Tuskless Walruses of Japan

There are a number of mammal "families" that contain just one living species. Of course, what constitutes a "family" of animals is an arbitrary distinction, rather like the one that says Pluto isn't a planet. So, really, all we're saying is that there are some mammal species that, in our subjective and pro-mammalian opinion, are do distinct and unusual that we feel they ought to be placed in a group all of their own. If they were insects, or snails, or something like that, we'd probably feel differently. But they're mammals, like us, so we don't.

One such family is the walrus family, containing - you've guessed it - the walrus (Odobenus rosmarus). Walruses are pinnipeds, which is to say, they are related to seals and sea lions. Just by looking at them, we can tell that walruses can walk on all fours like sea lions and fur seals, but unlike "true" seals, yet lack the external ears that sea lions and fur seals have and true seals don't. When you get to the interior structure of the skeleton, and so on, there are a number of technical differences that tell us that, no, this isn't just our imagination, they really are a bit different from all the others. And, there's, you know... the tusks.

As is so often the case, though, the walrus is merely the last representative of a group of animals that were once more diverse. Looked at over a much longer time period, the walrus family has had many different members, and likely at least as much diversity as, say, the seal family. We know of something like twenty fossil species spread over the last 15 million years or so, and, crucially, they are not all close variants of one another. Just as one example, the tusks are a relatively recent evolutionary innovation, and most of the older species don't have them.

Strictly speaking, the term "walrus" refers only to the living species - it's not the "common walrus" or anything like that, it's just "the walrus", so anything else ought to have a different name. The collective term for members of the family (however we choose to define that) is actually "odobenids". But, for simplicity, I'm going to call the others "fossil walruses". So there.

Walruses today live only in Arctic or near-Arctic waters, but fossil species have been found much further south, with a number, for example, being found in Californian deposits, and a few specimens from places such as Mexico, Florida, and Morocco. Over on the other side of the world, we also know of fossil walruses from Japan. One such, although by no means the first, was described in 2005, and given the name Pseudotaria muramotoi.

The remains of this animal were recovered from a bed of rock about 10 million years old on the west coast of Hokkaido, the northernmost large island of the Japanese chain. At the time, the creature was described as possessing a mixture of advanced and primitive features, suggesting that it represented an intermediary step between the early, quite sea lion-like walruses and the later, more specialised, forms. Then, just a few weeks ago, a second fossil was described from more or less exactly the same deposit, and it turns out to belong to a different species. (Given that it is the same deposit, it's unsurprising that one of the authors of this new paper is also the author of the paper describing the earlier one. The number of people with a specific interest in digging up Japanese fossil walruses is presumably not vast).

So we have two species of walrus living, in so far as it's possible to tell such things, in the same place at the same time. How do they relate to one another, and can they tell us anything about walruses at this time period - one that may have been quite crucial in the history of the group?

So far, only one specimen of this animal has been found. As is usual, it's incomplete, but it does include most of the skull, which is key to being able to prove firstly, that it is a walrus, and secondly, that it's different from any others we know of. On this basis, it has been named as Archaeodobenus akamatsui, meaning "Akamatsu's ancient walrus" in reference to a retired curator at Hokkaido Museum. 
The fossil also includes the lower jaw, and some bones from the shoulders, throat and torso, which allows us to get a reasonable idea of how big the animal was.

The estimate the authors come up with is a little short of 3 metres (10 feet), so this certainly isn't one of those instances where we're dealing with a tiny ancestor of some familiar, but large beast. Indeed, it's not too much smaller than a typical living walrus, although it may well have been considerably lighter, and more comparable to a sea lion in size. Of course, estimating the weight of an animal purely from its skeleton is largely a matter of guesswork, and we don't know, for example, how much blubber it had. Living as far south as it did (and not during an Ice Age), it plausibly had rather less than the living walrus, but it's not as if we know for sure.

The earlier fossil consisted only of a skull and a couple of bones from the neck, so it's even harder to estimate the size of that animal, but it's worth noting that the skulls of the two species are about the same size, so they probably weren't that different overall. But there is a significant caveat here: not all the bones of the skull in the Archaeodobenus specimen had closed up. That means it probably belonged to a sub-adult, which may not have been fully grown at the time of its death. In contrast, the Pseudotaria skull does seem to belong to a fully-grown adult, which might affect how meaningful the size comparison is.


At any rate, both fossils are notably primitive, lacking many of the features that identify later walruses, especially those most closely related to the living form. While most of these features relate to detailed shapes of the skull bones and vertebrae, an observer of the living animal would have immediately noted a much longer snout than in the living walrus, and also larger teeth.

Except, of course, for the tusks. The canine teeth of both fossil walruses were large and pointed, but no more so than in many other carnivorous animals, and it's certainly hard to describe them as tusks. One definition for "tusk" is that it's a tooth that projects outside the lips when the mouth is closed, and these almost certainly wouldn't have fit that description. Overall, the animals would have looked rather like modern sea lions, even though certain features of their skeletal anatomy suggest that they were hardly the most primitive walruses known, and likely at least some way along the line to the later, heavily tusked, forms.

Of the two, the new fossil, Archaeodobenus, has slightly more advanced features than its counterpart. But the two were discovered in the same fossil bed, so it's hardly likely that it is a direct descendant of Pseudotaria. Rather, the two must descend from the same ancestor (possibly through some intermediate steps) and it just so happens that Archaeodobenus developed more "modern" features than its relative.

We don't even know that, had we access to a time machine, we could have gone back to Miocene Japan and seen both species of walrus lounging about on the same shorelines together. It's perfectly possible that they did, and, for example, just happened to eat different foods or hunt in a different way, but, equally, they may have preferred slightly different habitats, or even lived in the same spots 100,000 years apart. Ten million years later, such fine distinctions tends to vanish into one single rock layer, beyond our ability to separate.

But that rock layer bears a little more examination. It is largely composed of a greenish mineral called glauconite, known to be formed in shallow, relatively undisturbed, seabeds. Which, of course, is the sort of place one might find animals like walruses. In this case, however, the rock layers immediately underneath this one suggest a localised retreat of the shoreline, exposing former seabed to the dry air. The glauconite layer, it is suggested, formed when the sea flooded back in over the course of a few hundred thousand years.

The suggestion is that, as the sea retreated, some early walrus, not yet discovered, found that its preferred habitat was shrinking or changing. Some members of that species became cut off from the others, living in separate bays a long distance apart. Over the million years of the retreat, these two groups evolved into different species, and when the seas flooded back in again, those two species ended up living together.

It's highly speculative, and there are a number of other possible explanations. As fossil walruses go, the two species do seem to be fairly closely related, but there could be other reasons for their divergence... although it's as good an explanation as any other at the present time. But, whatever the precise reason, the mere existence of two such distinct, yet related, species living alongside one another indicates that it was a time when a number of new walrus species were evolving. Many other walrus fossils, from other parts of the world, date to around the same time, so it may have been a particularly good time for them, and one where conditions were causing them to diversify.

The few walrus fossils we have older than these two typically show animals that resemble sea lions even more closely than do these two. That's not such an absolute rule that we can have any confidence that Archaeodobenus is at all likely to be the animal that proved the turning point from those animals to ones resembling those we're familiar with. But it was, at least, alive at round about the right sort of time, an example of the kind of change that would, in due course, lead away from fish-eating to the narrower shellfish-based diet of modern walruses.

[Image © 2015, Tanaka, Kohno, distributed under the terms of the Creative Commons Attribution License.]