Oligocene (Pt 17): Dawn of the Kangaroos

Ekaltadeta

Kangaroos are perhaps the single most iconic Australian mammals. As marsupials, we tend to think of them as lost relics of an earlier evolutionary period, and, indeed, they have been around for a long time. Of course, they have been evolving during that time, rather than standing still, but if we had a time machine, we could go back millions of years into Australia's past, and still find animals that were, more or less, kangaroos. But, obviously, there is a limit.

Exactly how far back that limit is partly depends on how kangaroo-like you want your kangaroos to be. But even then, there are some gaps in our knowledge that don't have direct counterparts on other continents. The obvious place to start is with the fossil record, and, here, at least, we can provide a clear answer. The oldest known fossil kangaroos date to around 28 million years ago, towards the end of the Oligocene.

But there is a problem with this. Specifically, 28 million years is the age of the oldest deposits at the Riversleigh World Heritage Site, Australia's largest fossil mammal site. Prior to that, there's a gap until we reach the time of the dinosaurs, from which we have very few fossil sites on the continent. Or, to put it another way, the oldest fossil kangaroos are 28 million years old because, when it comes to Australian mammals, the oldest fossil anything is 28 million years old.

Well, not quite, because there are a few fragments older than that from other sites, but those that belong to marsupials are too incomplete to assign to any living group. 

What we can do, however, is try to make estimates of when various marsupial groups evolved using genetic evidence. Calibrating using the fossils we do have, we can see how long it should have taken for two groups to evolve to their current state from a common ancestor, and therefore how old that ancestor must have been. When we do this, it turns out that the modern kangaroo family originated no more than 30 million years ago, and possibly less, so those oldest fossils probably are very close to the base.

However, that refers specifically to the modern family, and excludes several fossil species that are usually referred to as "kangaroos" as well as some living species such as potoroos, bettongs, and rat-kangaroos. According to genetic estimates, the last common ancestor of all of those lived either in the early Oligocene, or, more likely, in the latter half of the preceding epoch. The actual origin point of the group - the point at which "kangaroos" diverged from possums - would be earlier still, maybe 50 million years ago.

And we just don't have suitable fossils from that time period.

One of the earliest kangaroos we do know of was Cookeroo, which lived just as the Oligocene was fading into the Miocene around 23 million years ago. It was still quadrupedal, and similar in size to today's smallest wallabies (or, say, a Jack Russel). Although it had some skeletal features suggesting a link to the main living subfamily, so many others were primitive that it's more likely to have been outside all of the named subfamilies, probably living before they had diverged from one another. The few other late Oligocene members of the family, such as Wubularoo and Bulungamaya, are similarly difficult to place.

However, evidence that kangaroos more generally must have a history significantly older than this comes from the presence of others that do not fall within the main living family. Perhaps the most significant of these were the balbarids, or "fanged kangaroos". This side-branch of the regular kangaroos, known for having fang-like canine teeth, died out in the Middle Miocene, but all of the known genera date back to the Late Oligocene. These include Balbaroo itself, for which the group is named, and also Wururoo, Ganawamaya and Nambaroo. The latter, known from an unusually complete skeleton, was, like all fanged kangaroos, quadrupedal, but had opposable thumbs, suggesting that it may have been capable of climbing trees.

Perhaps the strangest kangaroo of the time, however, was Ekaltadeta, sometimes referred to as the "killer kangaroo". This is thought to be related to the living musky rat-kangaroo, a rat-sized omnivorous animal that is thought to represent the earliest living branch in kangaroo evolution. It was, however, far larger, although still small for a modern wallaby, and similar to Cookeroo at an estimated weight of around 7 kg (15 lbs). It possessed unusually large and sharp premolar teeth that likely enabled it to feed on flesh, hence the alternative name. While it was probably omnivorous rather than a true carnivore, and likely not large enough to be dangerous to anything bigger than a mouse, the idea of any flesh-eating kangaroo is admittedly an odd one.

In terms of iconic Australian animals, koalas can give kangaroos a good run for their money. They too, date back to the earliest Riversleigh deposits, including Nimiokoala, which is better known from the Miocene. The name literally translates as "excessive koala", and refers to its complex molar teeth. Compared with modern koalas, it was smaller and had a distinct, although not especially long, snout. The shape of its teeth and palate suggest that it had a much broader leaf-based diet than the living species, with specialisation on eucalypts likely first evolving in the group in the Middle Miocene. It also seems to have had less acute hearing, and so may have been less vocal than the living species.

While Nimiokoala survived for rather longer, Lumakoala is known only from the Oligocene, making it arguably the earliest known of all fossil koalas. With primitive teeth that imply a diet of soft leaves and probably the odd insect, it is also the smallest known koala, with an estimated weight of around 2.4 kg (5 lbs 5 oz.). 

The oldest known wombat is Rhizophascolonus from both Riversleigh and a site in South Australia. As the scientific name suggests, analysis of the structure of its teeth suggests that, like modern wombats, it had a diet that included a significant amount of plant roots.

Koalas and wombats are each other's closest living relatives, both belonging to a broader grouping that once included many other families that are now extinct. All of these, too, have their earliest fossils at the end of the Oligocene, or at least the very earliest part of the Miocene. Thus, like the kangaroos, the wombat/koala group must have a much older history to have been so diverse so early on.

Many of these early representatives of their groups are known only from fragmentary remains, often only their teeth, so that we can't say too much about them. The diprotodontids, which later included the "giant wombat" Diprotodon, are represented by Ngapakaldia, which, at around the size of a sheep, was still one of the largest marsupials of its day. The predatory "marsupial lions" include Lekaneleo, about the size of a domestic cat. One estimate shows that it already had, relative to its body size, the most powerful bite force of any known mammal, so, like its later relatives, it was probably already hunting creatures larger than itself.

Other groups did not last so long and so are less well-known, perhaps representing the last survivors of lineages that are otherwise missing from the fossil record. The wynyardiids survived into the Early Miocene, and are sometimes referred to as "tree wombats". Namilamadeta is an early example, and it probably resembled a possum more than a living wombat, although it may have been less arboreal. Their close relatives, the ilariids, are known only from the Late Oligocene, and include only Ilaria and Kuterintja. Two other small Oligocene marsupials, Marada and Mukupirna, are too incomplete to be clearly placed as anything more precise than being relatives of early wombats and currently remain the only known representatives of their respective families.

Other groups of marsupial can also trace their origins back this far. Marlu, Paljara, and Pildra are early possums, all placed within the family to which modern ringtail possums belong. They were much smaller than typical modern species, being thought to have weighed around 200g (7 oz.) and are much more primitive, with notably simpler teeth.

Galadi and Yarala are the earliest known members of the bandicoot lineage, although neither can be placed in the living family. Of the two, Galadi is thought to be closest to the living forms, and has adaptations that suggest it may have been more predatory than they are, although still omnivorous. At a little over 1 kg (2 lbs), it was around the size of the largest bandicoots today. Yarala is more primitive, a representative of some early offshoot of the bandicoot order, although its likely diet, consisting of insects and occasional plant matter, was not very different to that of its living relatives.

Carnivorous marsupials were represented, too, with some of the earliest examples being placed within the same family as the recently extinct thylacines. From what we can tell, the family changed relatively little in form over the following 24 million years, although the Oligocene species were smaller, at roughly the size of a fox. Nimbacinus is better known from the Miocene, where it became relatively predatory, but the Oligocene species is thought to have preyed on smaller animals. Of the other two genera known from the epoch, Ngamlacinus is thought to have been the most purely carnivorous, while Badjcinus had a strong and deep jaw similar to that of a modern Tasmanian devil, suggesting a diet including hard material, such as bones.

Other fossils are harder to place. Yalkaparidon is better known from the Miocene, but an early species did exist at the tail end of the Oligocene. There are enough remains to be confident that it was a marsupial of some kind, but beyond that, it's hard to see which, if any, known group it might belong to, leading to Australian researchers nicknaming it a "thingodont". Its large front teeth indicate that it probably bit into something hard, but the remaining teeth seem to be more adapted for a softer diet. It has been proposed that it might have gnawed into wood to get at tree-boring insects, leading to its description as a "mammalian woodpecker", but without remains of its paws, we don't even know if it could climb trees.

If Australia has limited fossil sites of the right age, Antarctica is even harder to explore, with most of its rock being buried under thick layers of ice. But it seems unlikely that, even this far back, there were mammals there, given that the extreme cold had just begun to set in and there would have been little to eat. The continent already had penguins to be sure, and probably more invertebrates on land than it has now, but that's about it.

The seas, however, were a different matter. With all of the continents now reviewed, it's time to take a look at the marine mammals of the Oligocene, in what will be the final post of this particular series.

[Picture by "Seismic Shrimp", from Wikimedia Commons.]