by Tim Flannery
So let us imagine Australia 4000 years ago. Some indication of what it was like can be gained from the historic situation in Tasmania (which dingoes never reached), and by looking at the impact of dogs today. From their arrival in 1803, Tasmanian settlers noted the abundance of marsupials, and in distinct contrast to the early mainland settlements, they gained much sustenance from wallabies, possums and kangaroos. Tasmania continues to be famed for its abundant wildlife, and marsupials are still part of the local cuisine.
And what of the scarcity of kangaroos on the dingo’s side of the dog fence? Surely this is a fine demonstration of the ability of dingoes to affect kangaroo numbers. Further clues can be found in remembering the naivety of island-dwelling bettongs which have remained innocent of dogs because of geographical isolation from 10,000 years ago, 6000 years before the dingo arrived.
Such evidence suggests that many marsupials were far more prevalent in Australia 4000 years ago. After 42,000 years of human hunting, a balance had been struck between the marsupials and human predators. Then man’s best friend arrives. With its exquisite nose the dog can locate wallabies, bandicoots and kangaroos effortlessly and, although the thylacine may have preconditioned them to the presence of a dog-like carnivore, the smaller marsupials had not yet learned to fear the scent and sight of a dingo. So they sit still and are killed, and the meat they provide is shared with the dog’s adopted human family. Certain types of game, such as the Tasmanian native hen, were driven to extinction before they could learn to flee, but many others were simply made more scarce, just as foxes and perhaps cats have made many species rarer in historic times.
But what effect could this process have had on people? Did over-harvesting the marsupials provide the Pama-Nyungan speakers and their dogs with an advantage? Imagine the boost given to a clan that could harvest meat twice as rapidly as its neighbours. They would dominate and spread, especially if, at the same time, the dogs were depriving neighbouring people of food by devastating the standing stock of marsupials. We do not know if it was the people or only their languages that spread, yet the vanished languages of southern Australia appear to have been as much victims of the arrival of the dog as were the thylacine and Tasmanian devil.
So what of the factors mentioned by Jones and Evans—changes in tool technology, plant use, and the increased size of gatherings for ceremonies and other purposes? A starting point is to look at the changes that occurred earliest—around the time the dingo arrived. The most noticeable change in the archaeological record at this time is the spread of the ‘small tool tradition’. This appears in northern Australia first, before spreading south. Interestingly, a similar development in tool tradition occured in southeast Asia at around the same time, so perhaps these tools represent another Austronesian-conveyed item which spread. There is no evidence that these tools offered any practical benefit; their spread may simply have been a matter of fashion. Aborigines also began to consume cycad fruit, which are toxic and require an extensive pre-treatment, a practice also known in parts of Africa, India, Guam and the Ryukyu Islands. Cycad fruit treatment is thus also a candidate for transfer by the Austronesians, but in this case the advantage is clear—in historic times cycad fruit fed large gatherings, enabling extended ceremonies to occur.
Two other changes came much later. In a comprehensive study of the archaeological record, Monash University archaeologist Bruno David has shown that by 1000 years ago Aboriginal people had begun to use large, flat grindstones to process grass seed. This innovation occurs around the time that trade increases. One trade item that is readily preserved in the archaeological record is large stone points. These were used to tip spears and daggers, and suitable stone for them was quarried at only a few locations in northern Australia. Again, they only appear during the last thousand years or so.
So how, after an occupation stretching back 46,000 years, are we to account for this sudden propensity by Aboriginal people to eat grass seeds and to trade? I feel that the dingo was, once again, the cause. Wherever herbivore numbers are great, grasses forgo sexual reproduction because their sexual organs are nibbled off before they can ripen. Instead they spread by underground runners. With kangaroo and wallaby populations suppressed by dingoes, perhaps Australia’s grasses were flowering and setting seed more successfully than ever before, making it worthwhile to harvest the seeds and grind them between flat stones.
In areas where there are no cycads, seed-cakes played a vital role in feeding participants at large ceremonies. They also made travel (and thus trade) easier, for they are durable and can be carried on long journeys. At base lay a shift in the relationship between the sexes, for grinding seed was women’s work, yet the product of their labour was a storable foodstuff that could be appropriated by men to provision the great ceremonies so central to Aboriginal life.
In this dingo-driven revolution we see a profound restructuring of Australia’s ecosystems and human cultures, which involved a further diminution of the role of large herbivores, and an increase in human population fuelled by harvesting newly available plant foods. This was a dramatic departure from what had gone before.
21
The Age of Mammals in Australia
Australia’s prehistory is as much about ‘black holes’ as it is about concrete knowledge. So vast are the gaps in the record that whole mammal families may have come and gone without our knowing. It’s like having a jigsaw when only one piece in ten is preserved. But in trying to understand Australia’s past, we must work with what we’ve got. Our current knowledge allows five basic stages to be discerned in the evolution of Australia’s mammals. These are:
1) The age of dinosaurs (between 100 and 115 million years ago), when Australia was inhabited by mouse- to rat-sized placental mammals and mouse- to cat-sized monotremes, but no marsupials.
2) The Murgon age (somewhere between 65 and 54 million years ago), when the last of Australia’s ancient, terrestrial placentals lived, and when the first (mostly rat-sized) marsupials had arrived from South America.
3) The age of the koala beasts (somewhere between 40 and 20 million years ago), represented by the Lake Pinpa fossils, when large ground-dwelling koala-like marsupials, and koalas, dominated the herbivore niche.
4) The age of diprotodontids (about 20 and 5 million years ago), as documented at Riversleigh and in central Australia, when wombatlike species were the dominant larger herbivores.
5) The age of kangaroos (5 million years ago to the present) when kangaroos proliferated into a variety of ecological niches.
One hundred and fifteen million years ago Australia was still part of the supercontinent Gondwana, and its flora and fauna were more cosmopolitan than today. The bones of dinosaurs are roughly a hundredfold more abundant in these deposits than are mammals, with only two basic types—monotremes and placentals—represented in Australia.
The monotremes are known from the jawbones of three kinds of animals—different enough to represent three separate families. One of these was a platypus-like species, a second was equally large with teeth superficially like a sea otter, while the third was a very primitive type about the size of a shrew. The larger monotremes were adapted to life in water, most likely rivers and lakes, and this humble triumvirate is the greatest diversity that monotremes ever achieved—a golden age attained while dinosaurs still ruled the Earth. The monotremes are an ancient group that with a single exception is known only from Australia. The exception is a platypus whose 63-million-year-old teeth were discovered in Patagonia. It seems to have migrated there from Australia at the very beginning of the age of mammals.
The discovery of placental mammals in Australia was, in contrast, a great surprise—even a shock—to the scientific community. This can best be understood by looking at the position occupied by Australia 100 million years ago. Then, there were two ‘supercontinents’, both of which would have been around 70 million square kilometres in extent (Eurasia, the largest modern continent, is by comparison 54 million square kilometres). Laur
asia was located in the northern hemisphere and consisted of what would become Asia (minus India, North America and Europe). The other continent, Gondwana, was located in the southern hemisphere and comprised the remaining major landmasses except South America (which for much of the age of the dinosaurs was the most isolated of all continents).
Until 1997 it was thought that the placental mammals had evolved in the northern hemisphere, where the great majority of species live and where the oldest fossils were found. But that year, after decades of fruitless excavation in the farthest corners of Australia, Nicola Barton, a member of Tom Rich’s field crew who was working at a place called Flat Rocks, 100 kilometres from Tom’s home in Melbourne, unearthed a jaw so small that half a dozen could sit on your thumbnail. On 8 March—the day of the discovery—the field crew threw a party, and without telling Tom of the discovery, invited him down. After placing the minuscule lump of sediment containing the jaw under a microscope, Tom sat stunned amid the merrymakers, repeating to himself, ‘Oh my Lord… Oh my God.’ His emotions were understandable: the Riches had long ago remortgaged their house to fund their search, and after twenty-three years of living his dictum about having ‘the will to fail’, Tom had been running on empty.
In the years that have passed since this discovery, over twenty more specimens have been unearthed, making Flat Rocks one of the most productive early Cretaceous mammal sites in the world. No matter that all these scraps would not fill a matchbox—these are the most significant mammal fossils ever recovered from Australia, for they have resulted in the overturning of a venerable scientific hypothesis and a revolution in our world view.
Such was the importance of the first specimen unearthed that its discovery was published in Science. The jaw was well preserved, and displayed a number of features that precluded it from being either a monotreme or a marsupial. Its molars were of a very advanced type that can cut, crush and puncture. Such teeth are known as tribosphenic, and only placental mammals and marsupials possess them; indeed they were a vital element of the success of these mammals, allowing them to open up a new ‘food frontier’ inaccessible to their competitors. The fossils had only three such molars, and a rear premolar that superficially resembled them. Such features are hallmarks of the placental mammals: a bit like our ‘kangaroo essentials’, they are features that define the placental lineage. Finally, the specimen lacked the ‘marsupial angle’, a bony strut unique to marsupials. Such a constellation of features indicated that our fossil was an early placental mammal—among the earliest found anywhere.
I spent a year at Harvard soon after this, and there discovered that our claim had been greeted with uproar and disbelief. Ever since the time of Huxley and Darwin, scientists had assumed that the ancestral cradle of the placentals lay in the northern hemisphere. The firmest objection to our identification was not our interpretation of the specimen itself, but to the notion that it had been found in Australia. The refrain I heard all too often was, ‘Everyone knows that Australia had no placental mammals except for some recently arrived bats and rats, so how could this fossil be one? It must be something else that just looked like a placental mammal.’ Among the more dogmatic denials I occasionally detected base motives. Palaeontologists must compete for funds to excavate, and this sometimes makes acknowledging a major discovery by another scientist difficult.
Even in Australia the discovery proved controversial, for Michael Archer has long discounted the possibility that the jaws might have belonged to placentals, instead suggesting (though never supporting) the idea that they are monotremes. Yet Archer is not worried by the thought of placental mammals being present in Australia in the distant past, for he has long championed a solitary, 54-million-year-old tooth from a site his team excavated at Murgon as belonging to just such a creature. My own guess is that this tooth, along with the twenty jaws from Flat Rocks, provides evidence for a placental mammal radiation in Australia which lasted for at least 60 million years before dying out by 54 million years ago. The extinction of placental mammals in Australia may be surprising but it is not implausible. Perhaps Australia’s infertile soils advantaged creatures with low metabolic requirements, like reptiles and marsupials, over the high metabolism, energy-hungry placentals.
In 2002, two independent studies of the relationships of the major types of placental mammals were published in the journals Nature and Science. Both analysed very long pieces of nuclear DNA—16,000 base pairs in all—using a technique called sequencing that, while time-consuming and expensive, is the gold standard for establishing evolutionary lineages. Both studies came to the same conclusion, an outcome that is as close to ‘revealed truth’ as science ever gets. So, what did the studies find?
An astonishing discovery never before even guessed at by anatomists and palaeontologists was that a group of African mammals which seemingly have little in common—including elephants, aardvarks, elephant shrews and the tiny golden moles—had all shared a common ancestor 90 million years ago (and here the ‘molecular clock’ could be calibrated against an extensive fossil record). This placental lineage, dubbed the Afrotheria, is the earliest split in the placental family tree. South America’s sloths, anteaters and armadillos (a group known as the Xenarthra) proved to be the next oldest lineage, their ancestor becoming isolated (probably in Antarctica rather than South America) around 80 million years ago. All remaining placental mammals, the studies showed, shared a common ancestor less than 80 million years before the present, which inhabited the northern continent of Laurasia.
The unavoidable implication of these studies is that the placental mammals had begun to diversify on Gondwana at least 90 million years ago, perhaps as the continent began to break up. In the light of this finding, the discovery of 115-million-year-old placental mammal fossils in Australia is to be expected. So it really does appear that the theory first postulated over a century ago by Thomas Huxley, that the marsupials are southern hemisphere in origin and placentals northern, needs to be turned on its head; for we now know that the marsupials arose on Laurasia, in the northern hemisphere, and it looks ever more likely that the placentals arose on Gondwana. Despite the palaeontological and molecular evidence that supports it, the Gondwanan origin of placentals is still a hot topic in palaeontology and I suspect it will be some time before the full implications of Tom Rich’s discoveries sink in.
A dark age of about 60 million years separates Tom’s contentious fossils from those constituting our next age. It is glimpsed through a single locality near the town of Murgon in southeastern Queensland, where clays laid down in a lake that formed in the crater of an ancient volcano have been dated to at least 54 million years ago (remember that this is a minimum date, and that the fossils may be much older). The mammals, which were all shrew- to rat-sized, are still in the minority in this deposit; though now it is not dinosaurs but crocodiles and giant soft-shelled trionychid turtles (extinct in Australia today, though surviving on other continents) whose bones abound. Most were marsupials, and the Murgon fossils likely represent the earliest marsupial immigrants to reach Australia, for they lived when Australia was connected with Antarctica, and possibly South America as well—as the teeth of several resemble South American species.
Alongside these early marsupials lived the last of Australia’s land-dwelling placental mammals, only two teeth of which have been unearthed. Their very rarity suggests that the newly arrived marsupials were in the process of displacing them. From then until the arrival of Australia’s first rats and mice around 4 million years ago, the continent was free of land-based placentals.
The fossils of one other placental mammal have been recovered from Murgon. The bones of an ancient bat are quite common, indicating that these creatures had arrived in Australia very early in the evolution of the group. Genetic studies suggest that bats originated in the northern hemisphere, so their presence in Australia so soon after their supposed origin is intriguing.
Between Murgon and the next window opening onto Australia’s past lies an
other dark age—this one lasting about 30 million years. This is extremely frustrating, for much occurred during this interlude, including the separation of Australia from Antarctica (45 million years ago), the onset of Australia’s drift northwards (at an initial rate of around twelve centimetres per year), and a series of abrupt and profound changes in the world’s climate. These events must have had a marked impact on Australia’s fauna, and one or more may have led to the evolution of the first kangaroos.
The window following that dark age is provided by the green clays of Lake Pinpa (guesstimated to be 25 million years old, but which may be anywhere between 40 and 20 million years in age), and it opens onto a greatly altered world. The land-based placental mammals have vanished from Australia and the marsupials have diversified into a number of ‘giant’ lineages. The largest of the Lake Pinpa herbivores were the calf-sized (but wombat-like in shape) ilariids, and the dog- to sheep-sized wynyardiids. The ilariids had ‘cuspy’ teeth strikingly similar to those of koalas and ringtail possums. Not many modifications are necessary to produce such teeth from those possessed by the marsupials living in Murgon times. Despite their large size the ilariids mark an early phase of marsupial evolution which was superbly suited to feeding upon leaves, which is perhaps why the koala (an ancient lineage that has not changed much since Lake Pinpa times) still thrives in Australia’s gum trees.
The wynyardiids, on the other hand, share similarities with wombats and diprotodontids, and may have given rise to both of these groups. As we have seen, kangaroos were present by Lake Pinpa times, indicating that they are an ancient lineage, albeit one that, unlike koalas, has changed markedly. Yet Lake Pinpa gives us only a passing glimpse at an era of great importance.
A gap of uncertain millennia separates Pinpa from our next window into the past, provided by a large number of fossil deposits from central and northern Australia, which are thought to be Miocene in age (around 25 to 5 million years ago). Even though some sediments were laid down in the same geographic region as Lake Pinpa, in them is a very different cast of animals, indicating that a dramatic and possibly abrupt climate shift had occurred. In what appear to be the oldest of these deposits (best represented in the Lake Eyre Basin) a few archaic marsupials such as wynyardiids linger, but alongside them we find primitive diprotodontids (the lineage that would give rise to the largest marsupial of all time), marsupial lions and various possums, some of which are still around today. There are also several kinds of bulungamayines and balbarines.