Sunday 15 May 2016

Toothless Old Fossils

Artwork depicting extinct animals almost invariably shows them physically fit, unless they're actively engaged in some life-or-death battle. They're usually either adults in the prime of life, or still juveniles (often newborn). And this is actually quite reasonable, since few animals get to live to an old age out in the wild. Most will, after all, be eaten before then, and if they have any serious illness or crippling injury, they're not going to last long. Senescence, the gradual decline of biological function associated with becoming elderly, is not something we see much in the wild, for all that it's common in humans and domestic animals.

Which isn't to say that we don't see it at all, especially in very large animals that have few, if any, natural predators once they reach adulthood. The same must also have been true in the distant past, and it's reasonable to assume that at least some fossils, at least of the larger animals, belong to elderly individuals. It isn't, however, necessarily going to be all that obvious, especially if the skeleton is incomplete in the first place. Nonetheless, a recent report does describe a jawbone that the authors believe belonged to an elderly animal, and, if they're right, this offers some unusual insights into the creature in question.

That's because the animal is Desmostylus, the best known of a group of extinct mammals named after it: the desmostylians. To understand why an example of an elderly desmostylian might be useful, however, we first have to look at what the animals actually were.

Desmostylians were fairly odd animals, from our modern perspective. They lived from about 31 to 7 million years ago, a long timespan, but one that's also a fair distance in the past. They have no living descendants, nor is there anything quite like them alive today.

When they were first discovered and described, back in 1888, they were thought to be sirenians, members of the group that today includes manatees. Now, manatees are themselves fairly odd creatures, fully aquatic animals that, unlike whales, seals, and the like, are grazing herbivores, not marine predators. Their discoverer, O.C. Marsh, only had a few teeth and some bits of the backbone, but based on the shape of those teeth, and the fact that he'd found them in marine deposits, he was confident that he had a herbivorous marine mammal. In which case, it had to be a sirenian of some kind, because what else was there?

It was he, as discoverer, that coined the name Desmostylus, which means "bonded pillar", a reference to the fact that the teeth look as if they're formed from a number of cylinders fused together. Not everyone was wholly convinced that the creature really was a sirenian, but, in general, most palaeontologists agreed, if only because it was hard to see what else they might be. It took until 1941 for someone to find a fossil that included the legs, thus proving that these previously missing body parts were... well, legs, and not flippers. Already raised to the level of a family in 1905, in 1959 R.H. Reinhart made the decision that anything with legs manifestly wasn't an early manatee, and raised them further, to the level of "order" - thus removing them from the Sirenia altogether and giving them their own group.

We therefore now recognise that there are four groups of fully aquatic mammals: the sirenians, the cetaceans (whales, dolphins, and porpoises), the pinnipeds (seals, sea lions, and walruses), and the long extinct desmostylians. Well, okay, and sea otters, but there's only one species of them. Admittedly, there is some question as to whether they were truly aquatic, or just amphibious, like hippos, but that particular debate is a matter for another post.

Instead, the question is where they fit within the mammalian family tree. After all, just because they aren't actually Sirenians doesn't mean that they aren't closely related to them. It's here that the new fossil comes in.

The fossil was uncovered from a rock formation in Orange County, California, not far south of Los Angeles. All the previous fossils described from this formation were either fish or invertebrates, although a number of mammalian fossils apparently still sit in museum back rooms waiting for somebody to write them up. Nonetheless, we have enough to date the deposit to some time around the mid Miocene, 14 to 9 million years ago.

The fossil only consists of part of a jawbone, but that's enough to identify it as Desmostylus, although not to narrow it down to a particular species (at least two, and possibly three, are known). What's significant about it is that, while the jawbone has tusks where they should be, there are no cheek teeth, nor the sockets that you would expect if they had simply fallen out after death. The authors' conclusion is that this was an elderly, toothless individual, that perhaps died of starvation after having found itself unable to eat.

Since most other desmostylian fossils belonged to young individuals, many of them not even fully adult, the researchers were able to compare these, and suggest how these animals' teeth developed throughout life. Infant desmostylians presumably had milk teeth, but these are soon replaced by the 'adult' set, with the first cheek teeth to appear being the premolars. The first true molar tooth appears at around the same time as the last premolars, and, as the animal approaches adulthood, the second molar appears behind it and pushes all the other teeth forward, so that the front-most premolar falls out. By the time full adulthood is reached, the third molar has appeared, pushing the whole chain of teeth forward, with the premolars falling out until only the molars remain.

This, they argue, would have been necessitated by the animal's diet of tough sea grasses, mixed in with gritty silt from the sea floor. This ground the teeth down, so that the animal needed to have the teeth turn up one at a time, so that there is always a fresh, relatively unworn set to use. Until, of course, the desmotylian becomes elderly, and the molars, which can no longer be replaced, wear down and vanish... and the animal dies.

This, of course, is not how mammalian teeth normally work. There's not usually a conveyor belt of new teeth popping up as the old ones wear down; you get one set after the milk teeth fall out, and that's it.

Not usually. Because this is how elephants work. The tusks, of course, are permanent, but behind that, the cheek teeth work in exactly the way described above, new ones pushing the old, worn, ones down until they fall out. Once the last ones have gone, the elephant finally expires, unable to chew its food. Moreover, while elephants are the best known mammals to have this particular pattern of tooth replacement, they are not unique. Sure, there's a wallaby that does it, but more significantly - so do sirenians.

In evolutionary terms, this is not so surprising. We've known since the late 1990s that elephants and sirenians are related, both belonging, along with a number of much smaller and more obscure animals, to a broader group called the afrotheres. Indeed, within the afrotheres, elephants and sirenians appear to be one another's closest relatives. That a feature which is (almost) unique to them is also shared with desmostylians lends support to the theory that these odd creatures really are related to sirenians, perhaps being an early offshoot of the line.

Counting against this, however, are a number of recent studies that use other evidence to argue that desmostylians weren't afrotheres at all, still less close relatives of manatees. Instead, the more modern view has been that they were very early relatives of the group of animals that today includes horses and rhinos. If that's right, then the similar pattern of tooth development is a coincidence, a case of parallel evolution with two groups of herbivorous aquatic mammals independently developing the same solution to the same problem.

That, at least, is the argument of the authors of the paper. But I can't help but wonder, do we really know that this jaw belonged to an old, senescent individual? It probably doesn't affect the argument about horizontal tooth development either way, but there are some oddities in the jaw, with regard to its thickness and the position of the tusk, that hint that it might not belong to any species of Desmostylus so far known.

There's also the question of how the animal survived without its chewing teeth for so long. Because clearly, if the tooth sockets had had time to vanish (as, indeed, they would do eventually) it must have managed to survive for at least a fair while after it lost its teeth. And during that time, it must have been eating something. But if one individual can survive for a time without teeth, it seems to me that it's not impossible that a species could do the same. Perhaps there's some very good reason why this can't possibly be the case, but when all we have is a partial jawbone, can we be absolutely sure that this particular one didn't belong to some unknown species that didn't have teeth, at least as an adult?

Which might sound a fairly odd suggestion, given what we understand of the desmostylian lifestyle. But I'm not sure that it's entirely impossible, and that's because manatees are not the only living sirenians. The other two recent species of sirenian both belong to the dugong family. One, the dugong itself, is still alive, but the other, Steller's sea cow (Hydrodamalis gigas), went extinct in 1768. Although closely related to manatees, and even more so to living dugongs, Steller's sea cow fed on soft kelp, instead of tough sea grass.

Steller's sea cow had no teeth, instead having ridged horny pads that presumably would not survive fossilisation. If horizontal tooth replacement could evolve twice, can we be sure that a sea cow-like lifestyle couldn't do the same? Perhaps it's a stretch, and I'm not even saying that it's particularly likely (it did, after all, have tusks, which the sea cow didn't), but it doesn't seem to me to be beyond the bounds of reason, either...

[Photo by "S. Rae", from Wikimedia Commons.]


  1. So, when in the course of Proboscidean evolution did horizontal tooth replacement arise? I have a vague recollection that it is del after the first fossil Proboscideans known. In which case… even if Elephants and Desmostylians (THANK YOU for posting about them! I've been fascinated by their obscurity ever since I first learned of their existence!) share this unusual scheme, and even if they ARE closely related, their last common ancestor may NOT have had it! (Back in the days of Henry Fairfield Osborne & co -- when serious palaeontologists could talk with a straight face about orthogenesis -- I suppose you could have gotten away with saying that a PROPENSITY TO EVOLVE horizontal tooth replacement was a "shared derived condition," but I think that would be hard to defend nowadays.)

    1. Well, one could easily imagine some genetic prerequisite for horizontal tooth replacement being a tethytherian synapomorphy. The trick would be demonstrating it in the absence of desmostylian DNA.