Header image

john hawks weblog

paleoanthropology, genetics and evolution

Photo Credit: Elephants. Joanne Goldby CC-BY 2.0

Genomes of straight-tusked elephants

Earlier this month in eLife, Matthias Meyer and colleagues published a cool paper: “Palaeogenomes of Eurasian straight-tusked elephants challenge the current view of elephant evolution”.

The straight-tusked elephants lived in Europe and western Eurasia as far east as India during the Pleistocene. Most people are familiar with other extinct elephant relatives, such as mammoths or mastodons. The straight-tusked elephants were not mammoths, and they are assumed to be much more like living elephants because they seem to have entered more northerly parts of Europe mainly during interglacial times. Paleontologists have noted that the straight-tusked elephants share some morphological features with Asian elephants, as mammoths do also. For some paleontologists, these similarities are so compelling that they have classified Palaeoloxodon as part of the Asian elephant genus, Elephas.

Ancient DNA evidence is breaking open the study of how the extinct relatives of living elephants moved, interacted, and evolved. First mitochondrial, and more recently nuclear gene sequences have revealed different populations that lasted more than a million years, yet hybridized and mixed where they met.

During the last ten years, it has become clear that populations of elephants in the central African forest have a long history as an evolving lineage distinct from savanna elephants across most of Africa. Forest and savanna elephants have increasingly been recognized as two species, Loxodonta cyclotis in the forest, and L. africana across the rest of Africa. No fossils have been attributed to L. cyclotis, and the L. africana fossil record is quite sparse up until 20,000 years ago.

Before that time, Africa itself was rich in elephants attributed to Palaeoloxodon, especially P. recki, which many sources identify as Elephas recki. P. recki has been identified from fossils as early as 4 million years ago, and as late as 300,000 years ago. Other African species of Palaeoloxodon (again, often classified as Elephas) have been interpreted as part of a single P. recki lineage, including the earlier P. ekorensis and the later P. iolensis. P. iolensis survived until around 30,000 years ago. With so many species identified as Elephas or closely related to Elephas, and with mammoths sharing so many features with living Asian elephants, the basic idea has been the Asian elephant branch of the elephant phylogeny was once global, with mammoths spread across the northern tier of Eurasia and across the Americas, extinct P. antiquus in western Eurasia, extinct P. namascus further east in Asia, and extinct P. recki in Africa. Paleontologists have speculated that P. namascus may itself be the immediate ancestor of living Asian elephants, Elephas maximus. Living African elephants, Loxodonta, were the odd elephants out.

Meyer and colleagues obtained mitochondrial genomes from four individuals of P. antiquus, three of them from Neumark-Nord, Germany, and one from Weimar-Ehringsdorf, Germany. They find that the Neumark-Nord elephants probably date to the last interglacial, around 120,000 years ago, while the Weimar-Ehringsdorf elephant dates to the previous interglacial, around 230,000 years ago. This is a pretty small section of the overall geographic range covered by P. antiquus:

Palaeoloxodon sites across western Eurasia, from Meyer et al. 2017
Figure 1 from Meyer et al. 2017, original caption: "Palaeoloxodon antiquus, geographic range based on fossil finds (after Pushkina, 2007). White dots indicate the locations of Weimar-Ehringsdorf and Neumark-Nord."

What they found was that P. antiquus mitochondrial genomes are not related to Elephas at all; they’re related to forest elephants:

Surprisingly, P. antiquus did not cluster with E. maximus, as hypothesized from morphological analyses. Instead, it fell within the mito-genetic diversity of extant L. cyclotis, with very high statistical support (Figure 2). The four straight-tusked elephants did not cluster together within this mitochondrial clade, but formed two separate lineages that share a common ancestor with an extant L. cyclotis lineage 0.7–1.6 Ma (NN) and 1.5–3.0 Ma (WE) ago, respectively.

That’s not a small difference. Living Asian and African elephants came from a common ancestral population more than 6 million years ago, during the Late Miocene. They are about as different from each other genetically as humans and chimpanzees. The fossil story was just wrong–and it’s as big a difference as misidentifying a Neanderthal as a fossil chimpanzee.

Elephant phylogeny from Meyer et al. 2017
Elephant mtDNA and nuclear DNA phylogeny from Meyer et al. 2017. The Neumark-Nord (NN) and Weimar-Ehringsdorf (WE) straight-tusked elephants are indicated. The mtDNA tree has a time scale (bottom) but the nuclear DNA tree has no time scale associated with it.

All four of the P. antiquus mitochondrial lineages are on the same branch as living forest elephants, and in fact some forest elephants have mtDNA genomes that are closer to P. antiquus than to some other forest elephants. In other words, the mitochondrial genomes of P. antiquus fall within the variation of L. cyclotis. Within this variation, the mitochondrial lineage of the earlier Weimar-Ehringsdorf elephant is part of a different clade than the three Neumark-Nord elephants, so the P. antiquus mitochondrial genomes are not a monophyletic group.

Now, we might well expect the story with the nuclear genome would be different for elephants. We know that the story for Neandertals is different considering the mitochondrial and nuclear genomes: the Sima de los Huesos nuclear genome groups clearly with later Neandertals even though the mtDNA of later Neandertals is more similar to that of living humans.

There is another reason why elephant nuclear and mtDNA genomes might be discordant. In humans, mtDNA is markedly less diverse than most parts of the nuclear genome, and mtDNA types occur across wide geographic areas. Elephants are the opposite. Their mitochondrial DNA exhibits substantially greater variation among populations than the average for the nuclear genome, because female elephants very rarely transfer between groups. Most gene flow in elephants is male-mediated, and male elephants sometimes disperse over very long distances. These contrasting patterns of nuclear and mitochondrial diversity in elephants are consistent enough to provide a way to “triangulate” the region that ivory samples originated (Ishida et al. 2013).

Meyer and colleagues cannot assess yet whether the Weimar-Ehringsdorf elephant would yield a divergent nuclear genome, because they didn’t get nuclear evidence from it. But two of the Neumark-Nord P. antiquus specimens yielded nuclear genome data and they are a close sister group compared to all the forest elephants. That is, the African forest elephants were much broader in their mtDNA phylogeny, and tighter together in their nuclear genome, just as one would expect from the mass of evidence about them.

Palaeoloxodon antiquus tooth, by Khruner (Wikimedia)
P. antiquus tooth. Photo credit: Khruner, CC-BY.

So, that’s an interesting data point about elephant evolution. A widespread extinct species of elephant, which on morphological grounds was interpreted as an Asian elephant relative, is actually related to forest elephants within Africa. Forest elephants today are a relative island species in central Africa, surrounded by savanna elephants. So from today’s standpoint, forest elephants look like a geographic and phylogenetic relict of a much more diverse lineage that once existed.

We already know that today’s situation did not exist earlier in the Pleistocene. In the past, many parts of Africa were inhabited not by today’s savanna elephants but instead by other extinct species, for much of the Early and Middle Pleistocene, P. recki. Savanna elephants are found in the fossil record as early as 500,000 years ago, but they are a relatively rare component of the elephant diversity in comparison to the extinct Palaeoloxodon species.

Of course, without ancient DNA evidence, it’s not certain that these other extinct Palaeoloxodon species are closely related to the forest elephants and P. antiquus.

Furthermore, the nuclear genome evidence presented by Meyer and colleagues does not establish whether the P. antiquus population may have exchanged genes with Asian elephants, thereby accounting for some of its anatomical resemblance to them. Hybridization has already been found to be widespread among the varieties of mammoths, and it continues to occur between savanna and forest elephants despite what appears to be a multi-million year separation. We might expect the same of other extinct elephant species.

When Eleftheria Palkopoulou presented on some of these data at a conference in 2016, she did talk about hybridization. Ewen Callaway reported on that conference presentation at the time: “Elephant history rewritten by ancient genomes”.

Palkopoulou and her colleagues also revealed the genomes of other animals, including four woolly mammoths (Mammuthus primigenius) and, for the first time, the whole-genome sequences of a Columbian mammoth (Mammuthus columbi) from North America and two North American mastodons (Mammut americanum). The researchers found evidence that many of the different elephant and mammoth species had interbred. Straight-tusked elephants mated with both Asian elephants and woolly mammoths. And African savannah and forest elephants, who are known to interbreed today — hybrids of the two species live in some parts of the Democratic Republic of Congo and elsewhere — also seem to have interbred in the distant past. Palkopoulou hopes to work out when these interbreeding episodes happened.

None of these scientific results concerning interbreeding and hybridization are in the new paper by Meyer and colleagues. So I expect we will see much more from these new genome sequences.