Everyone knows that there are two kinds of elephants in this world: Asian and African. The Asian is the only one that can be trained and the African ones live in harmony with their environment until hunters come by and shoot them. Scratch a little deeper, and the African bush elephant lives by destroying its environment and moving on to new areas, where it destroys that environment, cycling back to the original region over generational time; Both African and Asian elephants can be trained; and there are three, not two species of elephant in this world: Asian, African Bush, and African Forest. Once again, everything you know is wrong. But you knew that.

The paper under consideration today totally nails the taxonomic distinction between the two African elephants (bush and forest) like Nadia Comaneci nailed her landings, at least until further research proves otherwise. They are two different species (Loxodonta cyclotis and Loxodonta africana). I don’t think this research is likely to be overturned however.¹ It has long been suspected, and some have accepted it for quite some time now while others have not that Africa has two species of elephant. The DNA study in Rohland Et Al. 2010 is both an extension of prior work (anatomical and molecular) and new work and is reasonably conclusive. The paper was published 30 seconds ago or so as of this posting in PLoS Biology, and you can try this link to get your own copy of it. (If that link does not work let me know, it might take a few minutes to fire up.)

This issue is actually much more interesting than just whether or not two populations are distinct enough to be considered different species. If that’s all there was to it, there would probably still be an argument on our hands. There are lots of species with large distributions and significant morphological variation across space to allow for this sort of argument. For example, the Cape Buffalo (Syncerus caffer) and the African Forest Buffalo (also Syncerus caffer) look very different and even act somewhat differently, but they are almost certainly overlapping populations. Or at least, if you go from the cape to the forest and every now and then stop to look at the buffalo, you will see a gradation from large, black, large-horned beasts to smaller, browner, small-horned beasts, and they tend to migrate across habitats now and then.

But not so much with the forest vs bush elephant. I’ve seen many photographs of, many bones of, and many footprints of the elephants in the Semliki Valley, which is at the forest-savanna junction, and I’ve had a good look at living elephants in the Kasingi channel and vicinity, and those elephants are just like the ones I’ve seen at Nairobi, Amboseli, the Serengetti, Tarangire, Kruger, and photographs and bones I’ve examined from dozens of other places. They don’t even vary that much in size by region (but do within region), with some of the elephants on the border of the forest and bush being among the largest recorded. But just a couple of hundred kilometers into the rain forest is an entirely different kind of elephant, much smaller, and distinct in behaviors. The classic photos of these elephants are from regions far to the west, and my view of the Ituri elephants was always as a fast moving shadow. They are hunted there, so one is only allowed a close view of a living elephant under two special circumstances: You are about to kill it, or it is about to kill you. But, I’ve seen those freshly hunted by others, I’m very familiar with their prints, and thus have a sense of their size, and have handled their bones and, well, now and then eaten one. I have never held the opinion, once I started looking into this, that the bush and forest elephant were the same species. Having said that, the paper by Rohland Et Al shows a surprising time depth for their separation, but with a twist.

To understand the findings being reported, it is useful to first understand that when we consider the living elephants, we are seeing a depauperate taxon. If we go back just ten or twenty thousand years or so there were at least three major types of elephants alive beyond what we have now: Mammoths, mastodons, and gomphotheres. These were not three species, but rather, three taxonomic groups within which there were multiple species (depending on when and where you look). Indeed, Asian elephants, of the genus Elephas, are part of the group that dominated in Africa in the not too distant past. If you look at the African fossil record over the last several million years, you see mainly relatives of Elephas and the African form is rare or absent. More recently, what we think of as African elephants spread, and the “Asian” elephants retracted in their range.

The famous Mammoth, the big hairy one with the curvy tusks, the icon of the ice age (both the geological time period and the movie) is one of the Asian elephants, and yes, these Asian elephants were spread across Europe and North America as well. Think of it more as a northern Northern hemisphere group, surrounding the arctic circle, than African vs. Asian.

The Mammoths and the Gomphotheres, meanwhile, are separate groups that split off the elephant lineage somewhat earlier. A recent study of proboscidean (elephant) classification recognized 175 species or subspecies in 42 genera across 10 families, starting out as an aquatic life form and changing to mostly terrestrial, and at about 25 million years ago, diversifying into the pattern we see in today’s forms combined with the Pleistocene fossil forms (gomphotheres, mammoths, mastodons, African and Asian elephants, etc.)

From Shoshani and Tassy 2004: A cladogram of proboscidean taxa based on hyoid characters (from earlier work by Shoshani and Marchant).

Looking at the above diagram, you need to imagine the Loxodonta africana branch (third from the right) with two sub branches, one for the bush and one for the forest African form.

Rohland et al. analyzed both ancient and modern DNA to try to resolve the African elephant relationships. Another objective was to assess the relationship between the African and Asian elephants and the Mammoth, as prior studies had linked the Mammoth to either the Asian or African elephants (with odds favoring the Asian connection, as indicated in the particular example above). Obtaining fossil DNA was especially helpful, obviously.

The researches used a combination of modern DNA sequences and targeted PCR amplification of ancient DNA to develop a large data set of the American mastodon, the woolly mammoth, and all three groups of living elephants (Asian, African bush, and African forest).

The data strongly support that the Asian elephant is the sister species to the extinct woolly mammoth, and that the divergence of African bush and forest elephant happened as far back in the past as the divergence of Asian elephants and mammoths. That gives strong support to the species distinction among the African groups.

Demographic model for the history of the Elephantidae. See the original paper for a detailed explanation.

How far back? That’s tricky. The authors are very cautious in their estimates because key parts of the data have rather large error bars. They are very conservative in stating that the split between forest and bush elephants in Africa occurred some time between 1.9 and 7.1 million years ago.

But it was complicated. The theory is that Asian types of elephants dominated in Africa, and what was to later become the African elephants were divided into two populations, one ancestral to the forest group and one ancestral to the savanna, or bush, group. On the demise/removal/exit/whatever of the Asian elephants, the two groups maintained their genetic difference because of their habitat differences, but there was some overlap. Because of the occasional interaction between the two groups, estimates made using only mitochondrial DNA suggest a much more recent divergence.

The researchers provide a somewhat convoluted explanation that I don’t happen to like much for the difference between nuclear DNA and mtDNA, so I’ll give you what they say and then say what I think:

The finding of deep nuclear divergence is important in light of findings from mtDNA, which indicate that the F-haplogroup is shared between some forest and savanna elephants, implying a common maternal ancestor within the last half million years…. The incongruent patterns between the nuclear genome and mtDNA (”cytonuclear dissociation”) have been hypothesized to be related to the matrilocal behavior of elephantids, whereby males disperse from core social groups (”herds”) but females do not… If forest elephant female herds experienced repeated waves of migration from dominant savanna bulls, displacing more and more of the nuclear gene pool in each wave, this could explain why today there are some savanna herds that have mtDNA that is characteristic of forest elephants but little or no trace of forest DNA in the nuclear genome.

While it is true that elephants are “female bonded” it is also true that some of the more spectacular examples of long distance migration of elephants, anecdotal but nonetheless real, are of females. On a day to day basis, males may disperse, but they do so in search of con-specifics, which may have obviated a lot of inter-species interaction via male ranging. Furthermore, males coming from long distances are typically excluded by resident males, which attenuates their long distance genetic contribution. Females coming from afar are rarely excluded by local males, and less often by local females. Here, I’m speaking broadly of female bonded mammals, not specifically of elephants.

Given that within elephant groups a subset of females reproduce, based on female dominance hierarchies, it is not that unlikely for a long-distance dispersing female to end up making a larger than average contribution to the mtDNA (passed on via females) lineages of a distant region. This explanation requires a more unlikely event but a less twisted explanation. Occam’s razor, which can cut both ways in a case like this, is probably not too important. We are trying to explain a quirky result. Historical contingencies rule.

The problems with the history of elephants in Africa is actually the tip of a rather large iceberg. Despite the strong theoretical argument that Africa must have had refugia of forest patches during the dozen or so periods of the Pleistocene when the forest contracted, especially the last few when the forest must have nearly disappeared, no one has found evidence of any paleorefuge. This means that there are huge gaps in our knowledge of ancient African habitats. I’ve argued before that mid to late Holocene human activities may have erased a huge habitat in Africa, which I called Habitat X, intermediate between forest and grass-dominated parkland. Along with this poorly defined middle ground could have been huge areas of fruit-rich but also grassy habitats with limited fire (an important part of Africa’s savanna ecology today) which could have been more like some Asian habitats today. Missing habitats? Missing Asian elephants? Coincidence? Probably, but it is fun to speculate.

The paper released today is an important and well executed extension of a string of research on the topic of Elephant phylogenetic history. It seems to end the debate over the species status of the African forest elephant, and to raise interesting questions about the dynamics of elephant biogeography over the last few million years.

¹There is a caveat. I have not yet examined the “supplementary information” related to this research, which includes the details on where the samples were collected. If the forest elephant samples are all from Cameroon, Gabon and vicinity and do not include Ituri elephants, than I reserve judgment that the pattern shown in the figure and discussed in the paper is complete. The Ituri elephants may show a closer relationship to the savanna elephants, or may end up being a third population.

Photograph of the African elephants by the author. All other photographs (C) creative commons, from Wikipedia (click the pic to see the original).

SHOSHANI, J., & TASSY, P. (2005). Advances in proboscidean taxonomy & classification, anatomy & physiology, and ecology & behavior Quaternary International, 126-128, 5-20 DOI: 10.1016/j.quaint.2004.04.011

Rohland, N., Reich, D., Mallick, S., Meyer, M., Green, R., Georgiadis, N., Roca, A., & Hofreiter, M. (2010). Genomic DNA Sequences from Mastodon and Woolly Mammoth Reveal Deep Speciation of Forest and Savanna Elephants PLoS Biology, 8 (12) DOI: 10.1371/journal.pbio.1000564

Coverage of this story elsewhere (Have you blogged it? Please add your link to the comments!):

• Nature News: African elephants are two distinct species
• Harvard Gazette: Seeing double: New study shows Africa has two elephant species … and why that matters

Have you read the breakthrough novel of the year? When you are done with that, try:

In Search of Sungudogo by Greg Laden, now in Kindle or Paperback
*Please note:
Links to books and other items on this page and elsewhere on Greg Ladens' blog may send you to Amazon, where I am a registered affiliate. As an Amazon Associate I earn from qualifying purchases, which helps to fund this site.