Sunday 1 September 2019

The Many Kinds of Extinct Marsupial

Simosthenurus, a short-faced kangaroo
Last week, I looked at the diversity of marsupial species that can currently be found in the world and, in particular, how the increasing use of molecular analysis over the last quarter-century or so has improved our understanding of their relationships. We don't have such an advantage when it comes to looking at fossil species and it doesn't help that the fossil record is necessarily incomplete. Indeed, the latter is particularly true in Australia, which, perhaps partly just because it's a smaller continent, doesn't have as wide a selection of rock deposits of the age we're interested in here as some others do.

Nonetheless, because marsupials have been around for a very long time, we do know a fair bit about their fossil history. So, with the aid of the same review that I used last week, let's take a look at some of the kinds of marsupial that are no longer with us.

Not least because recent fossils are more common, and typically in a better state of preservation, than older ones most of the fossil marsupials we know of belong to groups we are still familiar with today. There are many fossil wombats, possums, and koalas, for example. Even here, however, there can be some unexpected diversity, with perhaps the most significant being the short-faced kangaroos. These were often larger than living kangaroos and, according to some analyses, too large to hop, perhaps walking about on their hind legs in a manner not unlike that of humans. They are significant in part because they represent a rare case of an exclusively fossil taxon that has had some of its DNA analysed - this demonstrated that the short-faced kangaroos do, indeed, form a separate branch within the larger kangaroo family, not closely related to any of the living forms.

However, over the years, a number of fossil marsupials have been uncovered that are considered odd enough not to belong to any living family. The balbarids, for example, were previously considered to be odd, primitive, members of the kangaroo family, but have been considered a distinct group since around the turn of the century. They died out over 10 million years ago, but are notable for the fact that they still walked on all fours and the fact that at least one species (Balbaroo fangaroo) had large fangs in the upper jaw.

Aside from the recently-extinct thylacines, however, perhaps the best known of these extinct marsupial families is that of the "marsupial lions". While the last example, which died out not long before the Ice Ages, was the size of a leopard, most were much smaller, often tree-climbing, animals, and none looked much like cats. Despite being predatory, their closest relatives today are the wombats, with the large, rabbit-like teeth of those animals being modified in the "lions" into sharp blades that compensated for their lack of canines.

In fact, the wombats have a number of extinct relatives that don't belong to either of the living groups. These others all seem to have been herbivorous, and they include groups such as the ilariids and wynyardids, which included at least some representatives much larger than living wombats - approaching the size of dogs. Largest of all were the "giant wombats", with one species estimated to have weighed over two-and-a-half tonnes. Despite the name, these are not true wombats, as we know them today, but relatives that diverged over 25 million years ago.

Including all of these groups, there are currently six recognised families of extinct wombat-like animal (in the sense of animals thought to be closer to wombats than koalas are). One of the other two includes just one known species, Marada arcanum, a primitive Oligocene animal that apparently left no descendants. The other includes the rather weird "marsupial tapirs", herbivorous pony-sized beasts that seem to have had a short trunk-like snout and a long, extrusible tongue - at least so far as we can tell from the bones alone.

Other named families of Australian marsupial are less well-known, often being described only on the basis of a few teeth or a partial jaw. These include the "sprite possums", mirilinid possums, and miminipossums, and, known from a more complete skull, the small, short-snouted Yarala, which also gets its own family. The first three of these are, as the names suggest, related to possums, while the last is closer to bandicoots. The odd "hammer-toothed marsupial" Malleodectes, is known to be related to the carnivorous branch of the Australian marsupials, although exactly how is unclear. Relatively small, its flattened teeth suggest it crunched up something hard; since it seems to have been a carnivore, our best guess is snails, but we don't really know. Another animal with flattened teeth, Numbigilga, is even harder to place; it may be something to do with bandicoots, despite being a herbivore that may have fed on something like nuts.

Yalkaparidon is another oddity, sufficiently strange that it is considered not only to belong to its own family, but to be the only known representative of an entire order of marsupials that no longer exists. It's perhaps more related to the modern herbivorous species than to the carnivorous ones, but even that is unclear. The peculiar shape of its teeth and skull suggest that it may, like the living aye-aye of Madagascar, have fed on wood-boring larvae.

The further we go back in time, the harder it is to place fossils within the relevant family tree, for the simple reason that the animal in question may have lived before the branches we know of split apart. Because Australia has so few fossil sites from the first half of the Age of Mammals, there are precious few fossils that fit this description on the continent. Among the exceptions is the oldest known undoubted marsupial, Djarthia, which lived in eastern Queensland over 50 million years ago. So primitive that it was initially hard to distinguish from the South American marsupials of the day, more complete fossils have recently confirmed that it was related to the living Australian marsupials, rather than representing a separate immigration event that subsequently failed. It likely lived not long after such animals first entered the continent.

They did so from South America, which has a rather longer marsupial fossil record than Australia does. Given the arrangement of the continents at the time, they must have done so by crossing Antarctica, before that continent froze over. Despite the obvious difficulty of collecting them, we do, in fact, have some mammalian fossils from this period of Antarctican history. None of them are definitively close to Australian forms, although the possibility has been raised. On the other hand, some of the fossils that can be identified are close relatives of the Monito del Monte, a South American animal which (as I noted last week) is more closely related to Australian species than to its own neighbours.

While there are fossil opossums and the like in South America, when it comes to other groups from that continent, we run into a question of terminology. This is because the term "marsupial", as used by biologists, strictly speaking refers only to members of the crown group. That is to say, an animal is only considered a marsupial if it is descended from the last common ancestor of all living marsupials. If its ancestors branched off before that common ancestor evolved then, technically, speaking, it can't itself be a marsupial.

Unlike the Australian species, many South American fossils fall into this category - animals that are more closely related to marsupials than to anything else alive today, but that are not themselves marsupials. The collective term for these animals - marsupials, plus their close relatives - is metatherian. This is used in distinction to the eutherians, which is the group containing the placental mammals and all their extinct relatives. (Monotremes, of course, are neither and, if you go far enough back in the fossil record, there are other exceptions, too).

Looking at all of those animals would lead us down an entirely separate rabbit-hole. We do have a number of fossil opossums, as well as relatives of the modern shrew-opossums, but there are no generally accepted families other than these two that we can definitely say count as true marsupials. One possible exception are the animals known are argyrolagoids, small jumping animals with ever-growing teeth that rather resembled rat-sized jerboas. Although usually considered to be non-marsupial metatherians, it has recently been proposed that they might be true marsupials, probably related to the shrew-opossums.

While the most spectacular non-marsupial metatherians - the large carnivorous forms such as Borhyaena and Thylacosmilus - are all South American, examples are known from every continent except, ironically, Australia. With all their relatives on that continent, as well as the oldest living branches of the group, it seems likely that marsupials proper evolved in South America. While everyone agrees that Djarthia is a marsupial, there is less agreement as to the oldest South American marsupial, since it's not always clear exactly where some of the early metatherians from that continent fit in the family tree; some candidates are older than Djarthia, some are not.

At any rate, the first metatherians hailed from the Northern Hemisphere, and probably only entered the south after the extinction of the dinosaurs. Their history stretches back much further than that, with fossils dating back 110 million years to the middle of the Cretaceous period, and some molecular estimates giving an even earlier date for their origin.

Since then, eutherians seem to have mostly had the upper hand. But even so, the metatherians, marsupials included, seem to have done well enough by themselves.

[Photo by "Ghedoghedo", from Wikimedia Commons.]


  1. Metatherians and hadrosaurid dinosaurs had similar distribution. They probably followed the same route from North to South.

    1. That seems likely, although, obviously the hadrosaurs were a lot earlier, since there are no known metatherians from Mesozoic Gondwana. Interestingly, though, we do know of some fossil metatherians from Eocene Africa (e.g. Peratherium) but not, so far as I am aware, any African hadrosaurs.

  2. You've made my day with Balbaroo fangaroo. Many thanks