his colleagues discovered the three diploid species of sunflowers, and that work is deservedly famous. But even he admits that hybridization between plant species in nature is rarer than most people suppose, and that diploid hybrid species of plants are quite uncommon (although polyploid hybrid species are not).

Fig. 2. The diploid hybrid species Helianthus anomalus studied by Rieseberg and his colleagues. The species derives from hybridization between H. annuus and H. petiolaris. Its novel combination of genes enabled it to colonize sand dunes where the parental species can’t live.
At the end of his piece Carroll cites the recent discovery that the genome of modern humans carries a small amount of DNA from Neanderthals. This suggests (although there is dissent) that humans coming out of Africa in our most recent migration, about 60,000 years ago, hybridized with Neanderthal populations that derived from an earlier migration. But this says little about diploid hybrid species, for it’s merely hybridization between different human populations. (Recall that both Neanderthals and the ancestors of modern humans are both considered members of the species Homo sapiens.)
The entire recent history of our species involves hybridization between previously isolated populations. Those populations may have been on the way to becoming different species, but they never got there because they weren’t isolated long enough to become reproductively incompatible, and our species also devised means of transportation that linked the populations together, causing interbreeding. Bit by bit, the human gene pool is melding. About 20% of the genes in African-Americans, for example, came from matings with whites after they were brought from Africa as slaves. This is precisely the same phenomenon observed in Neanderthals/modern humans, except that those populations had been isolated a bit longer. It’s still “hybridization”, but it’s of no relevance to speciation.
Hybrids continue to fascinate both biologists and laymen alike. I’m not sure why—perhaps because they’re seen as a violation of the “natural order.” And they are certainly interesting, and of some evolutionary importance. But for the nonce I think biologists go overboard when claiming that hybridization will completely revise our view of nature, or of evolution.