Header image

john hawks weblog

paleoanthropology, genetics and evolution

Photo Credit: Elephants. Joanne Goldby CC-BY 2.0

Hybrid origins of the straight-tusked elephants

Elephants are one of the most important comparisons for human origins. Like humans, they’re long-lived animals that have complex social behavior, they require extensive home ranges and sometimes migrate over long distances.

What genetics has discovered about their evolution and diversification over the last few years provides some fascinating parallels to human evolution in the Pleistocene. Mammoths are their own fascinating story—I wrote about them several years ago, and again in 2016, and the story continues to develop.

But in the last year, the other ancient elephants have been at the forefront of new discoveries. In particular, the “straight-tusked elephant”, Palaeoloxodon antiquus, has yielded two ancient genomes that have disrupted what paleontologists thought they knew about Pleistocene evolution.

Straight-tusked elephant skull and reconstruction from Torralba, Spain
Straight-tusked elephant skull and reconstruction from Torralba, Spain. Photo: José-Manuel Benito via Wikipedia

Last summer I wrote about sequencing work on the ancient straight-tusked elephant: “Genomes of straight-tusked elephants”.

At that time, Matthias Meyer and colleagues had demonstrated that the genomes of two individuals of Palaeoloxodon antiquus from Germany were closer to African forest elephants (Loxodonta cyclotis) than savanna elephants (Loxodonta africana).

That was newsworthy. Paleontologists had mostly thought that Palaeoloxodon was related to the Asian elephant and mammoth clade. It turns out that it’s one of the African elephants. The result also emphasized the deep phylogenetic separation of the savanna and forest elephants in Africa. Those two living African elephant populations, once assumed to be part of a single species, are substantially different genetically from each other, perhaps as much as mammoths and Asian elephants were.

It also raised new questions about the relationships of the extinct African lineage, Palaeoloxodon recki. Once widely known as Elephas recki, this was the major component of the elephant fauna in the Pleistocene African fossil record. Today’s savanna elephants, L. africana, are virtually unknown through much of the African Pleistocene. Nobody really knows where the living lineage may have been hiding, nor does anybody know why P. recki might have become extinct.

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

In my post last June, I hinted that there might be more to the story. Ewen Callaway had reported on a conference presentation by Eleftheria Palkopoulou that discussed evidence for hybridization among these ancient elephants: “Elephant history rewritten by ancient genomes”.

Now, Palkopoulou’s analysis has been published in PNAS: “A comprehensive genomic history of extinct and living elephants”. The title is a bit overblown in my opinion, because I have many questions that the new paper doesn’t answer. But the paper does add two important details to Meyer’s results from last year.

First, Palkopoulou and colleagues show that the straight-tusked elephant genome from Neumark-Nord, some 120,000 years old, is not a simple branch of the elephant phylogeny. This individual’s ancestry derives mostly from a branch that stemmed from the common ancestors of savanna and forest elephants. But it also has substantial ancestry from woolly mammoths. And up to a third of its genome came from a population genetically similar to today’s forest elephants from Sierra Leone, in West Africa.

A third of the genome is pretty high to be interpreted as a “ghost population”. The straight-tusked elephant population of Europe in the early Late Pleistocene was apparently a mixture of two source populations, one with a long independent evolutionary history, and one with continuing strong genetic connections to African forest elephants.

This strong African forest connection was not with every population of forest elephants. The Sierra Leone L. cyclotis individual in the study bears strong similarity to the ancient straight-tusked elephant, but the Central African Republic-sampled L. cyclotis genome does not.

Second, Palkoupoulou and coworkers used a combination of analyses to jointly examine the effective population sizes of elephant species and genetic divergence times between them. There are lots of details in this analysis, with so many lineages sampled, and I wouldn’t trust many of these details too far until more individuals are added to the African elephant and Palaeoloxodon samples.

Still, these analyses reinforce what the evidence for introgression shows. The two sampled forest elephants demonstrate a long divergence, with an estimated divergence time between 463,000 and 609,000 years ago for the populations that these two sampled individuals represent.

These two populations of forest elephants from different parts of Africa are around as different from each other as Neandertals and Denisovans were.

Clearly, we are not going to understand the evolution of the forest elephants, or their connection with straight-tusked elephants, until geneticists sample a lot more of them. Two genomes from each are not enough.

Another detail of the new analysis bears upon the long-time absence of the savanna elephant from the Pleistocene African fossil record:

The two savanna elephants had lower Ne relative to forest elephants for hundreds of thousands of years (Fig. 4D), potentially reflecting ecological competition from the African elephant Palaeoloxodon recki (including Palaeoloxodon iolensis) that dominated the African savannas until the Late Pleistocene (2, 19), or the high levels of male–male competition documented in this species.

That’s a possibility. I find it fascinating that the savanna elephant lineage is very ancient indeed, separated from forest elephants for the last 2 million years. The identity of the P. recki population remains obscure, and the great difference between today’s forest elephant samples suggests that a better sample of elephant DNA from across Africa may yield additional genetically differentiated lineages. It’s even conceivable that some lineage of forest elephant might turn out to be a close relative of P. recki or P. iolensis.

Or then again, maybe P. recki will turn out to be a true ghost, not closely related to P. antiquus at all. As I wrote last year:

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.

I just want to reiterate this sentiment. Discovering that P. antiquus isn’t what paleontologists once thought does not give me confidence that paleontologists really know where P. recki or P. namascus belong. For that matter, I have no confidence that P. recki within Africa is really a single lineage. Until recently, most biologists considered L. africana and L. cyclotis to be a single lineage.

I can’t wait to see results from a bigger sample of ancient elephants. The story of the straight-tusked elephants is likely much broader than two German skeletal samples. There are many challenges to ancient DNA study in temperate and low-latitude situations, but if there’s any species with plenty of tissue to sample, it should be elephants.