In my book Speciation, written with Allen Orr, we give some estimates about how long it takes to make a new species. These estimates vary, of course. In the case of speciation that involves instantaneous genome doubling, as in auto- or allopolyploidy, a new “hybrid” species can arise in as few as three generations. But under normal conditions the process usually takes hundreds of thousands to millions of years. The record we found for “normal” speciation is the production of five species of cichlid fish in lake Nabugabo, a satellite of Lake Victoria cut off by a sand bar. In that case, each of the five species has its closest relative in Lake Victoria, and the differences between these “sister species” is not large, involving mostly changes in color.
Now, however, a case of equally rapid speciation, involving even more genetic change, has been reported in a new paper in the Proceedings of the Royal Society by Jonathan Puritz et al. (reference and link below; free download). The two species are starfish, Cryptasterina pentagonia and C. hystera. They both live on the northeast coast of Australia, separated by about 375 km. In the figure below, the distribution of C. pentagonia is shown in blue and that of C. hystera in red (the latter has a more limited distribution):
Although the species are ecologically similar and look pretty much the same, there’s a profound difference in their life histories. C. pentagonia is a “broadcast spawner” that spews its gametes into the sea (eggs and sperm unite in the water column), while C. hystera is not only a brooder, keeping fertilized eggs in a body pouch until the larvae hatch, but is also “selfing”: an individual is a hermaphrodite that fertilizes its own eggs. This is a pretty profound change that involves many aspects of morphology and physiology.
Nevertheless, genetic evidence establishes these two species as each other’s closest relatives (i.e., “sister species”). The genetic analysis involved five “microsatellite” loci and two loci on the mitochondrial DNA (remember this week’s caveat about mtDNA), but they also sequenced two gene fragments from nuclear DNA, so it’s a pretty good genetic sample. What they found is this:
- The selfing species has almost no genetic variation. That’s what you’d predict since self-fertilization erodes variation, but there’s virtually zero variation in C. hystera.
- Conservative age estimates of the divergence between these two species, based on the “molecular clock” (using DNA divergence to estimate age) range between 905 and 22,628 years. The best estimate is ca. 6000 years (6159, to be precise). That is an extraordinarily fast rate of speciation given the extensive restructuring of the reproductive system involved. I almost have trouble believing it could happen so fast.
- How do we know these are true species? Under the “Biological Species Concept” (BSC), the thing that counts is that the entities be reproductively isolated; that is, they can’t exchange genes. Well, that’s true to some extent in this case. One can’t really claim that the selfing “species” is reproductively isolated from its nonselfing relative because individuals of the selfing species are just as isolated from each other as they are from members of C. pentagona (selfers don’t mate with any other individuals). But individuals of C. pentagona are reproductively isolated from the selfer, because in principle they could coexist with that species and deluge them with sperm and eggs, but that wouldn’t produce hybrids since individuals of C. hystera fertilizes only themselves. This is a mistake we made in our book when claiming that the evolution of self-fertilization from a species that “outcrosses” does not cause reproductive isolation. It does—but in only one direction.
- What were the evolutionary forces that led to the evolution of selfing from normal outcrossing? (That was the ancestral condition, as we know from phylogenetic analysis.) The authors don’t know, but one possibility is the association of self-fertilization with cold waters in marine invertebrates. The southern species, which selfs, happens to be located south of a geographic break in water temperature where the ocean becomes colder along the cost of Australia. Another reason is “reproductive assurance”: if the selfer descended from a small number of colonists, or perhaps only one, then there’s a strong selective pressure to fertilize yourself (if you don’t have another individual to mate with you don’t leave any genes unless you can self).
This is a good paper, and I hope the young age holds up with further molecular analysis (more is needed). I have only one beef, and that’s about the discussion, where the authors say this:
The timing of these evolutionary changes is critically important for testing and rejecting some alternative hypotheses for speciation in Crypasterina. In particular, our results are not consistent with the slow, gradual loss of shared alleles (and of reproductive compatibility) via genetic drift in allopatry (classic geographical speciation), or with divergent adaptation to broadly sympatric microhabitat differences such as different intertidal heights.
The authors have made a serious mistake here by associating “allopatric speciation” (that is, the creation of new species after populations are isolated from each other by geographic barriers) with genetic drift, and “sympatric speciation” (speciation in small areas without geographic barriers) with natural selection. I have found this error pervasive in the evolutionary community: it was mentioned several times at the evolution meetings in Ottawa. The fact is that speciation of geographically isolated populations can occur by either genetic drift or natural selection, and everybody really knows this. That’s what happens, after all, when a colonist invades an island and forms new species after much modification by natural selection in a new environment. For reasons that Allen and I discuss in our book, we think that natural selection is far more important than genetic drift in cases of allopatric speciation. In sympatric speciation, selection is also likely to be involved because that form of speciation requires evolutionary forces strong enough to sunder an interbreeding population, and selection is much more likely to do that than is random genetic drift. But let’s get the record straight here. If you’re an evolutionary biologist, burn this sentence into your brain:
Just because speciation occurs among geographically isolated populations, that does not mean that the evolutionary force producing those new species was genetic drift rather than natural selection.
Got it? Now tell that to your colleagues.
I’m not sure where this mistake comes from, but it seems to be promulgated by the concentration of evolutionists’ effort on cases of speciation that occur without geographic isolation: sympatric or parapatric speciation. Because those often involve ecologically-based natural selection, somehow that’s been taken to mean that speciation that does involve geographic isolation doesn’t require ecologically-based natural selection. Serious error of logic.
J. B. Puritz et al. 2012. Extraordinarily rapid life-history divergence between Cryptasterina sea star species. Proc R Soc B 2012 : rspb.2012.1343v1-rspb20121343.