Sunday 7 July 2024

Giant Kangaroos: Were They Utterly Hop-less?

Most people have a clear image of what a kangaroo is: a large herbivorous animal that carries its young in a pouch and that moves by hopping about on its hind legs. There are three living species of true kangaroo, and this description does, indeed, accurately fit all of them. However, the kangaroo family, or Macropodidae, is much larger than this, with a total of 63 species, most of which can generically be described as "wallabies". 

There are also several fossil species known, stretching back to the late Oligocene, over 25 million years ago. Having only skeletal remains, and often partial ones at that, while we know that they had sufficient similarities to be placed in the same family and their teeth indicate they were herbivorous, what about the other two features: pouches and hopping? 

Here, one approach is to use the concept of "extant phylogenetic bracketing". The idea behind this is that you take the living members of a group of animals that have some particular feature in common and arrange them on a family tree. Then you look to see where the fossil animal you are interested in fits in that family tree, based on skeletal similarities. (Literally hundreds of them, if you're doing this properly, which requires some significant computer power). If the fossil animal is "bracketed" by living species that all had this feature, then, unless there's some good reason to suppose otherwise, it probably did, too.

So, for instance, if we find a fossil zebra, and it's clearly descended from the last common ancestor of all living zebras. then it probably had stripes. We can't know that for sure, because it might be the one exception, but it seems pretty likely.

Using this method, we can be confident that extinct members of the kangaroo family did, indeed, have pouches. Not only is this true of all living species in the family, it's true of virtually all living marsupial species in any family; fossil kangaroos are so far inside the phylogenetic bracket that it would be truly astonishing if they didn't have pouches. Evolutionarily speaking, even if the pouch doesn't fossilise, it'd hard to see how it possibly be missing. (That the epipubic bones that help support the base of the pouch are present is also a bit of a clue, although technically not proof, since pouchless male kangaroos have them too).

But what about hopping? You might take a look at wallabies and say "well, they all hop, so surely fossil relatives must have, too?" Except, well... it rather depends on what you mean by "hop".

It is true that all living members of the kangaroo family are capable of bounding along the ground bipedally when they are moving fast. For that matter, so can all the members of the closely related potoroo family (Potoroidae). But that doesn't necessarily mean that this is their preferred method of locomotion. If we take a look at the macropodid family tree, what we get is as follows. Here, animals that regularly hop are shown in blue, and extinct groups, with uncertain locomotion, in grey. The others can hop, if they really must, but they generally don't; tree kangaroos typically just walk, especially when they are on a branch, while quokkas prefer to use all four legs to bound along, rather than just their hind ones. (Some potoroo species do this, too, while the next family out in the tree is physically incapable of hopping at all).

So what we see is a slightly unclear pattern. The common ancestor of the family could probably hop, but some of its descendants developed this as the primary means of locomotion, while others either didn't or did at first but then lost the ability somewhere along the way. That makes it harder to say that any given fossil species was necessarily more like, say, a red kangaroo than it was like a quokka.

We also have that key qualifier I mentioned earlier: "unless there's some good reason to suppose otherwise". Hopping is a surprisingly efficient method of locomotion, in terms of energy use, actually being better than quadrupedal motion for medium-sized animals at high speeds. But there are limits, especially when it comes to larger animals. It has been estimated that the optimum size for a hopping animal is around 50 kg (110 lbs). While this is larger than most grey kangaroos, red kangaroo males can reach 90 kg (200 lbs), so it clearly still works at that size, even if this does put some strain on the musculoskeletal system. 

In fact, the maximum size for an animal that hops like a kangaroo is probably around 160 kg (350 lbs), beyond which we suspect their tendons would rupture. This explains why there are no kangaroos the size of cows. Significantly, some of the largest extinct kangaroos, belonging to the short-faced subfamily, reached up to 230 kg (505 lbs). While it was originally assumed, due to phylogenetic bracketing, that they must have hopped, later analysis showed that they can't have done and some recently discovered fossil tracks seem to bear this out.

Not all fossil kangaroos belong to the short-faced group, however. Bohra was a close relative of  the living tree kangaroos and probably had a similar gait, while Congruus is also thought to have been capable of climbing trees, which would rather limit the opportunities for hopping while up there. For the giant kangaroo Protemnodon, however, the situation is less clear.

Protemnodon was long thought to be a close relative of the banded hare-wallaby, the most basal member of the kangaroo family but the recovery of DNA from recent fossils demonstrated that it is a very close relative of the true kangaroos. Phylogenetic bracketing alone would therefore suggest that it should have hopped, especially since at least some species lived in the open, arid environment of the Australian Outback. But, since it weighed up to 130 kg (285 lbs), that may be a stretch... but not necessarily impossible.

Fortunately, because the last giant kangaroos died out as recently as 10,000 BC, many of their fossils are relatively complete. While we can't watch them to see how they moved, we can compare the proportions of their limb bones to those of living species to determine what is most likely. While there have been previous efforts - that have largely failed to come up with a definitive answer - a new one was published a few weeks ago, taking a look specifically at a fossil of the species Protemnodon viator.

The study used principal component analysis to compare the fossil bones with those of both living and other extinct species. This is a mathematical method that boils down a set of multiple different variables to just two, allowing you to plot them against one another in a graph, so that similar combinations (in this case, species) will cluster together. 

When they did this, they found that the known animals formed three distinct groups. One group held all the hopping animals, from the red kangaroo to the tammar wallaby and the rufous rat kangaroo. A second held those kangaroo-like animals that were nonetheless primarily quadrupedal, such as quokkas and tree kangaroos. The short-faced kangaroos, ended up in the third group, reinforcing the idea that they weren't using their hind legs to hop, but to take long bipedal strides. (This probably looked a bit weird to modern eyes, since no living mammal does this).

Having demonstrated that we can predict the motion of living animals by where they fall on the graph, all we need to do is to plug in the figures for the limb proportions of Protemnodon and see which group it falls in. And the answer is... none of them. It has no clear living analogue and doesn't resemble the extinct short-faced kangaroos, either.

This is due, in large part, to a combination of relatively short feet (like the quadrupeds) and long arms (like the bipeds). The former is probably more significant for the gait since it would make it harder to push off from the ground while hopping. So, while it probably could and did hop from time to time, this is unlikely to have been its primary mode of locomotion. Most of the time, it likely walked on all fours, and the higher body weight may help explain the powerful arms, which would often have been needed for support. An alternative suggestion for the muscular arms is that the specimens examined happened to have been male, and used the arms while fighting for mates, as living kangaroos do, but to a greater extent.

One possibility, which the authors argue is supported by some features of the hips, is that the animal's motion was closer to a gallop than a bound, moving the legs separately, rather than pushing off with both at once. The unusual hooked claws may have helped it walk on uneven ground, while the shape of the shoulder joints and arm bones suggests that it may have used the arms for digging, perhaps to get at edible roots, or to den like a bear.

All in all, despite looking superficially kangaroo-like, but with shorter feet, stronger arms, a longer neck, and a much greater size, the giant kangaroo wasn't quite like anything we have today. The 63 living species of the kangaroo family are surprisingly varied, but the extinct species may have been even more so.

[Photo by "Paleocolour" from Wikimedia Commons. Cladogram adapted from Llamas et al. 2015].


  1. One would have thought that humans would have been an obvious example of extant mammals moving by long bipdal strides. But I guess short-faced kangaroos walked in a decidedly non-human manner?

    1. Yeah, I think so. It's the "long" part of it that's odd, as I understand. A human could walk like that if they wanted to, but they normally don't.