Australasian Science: Australia's authority on science since 1938

Ancient Swingers

Nimbadon skulls representing developmental ages

AL90 has produced dozens of exceptionally well-preserved Nimbadon skulls representing developmental ages from pouch-young (left) to mature adults (right). Credit: K.Black/H. Godthelp

By Karen Black

Skeletons unearthed in a fossil cave in north-western Queensland reveal that 15 million years ago Australia’s ancient forest treetops were home to mobs of 70 kg wombat-like marsupials – the largest tree-dwelling marsupial herbivores to have ever lived.

Entombed in a fossil cave in remote north-western Queensland lie the skeletons of hundreds of bizarre as well as familiar extinct creatures that occupied the ancient forests of Australia 15 million years ago. The World Heritage-listed Riversleigh cave known as AL90 is providing a unique window into our ancient past at a critical time in Australia’s history when our continent began its transformation from a land dominated by lush forests to the cooler, drier, climatically unpredictable habitats we know today. It was a time of significant faunal turnover that saw the extinction of a range of archaic mammal groups and the emergence of Australia’s iconic and globally distinctive marsupial fauna.

One of the greatest success stories of Australia’s climatic transformation was the spectacular radiation of kangaroos, with 70 living species that have evolved to occupy a wide variety of habitats ranging from deserts to wet tropical rainforests. Among their ranks are Australia’s largest living terrestrial (e.g. red kangaroo) and arboreal (e.g. Bennett’s tree kangaroo) mammals, with large males of both species clocking in at 85 kg and 14 kg, respectively.

Prior to about 30,000 years ago, the largest mammals in Australian ecosystems were not kangaroos but a group of wombat-like herbivores called diprotodontoids. These marsupials are common and widespread in fossil assemblages throughout Australia and New Guinea, and have a fossil record dating back at least 25 million years with over 38 species documented.

The most famous was the largest marsupial ever to have walked the earth, Diprotodon, which stood 1.8 metres at the shoulder, weighed in at a massive 2.5 tonnes (AS, June 2008, pp. 20–22) and was Australia’s answer to the African rhinoceros. In fact, diprotodontoid species, which range in body size from approximately 50 kg to 2.5 tonnes, are often colloquially referred to as marsupial rhinos, marsupial sheep, and, for some of the more bizarre trunked forms, marsupial tapirs.

Because of their large size and abundance in the fossil record, it has generally been accepted that most diprotodontoids were terrestrial and roamed in herds or mobs like modern-day kangaroos. But a recent study of skeletons of a small diprotodontoid called Nimbadon from AL90 cave has turned many of these long-held preconceptions about the lifestyles of these ubiquitous marsupials literally on their heads.

The AL90 Cave

My first trip to AL90 was in 1994 as a PhD student commencing a study of the diverse array of Riversleigh diprotodontoids. I could barely contain my excitement – my study site was to be a fossil cave deposit that was millions of years old. I imagined clambering into subterranean depths riddled with glistening crystalline stalactites and stalagmites and searching for the skeletons of beasts never before seen by human eyes.

My illusion was shattered when instead I was greeted by a large limestone ledge and a small pile of rubble from the previous year’s field trip that had since been cemented into a termite mound.

AL90 was, in fact, a fossilised cave where the roof has eroded away and the cave has undergone cycles of infill and erosion over millions of years. What we see today are remnants of the cave floor, walls and cave-fill sediments with former cave structures such as flowstones and stalagmites preserved, along with the fossils, in solid limestone, including the skeletons of thousands of tiny bats that had once roosted in the cave 15 million years ago.

Armed with a light explosive, our team from the University of NSW began to delicately section the large limestone ledge to reveal what lay below the myriad bat bones. As the dust settled after the initial blast, the truly spectacular nature of AL90 was revealed.

The explosives had cracked the ledge along a weak point in the limestone which just happened to be through a complete Nimbadon skull that was now sliced into two perfect halves. Each successive blast revealed more bounty, including articulated Nimbadon skeletons with the bones still lying in the position in which the animal had died.

Several partial skeletons and a phenomenal ten skulls were recovered that year. In one day at AL90 we had more than tripled the number of diprotodontid skulls that had been collected in more than 25 years of field work at Riversleigh.

With each subsequent field season, AL90 began to look more and more like the cave it actually was. It now measures 8 metres long by 5 metres wide and 2 metres deep, and is by far one of the largest excavations at Riversleigh.

Although it would seem to be a diprotodontoid graveyard, a range of other mammals have been found. Several skeletons of a small, primitive galloping kangaroo have been recovered as well as that of a fox-sized thylacine that appears to have been an agile tree-climber, similar to modern-day quolls. Other mammals include tiny koalas one-third the size of their living relative, as well as bandicoots, ringtail and brushtail possums, and six species of leaf-nosed bats. Frogs, skinks, agamid (dragon) lizards, snakes and several species of bird, including honeyeaters, a flightless rail and a small emu-cassowary cousin, are also present.

The Nimbadon Fossil Sample

To date, more than 30 Nimbadon skulls and numerous skeletons representing developmental ages from tiny suckling pouch-young to elderly adults have been recovered from AL90, making it arguably one the best-represented fossil marsupials known.

The exquisitely preserved developmental series of skulls has allowed the first detailed study of growth in a fossil marsupial, and has revealed that Nimbadon developed in much the same way as modern kangaroos. Early growth of the skull was centred on the bones surrounding the mouth, a phenomenon that allows the baby marsupial, which is born at a relatively embryonic stage, to attach to the mother’s teat and begin suckling.

Nimbadon possessed low-crowned lophodont teeth, which indicate it was a browser on relatively soft vegetation. Its skull, although relatively primitive in form compared with other diprotodontoids, was unusual in having more forward-facing eyes and a short, bulbous snout with inflated nasal sinuses. Both of these features hinted that Nimbadon was interacting with its environment in a manner quite different to other diprotodontoids.

It was not until we began to analyse the skeleton that the truly unique nature of Nimbadon came to light. An animal’s skeleton is uniquely adapted to the environment in which it lives and the activity it performs within that habitat. Whether it be swinging through the tree tops, running across open plains or burrowing underground, an animal’s locomotory repertoire is reflected in structural differences in the skeleton. The length of bones, how they articulate with one another, their range of motion and the size and distribution of areas for muscle attachment can tell us volumes about an animal’s lifestyle. By comparing fossil skeletons with those of living animals of known habitat and lifestyle, we can make relatively accurate deductions about the lifestyle and ecology of a long-extinct beast.

We compared the Nimbadon skeleton to those of a range of modern and extinct marsupials of varying lifestyles, including its closest living relatives, the wombats, as well as the koala, the brushtail possum, bandicoots, marsupial lions and other diprotodontoids. Strikingly and somewhat unexpectedly, we found that Nimbadon’s skeleton, in particular its limbs, hands and feet, were most similar to the living koala.

Like the koala, Nimbadon possessed powerful forelimbs with highly mobile shoulder and elbow joints that would have allowed significant extension and rotation of the arms, which are essential for navigating through a complex three-dimensional landscape such as a tree canopy and for balancing and maintaining pressure on a branch. Nimbadon’s hands and feet were extremely large, with long, flexible fingers and toes and semi-opposable first digits. Combined with the deep carpal tunnel of the wrist (to house the tendons of muscles that flex the fingers), these features suggest that Nimbadon possessed an exceptionally powerful grasp.

But perhaps most striking of all of Nimbadon’s adaptations for life in the trees were its massive, sharp, recurved claws on both its hands and feet. At the base of each of these claws was a large process for the attachment of tendons that, when flexed, were capable of applying a powerful force to deeply penetrate the tree trunk during climbing.

Our study showed that claws of this kind were only found in species that live in or regularly climb trees. In fact, Nimbadon’s claws are identical in shape and structure to those of the koala, albeit far larger. The numerous similarities between koala and Nimbadon skeletons suggests they functioned in much the same way and that, like the koala, Nimbadon also adopted a trunk-hugging method of climbing.

Interestingly, however, we found that Nimbadon was unique in having the shortest hindlimbs relative to its forelimbs of any other marsupial. These same limb proportions are found today in animals such as sloths and some apes that regularly suspend themselves from tree branches using their powerful forelimbs. A shortened hindlimb may function to reduce body mass and consequently the energy expended during these suspensory behaviours. So it would appear that Nimbadon was not only a tree-hugger, but also more than likely a tree-hanger, a behaviour no longer regularly used by any marsupial.

Body weight estimates for Nimbadon range between 54 kg and 88 kg. While the thought of a 70 kg wombat-like arboreal marsupial may take some getting used to, large arboreal or climbing placental mammals are well-known. The largest living arboreal mammal is the orangutan, with males reaching weights of up to 118 kg in the wild.

Bears are habitual climbers, with many of the smaller rainforest species spending a considerable amount of their time in trees while some even build nests. The Malayan sun bear, for example, is the smallest (27–65 kg) of living bears and inhabits the lowland tropical rainforests of South-East Asia, Sumatra and Borneo. Although sun bears primarily feed on insects, a significant proportion of their diet is fruit, particularly figs, making them significant seed dispersers in their rainforest habitats.

It is quite possible that, like the sun bear, Nimbadon may have supplemented its diet with fruit and possibly played a role as a large seed disperser in Australia’s ancient forests. Perhaps the characteristic features of Nimbadon’s skull, such as the bulbous snout and enlarged nasal sinuses, may be related to the enhanced olfactory capacity needed to detect rainforest fruits. In any case, Nimbadon’s climbing ability would have allowed it access to multiple layers of the rainforest canopy for not only food (thus reducing competition for resources with terrestrial kangaroos) but also for protection from predators.

So how did so many tree-dwelling Nimbadon come to rest in an ancient cave? There are several clues hidden within the structure of the fossil deposit and within the nature of the fossils themselves. We have been able to determine that AL90 cave’s entrance was vertically orientated, with its opening at ground-level. The entrance was more than likely obscured by vegetation and would have acted as a kind of pit-fall trap, with unsuspecting animals dropping in through the entrance. A similar method of entrapment can be seen by the remains of animals found in the numerous sink-holes of the limestone karst terrain in the Riversleigh region today. Nimbadon was probably quite cumbersome on the ground and while travelling from one tree to another may have fallen easy victim to such a trap.

The abundance of larger animals in the AL90 fossil deposit, such as kangaroos and Nimbadon, compared with smaller species such as possums and bandicoots, suggests that these smaller animals may have been able to escape the cave, perhaps by scaling the walls or through a smaller, secondary opening.

The completeness and articulated nature of the Nimbadon fossil skeletons indicates that these animals died in the cave rather than simply having their bones washed in through the cave entrance. In fact, the position of a number of skeletons within the fossil deposit suggests that some individuals survived the fall, only to curl-up and die from starvation in some dark recess along the cave wall.

It is difficult to tell whether the skeletons accumulated in the AL90 cave over seasons, decades or even thousands of years. Accumulation over a relatively short period of time would suggest that Nimbadon was relatively common in the surrounding environment and raises the intriguing possibility that Nimbadon may have roamed in mobs or large family groups within the tree-tops.

The Demise of Nimbadon

While the exact cause of the disappearance of Nimbadon from the fossil record is not known, our understanding of modern ecosystems has shown that large-bodied, highly specialised species with limited niche breadth are highly sensitive to habitat change. The progressive drying and cooling of the Australian continent beginning in the middle Miocene around 15 million years ago resulted in the spread of more open forests in central and northern Australia and the retraction of rainforest to the coastal margins. The primary habitat of Nimbadon was being lost, with rainforested areas possibly too fragmented to sustain the requirements of such a large arboreal species. While several diprotodontoid species survived and thrived during this climatic transition these were larger, generalist, terrestrial species that show numerous adaptations in their skulls and dentitions (including larger, more complex, higher crowned teeth with thicker tooth enamel) in order to process the tougher, drier plant material associated with the gradual aridification of northern and central Australia.

As unique and ecologically diverse as our fauna is today, the fossil record indicates that modern Australian ecosystems have suffered an even greater loss in ecological diversity than previously expected. Understanding the nature and rate of change in past palaeocommunities, such as what has been documented at AL90, is increasingly important for recognising and understanding future challenges likely to be faced by Australia’s surviving biota in the face of ongoing climate change.

Karen Black is an Australian Research Council Postdoctoral Fellow in the School of Biological, Earth and Environmental Sciences at the University of NSW.