If you follow the popular science press, or read the evolution blogs, you’ve probably heard that, according a new paper in PLos Biology by Rohland et al., scientists have added another species of elephant to the two we have already. Up to now we all knew about the African elephant (Loxodonta africana) and the Asian elephant (Elephas maximus). Besides being located on different continents, these two species differ in size (the Asian is smaller) and other morphological traits (Asians have much smaller ears and, unlike Africans, only the males are tusked).
Previously, though, the Africans were divided into two subspecies, L. africana africana, called the “savanna” or “bush” elephant, and L. africana cyclotis, the “forest elephant.” These subspecies differ in morphology (though not as strongly as the African vs. Asian species), ecology—well, at least location; they are, as their names indicate, found in different habitats, although there is some co-occurrence (and putative hybridization)—and social behavior. (There’s no strong evidence that the morphological differences between the subspecies have a genetic basis rather than being induced by different ecologies (I suspect they are genetic, especially if the differences persist when the subspecies are raised in the constant environment of a zoo). I’m a bit warier about social differences, which can originate and be passed on culturally rather than genetically.
A paper by Roca et al., published in Science in 2001, examined the mitochondrial DNA of these two subspecies and found that they were quite divergent—far more diverged genetically (and presumably in time, since DNA differences largely reflect the time since populations split from a common ancestor) than was previously suspected.
The new paper by Roland et al. extends the earlier study to both mitochondrial and nuclear DNA, and confirms the earlier finding that the savanna and forest elephants are anciently diverged—perhaps by as much as 2 to 7 million years. They are still each other’s closest relatives, though, and are both more distantly related to the Asian elephant, which diverged from its own closet relative, the extinct wooly mammoth (Mammuthus primigenius), at about the same time that the two African groups diverged from each other.
Based on the long divergence between the two African populations, and their morphological differences, Roland et al. suggest that the savanna and forest elephants should be recognized as different species, bringing the total living species to three.
The paper is online for free, and if you want a shorter summary of the results, Greg Laden’s blog has a good discussion. I would like, though, to make two points about these results.
First, although I don’t have a huge quarrel with designating these forms as new species, I’m not particularly keen on designating populations that are geographically isolated, and have some morphological differences, as new species based largely on divergence time. That’s because I adhere (for reasons I outline in the book Speciation that I co-wrote with Allen Orr) to the “biological species concept” (BSC), which recognizes populations as different species if they exhibit reproductive barriers between them that prevent hybridization when the two populations co-occur. I outline my reasons for adhering to the BSC in chapter 1 and the Appendix, based on its connection to what most evolutionists see as “the species problem”—the existence of discrete, objectively recognizable groups in areas where they co-occur.
Now the savanna and forest elephants are mostly isolated in space, but there is some evidence that they hybridize when they do co-occur, though I can’t find information about whether those hybrids are fertile. (If they aren’t, then the two groups have no possibility of exchanging genes and are definitely different biological species). But their hybridization where they co-occur makes their status as biological species questionable.
What about the ancient genetic divergence and morphological differences? Well, unless the morphological differences are correlated with reproductive barriers, they’re not good indicators of reproductive incompatibility and status as species under the BSC.
We all know that two populations instantly recognizable by different appearances aren’t necessarily isolated reproductively.
Take the different human populations (“races” or “ecotypes,” if you will) that were pretty isolated before modern transportation began moving their members about. You would have had no trouble telling apart populations from eastern Asia from those of the Yucatan from those of sub-Saharan Africa. Yet we don’t think of those human populations as members of different species. Why not? Because we know now that they’re all reproductively compatible with each other. Any human male can in principle produce a fertile child when mating with any human female, regardless of where they come from. Are the morphological differences between savanna and forest elephants greater than those between human “races”? Who knows?
What about divergence time, as indicated by genetic difference? That’s a bit dangerous, too, because while genetic difference between populations can indicate the time since they last shared a common ancestor, it doesn’t necessarily say anything about reproductive compatibility or the possibility of gene exchange. Reproductive incompatibility is probably a byproduct of divergent selection, and although two populations are anciently diverged, they didn’t necessarily experience divergent pressures of natural or sexual selection that would cause reproductive incompatibility as a byproduct. If their environments are largely identical, for instance, there may be few types of selection that would make the populations diverge in a way that could produce reproductive barriers.
As an example, the African helmeted guinea fowl can produce viable hybrids (I don’t know if they’re fertile) with the domestic chicken, yet those species are separated by at least 55 million years. That’s a divergence equivalent to that separating humans and lemurs, taxa which of course can’t produce hybrids! The point is that different groups of animals and plants diverge evolutionarily at different rates, depending on the environmental differences they experience, as well as on other factors. And this means that it’s dangerous to infer anything about reproductive compatibility from divergence time alone—at least when that diverence time is a few million years or less.
Now Roland et al. could have at least alluded to this problem, since most evolutionists do employ the BSC, but nowhere in their paper will you find a mention of reproductive barriers. Their view that these populations should be different species comes solely from divergence time (and perhaps consideration of their morphological differences). In fact, the title of their paper is “Genomic DNA sequences from mastodon and woolly mammoth reveal deep speciation of forest and savannah elephants.” Deep speciation? What they really mean is “deep genetic and temporal divergence” between the elephant groups. And had I written the paper, those are the words I would have used.
Modern systematics, should, I think, be informed by evolution, at least to the extent that if you classify geographically isolated populations as different species, you should have a reasonable inference that they wouldn’t be able to exchange genes if they lived in the same place. (This is not impossible in some groups. In flies, for example, if you cross two allopatric groups and find that their hybrids are inviable or sterile, you are nearly 100% certain that they are biological species). It’s my opinion—and I know that others differ—that modern systematics needs to be informed by information about reproductive compatibility. And indeed, some modern alpha taxonomists (that is, those who name and classify different species) already do this. Roland et al. not only didn’t do this, but don’t even mention the issue. (This is, by the way, an ancillary point and not a severe criticism of their very nice paper.)
I’m not going to quarrel with the idea that there are now two species of African elephants. It would have been nice, though, had there been some discussion of their reproductive compatibility, and of the status of the hybrids when those two “species” co-occur.
My second point is that in many cases, and maybe even this one, species are named not only because of their biological properties, but for political reasons: it’s easier to get government protection for entities that are considered members of different species (or, in the US, of different subspecies) than simply different populations of the same species. This issue makes biologists more willing to name new species than they otherwise might have been.
As New York Times reported when discussing the elephant work:
The evidence means that conservation efforts may need to be re-evaluated, said Alfred Roca, an animal scientist at the University of Illinois and another author of the study.
“Conservation efforts tend to focus on savanna elephants,” he said. “But the forest elephant also really needs to be a priority for conservation.”
And, from National Geographic’s piece on the elephants:
Raphael Ben-Shahar, an elephant expert at Oxford University, says, “Up until DNA fingerprinting tests, species were defined on the basis of morphological and anatomical differences.” Using the old classification yardsticks, the forest elephant was merely a subspecies of the savanna elephant. However, there was widespread disagreement among taxonomists as to whether the differences between the two elephant types were significant enough to denote separate species, he says.
The DNA evidence should put the controversy to rest. [JAC note: DNA differences are not sufficient to put the controversy to rest, and a good reporter should have known that.]
“The impact on management strategies if there are two elephant species in Africa is huge,” says Ben-Shahar. “Now we will have two species that are less numerous than was thought before.”
Remember: we have known for decades that the forest and savanna elephants are morphologically different, and have known since 2001 that they are anciently diverged as well. These conclusions have not changed at all with this new paper.
Let’s face it: we biologists want to conserve everything, and will glom onto any strategy that lets us save as many populations as possible. That’s not a bad thing, for as a biologist I feel that the diversity of life has inherent value, both in terms of us not having the moral right to screw around with other equally-evolved forms, and in terms of biologists’ selfish desire to keep things around to study and learn more about nature.
But biologists often have to hide this motivation: we must pretend that we’re saving populations because we need to retain genetic diversity, or prevent inbreeding, or save rare alleles that could bring back a larger species. We can never divulge the real reason why many of us want to save things like the elephants—because they have inherent value as organisms, and because they’re fascinating. That’s why many conservation biologists are busy worrying about the species and subspecies status of plants and animals: they secretly treasure them for their own qualities, but have to make a different case to the government and public about why they need to be saved.
This isn’t just a theory of mine: I’ve talked to many conservation biologists who admit that the real reason they want to save species differs from the rationale they must offer the public and the government. (Not all conservation biologists are like this, of course: some really are motivated by the reasons they give the public and official agencies.)
All conservation biologists are doing a wonderful thing: they’re keeping our own oafish, selfish, and greedy species from driving every other species to extinction. Plants and animals have no defense against the depredation of humans; they have only the law—and the morality of right-thinking folks—to protect them. In some ways conservation biologists are the most important of all biologists, for without their efforts what would we have to study? Without them, ecologists and evolutionists would eventually be reduced to studying their own intestinal flora.
What a shame, though, that we have to manipulate biological nomenclature, and dissimulate about our motivations, to keep other species alive! Shouldn’t we name species based on biology rather than politics?
h/t: Geoff North
Roland, N., D. Reich, S. Mallick, M. Meyer, R. E. Green, N. J. Georgiadis, A. L. Roca, and M. Hofreider. 2010. Genomic DNA sequences from mastodon and woolly mammoth reveal deep speciation of forest and savannah elephants. PLoS Biology 8(12): e1000564. doi:10.1371/journal.pbio.1000564.
Roca, A. L. N. Georgiadis, J. Pecon-Slattery, and S. J. O’Brien. 2001. Genetic evidence for two species of elephant in Africa. Science 293:1473-1477.