Death Down Under: A Deep Look At Australia’s Megafaunal Mystery

There are few places on the planet as distinctive as Australia. With its isolated geographical location and iconic landscapes, it captures the imagination of people the world over. Picturesque red deserts, glistening beaches, and Eucalyptus woodlands may provide its most characteristic scenery, but arcane rainforests, snow-covered mountains, and much more can be found on the continent as well.

As fascinating as today’s Australian fauna are, something is missing. Few of them classify as megafauna, or animals larger than 45 kg. Up to 23 out of 24 genera of Australian land animals weighing more than 45kg disappeared towards the end of the Late Quaternary, mostly by 40 thousand years ago¹. Among the genera that were lost included the largest known marsupial, a giant flightless bird, a massive relative of the Komodo dragon, a marsupial version of the lion, and a terrestrial crocodile. It is hard to imagine the continent once hosted such beasts, given that the largest land animal there today is the red kangaroo.

What happened? The Australian extinctions serve as both an extension of and a focal point within the broader debate surrounding the disappearance of most of the world’s megafauna during the Late Pleistocene across multiple continents, with humans and climate change of course being the main suspects. Those who blame humans argue that overhunting and/or landscape modification were the root cause of extinction in Australia, while those who favor climatic explanations argue that it was instead increasing aridity that pushed the fauna of this already dry continent over the edge.

After having done extensive research on the topic, I will argue comprehensively for the idea that humans played the primary role in the disappearance of megafauna, focusing mainly on the following points:

1. Aboriginal arrival closely coinciding with a clear extinction wave
2. Serious flaws in climate arguments, including a lack of evidence for major climatic instability during the extinction window 50-40 thousand years ago(kya)
3. High susceptibility of Australian megafauna to human predation
4. Potential contribution of anthropogenic modification of Australian landscapes to extinctions

I recognize that it may seem audacious to claim climate played a minor role at best given how controversial the topic is. This is why I’ve put great effort into finding as many sources as possible and using logical reasoning to strengthen my case. Uncertainties in the research will be made clear, as will my own personal speculations.

Before going further, I would like to note that the geographic terminology for this part of the world can get confusing. There are several names for broader regions that include Australia. For the purpose of reducing confusion, I will use Australia synonymously with mainland Australia while Australasia(though more of a biogeographic term) will be used to refer to the modern region encompassing Australia, New Guinea, and Tasmania. Sahul will refer to the prehistoric continent that united all three.

Additionally, this post and its reference list are both quite long so I decided to put the references behind a spoiler tag at the end for the reader’s convenience. Clicking on the button will reveal them. After that, you can click on the numbers to jump to the corresponding references.

With that said, let’s get to the bottom of this mystery in the land down under.

Background

To assess what happened to the megafauna, it may be best to set context by describing Australia on the eve of the extinctions. It was the first half of the period known as Marine Isotope Stage 3, an important phase of the last glaciation. It was cooler than today but not as cold as the Last Glacial Maximum(LGM) tens of thousands of years later. Like today, Australia hosted desert, savannah, dry scrub, swamp, and forest. Open landscapes were even more dominant than at present.

Lower sea levels during glacials exposed continental shelves and created land bridges connecting mainland Australia to several islands in the Australasian realm, notably New Guinea and Tasmania. A land bridge between Australia and New Guinea was present by the time humans arrived, while the land bridge between Australia and Tasmania formed slightly later. The combined landmass created a paleocontinent known as Sahul.

Extinct megafaunal species present at the time included Diprotodon optatum, a nearly 3-ton animal resembling a giant koala and perhaps the largest known marsupial; Genyornis newtoni, the last known member of a group of large, flightless Australian birds known as thunderbirds; and Thylacoleo carnifex, also called the marsupial lion. An unnamed species of Meiolania, a genus of large turtles widespread across Australasia, was present as well. It may have been the biggest of its genus at 2-3 meters long.

Quinkana fortirostrum was a terrestrial crocodile of the Mekosuchine clade. Varanus priscus, also known as Megalania, was a massive relative of the Komodo that may have been 5 meters long. Additionally, there were multiple species of large kangaroos belonging to the extinct genera Procoptodon (short faced kangaroos) and Protemnodon (giant kangaroos), while the extant Macropus genus (modern grey kangaroos) contained more members than the current two species.

A biogeographic barrier known as the Wallace line separates Australasia from southeast Asia. As a result, Australasia has been isolated for tens of millions of years, which is why the overall composition of its flora and fauna diverged from other realms to such a significant degree. It is the only place where marsupials are the dominant form of mammals, and no placental land mammals besides bats and rodents were present before humans and the animals we introduced like dingoes. This is important context when we get to discussing the fate of those fauna.

Now that we understand what kind of continent humans were about to discover, we can move on to pinpointing when humans actually got there and the implications of it.

The Arrival

The trouble with tying human arrival to the megafaunal extinctions is that scientists aren’t entirely confident as to when humans first arrived on any given continent. The information is important because conventional logic on this topic would suggest that extinction should immediately follow human entry. Time between arrival and extinction is not the end-all-be-all as there are many factors which can delay extinction at the hands of people, but it is useful to get an idea of when people arrived nonetheless. So how do we figure out when people actually got to Australia?

Since archeologists constantly debate the dates of very old archeological sites, genetics is absolutely pivotal in solving this puzzle. Genetic research shows that Neanderthal ancestry introgressed into the common ancestors of all non-Africans in the Middle East around 54-49 kya². Shortly after, the East Eurasian(aka Eastern non-African) clade split from West Eurasians. Australasian populations fit within the East Eurasian clade, along with East and Southeast Asians.

Next came the split of the ancestors of Australasians from the rest of the East Eurasian clade. After ~49 kya, the ancestors of Australasians absorbed a large amount of Denisovan ancestry, likely in southeast Asia. From there they went on to Sahul; a study on Aboriginal mitogenomes indicates a rapid colonization of Sahul from the northern part of the continent soon after 50 kya³. This aligns with the oldest reliable archeological sites in Australia which date to around 48-47 kya^{4 5}.

These key events starting from Neanderthal admixture into the ancestors of non-Africans and ending with the settlement of Sahul happened in short succession. This was a rapid eastward migration across Eurasia and into Sahul during the Initial Upper Paleolithic, a period of technological and cultural leaps among humans. Critically, all of this suggests Australia was settled slightly later than 50 kya, not earlier.

It is commonly claimed that people arrived in Australia around 65 kya, which if true could imply an extended coexistence between humans and megafauna. However, this figure is based solely on the controversial Madjedbebe site in northern Australia, whose contents are probably much younger⁵. There are no reliable dates for Homo sapiens presence that old in most of Eurasia, and this date doesn’t align with what we know about Aboriginal genetics as outlined above.

Hypothetically speaking, if people were indeed present in Australia 65 kya+, the human group(s) present there at the time most certainly were not the ancestors of modern Aboriginals. More importantly, the lack of associated archeological sites left behind by these alleged pre-Aboriginal arrivals would suggest they were probably too small in number to drastically impact their environment anyway. The implications are that we can largely ignore Madjedbebe and reasonably set the date for Aboriginal settlement to about 48-47 kya.

Pinpointing the Time of Extinctions

Now that we have a timeframe for when Aboriginals arrived, we need to determine when the extinctions(or at least most of them) took place. This is important to do for two reasons:

1. Some argue that the extinctions were staggered and that there wasn’t one brief period when Australian megafauna died off en masse⁶. Rather, increasingly intense climate cycles over the past few hundred thousand years wiped out megafauna in many waves such that there were hardly any present by the time humans arrived.
2. The same people also claim that those surviving megafauna then coexisted with humans for many thousands of years before disappearing, suggesting that humans were not responsible because they should have immediately caused extinction upon arrival as opposed to delaying it⁷.

Regarding the first notion, which I would call the “attrition hypothesis”, it is not clear if Middle Pleistocene species were mostly extinct by the time humans arrived. Evidence from places like Naracoorte cave indicating that many species indeed made it well into the Late Pleistocene⁸. The Signor-Lipps effect combined with an especially scanty fossil record for Australia is likely causing the extinctions to appear far more staggered than they really were. Towards the end of the next section, I will explain why it is unlikely that such high, staggered losses took place.

As for the second, there are major doubts about the validity of late-surviving megafauna in Australia, with outlying dates usually resulting from erroneous dating methods⁹. Further, a few late surviving outliers do not diminish the big picture of an extinction pulse taking place at a given time. In fact, due to the great size of Australia, it would be unsurprising if some stragglers did hang on for much longer than others in some locations even if animal communities were wiped out quickly overall.

Going back to the original question, can we determine when the megafauna, or at least most of them, vanished? Coprophilous fungi, or dung fungi, can give us major clues. Dung fungi are associated with the digestive process of herbivorous animals and can be used by scientists to track the abundance of megafauna at locations through time. Higher levels of dung fungi usually indicate higher animal density; a crash in levels with no recovery can indicate extinction. Though concerns exist on the reliability and utility of dung fungi as a proxy, taking certain precautions can resolve them¹⁰.

A number of studies using dung fungal analysis have been published for Australia. Johnson et al. 2015 reported a decline in dung fungi at Lynch’s crater in northeast Australia around 40 kya¹¹. A 2017 study analyzed an ocean core off Cape Pasley in southwestern Australia while a 2023 study analyzed another near Kangaroo island( recording data from the Murray-Darling basin in interior southeastern Australia)^{12 13}. Both ocean cores revealed that dung fungal spores plummeted to low levels around 43 kya. In Caledonia Fen, a high-elevation site near the coast of southeastern Australia, dung fungi plummeted somewhere between 52-46 kya¹⁴.

The authors of the Caledonia Fen paper caution that the time scale for their site was not fully resolved(hence the range). Also, it is possible that the timing of fire and vegetation changes at Lynch’s crater took place slightly earlier¹⁵ than suggested by Johnson et al., which I imagine may apply to the timing of spore collapse there too. Regardless, true dates will not be very far off and these studies took care to avoid biased results in dung fungal analysis.

In each case, dung fungi levels fluctuated to some degree but remained high overall, and did not fully recover after crashing for the rest of the Pleistocene. Crucially, extinction estimates from fungal analysis are not far from those derived from reliable fossil analysis suggesting that extinctions were mostly complete by ~41 kya¹⁶. There has been difficulty in precisely dating some fossils and sediments due to their considerable ages, but an extinction pulse between 50 and 40 kya is near-certain. This shows that megafaunal collapse in Australia closely followed human entry and was abrupt rather than staggered, even if final extinctions of some species possibly occurred later.

Glaring Problems in Climatic Arguments

Could climate change be to blame? Aridification, or increasing dryness over time, is the form of climate change cited to explain Australia’s extinctions. It serves as a convenient climatic explanation for two reasons¹⁷. First, Australia is an already dry continent and it is plausible that an increase in aridity -gradual or sudden- could push its animals over the edge, mainly by diminishing water supplies. Second, the extinctions occurred during a glacial period, and a common but outdated belief is that glacials were markedly drier than interglacials almost everywhere.

A commonly cited study by Hocknull et al. argues that “sustained environmental deterioration” due to aridification from ~48 kya to ~30 kya was the cause of megafaunal collapse¹⁸. The authors cast doubt on the reliability of dung fungal analysis, and date sediments associated with megafauna to claim that extinctions occurred after 40 kya. Yet, in the previous section we pointed out that dung fungal analysis can be reliable and indicated megafaunal collapse by 40 kya, not after. But what can we say about climate change in Australia during the actual extinction window from 50-40 kya, as well as more broadly?

Recognizing scale is crucial: demonstrating evidence of some climate change does not prove it was extreme enough to cause widespread megafaunal loss. Given a timeframe of many thousands of years and a landmass as large as Australia, one will inevitably find some evidence of localized climate change. This occurs even in our stable Holocene. Those who advance the aridity hypothesis, therefore, cannot simply cherry-pick examples of regional drying trends but must prove that aridification was both severe and ubiquitous enough to cause sudden continent-wide extinctions of several genera, many of which were dry-adapted¹⁹.

No evidence exists for such a drastic climate shift; paleoclimate simulations do not indicate anomalous aridity in Australia from 50-40 kya²⁰. A hydroclimate analysis(involving water bodies) suggests slightly wetter conditions overall than present²¹. While it is true that the monsoonal region of north Australia experienced a drying trend ~48 kya, it remained wetter than it is at present for thousands of years afterward. Currently, the region is wet enough to support crocodilians. Overall continental drying only set in after 40 kya(following dung fungi decline), and there is no proof that it was unusually severe.

Ice core records from Antarctica reliably track climate change in the southern hemisphere including Australia. The graph below shows normal temperature shifts in Antarctica during MIS 3(~57-25 kya) overall, especially compared to Greenland. Climate patterns on our planet are interconnected-extreme climate change on one continent will happen in the context of major global or hemispheric changes, and will tend to be associated with major shifts in temperature. It is therefore unlikely that one continent would experience sudden hyper-aridification on its own. The study by Hocknull et al.¹⁸ fails to explain how the purported severe climatic downturn in Australia (48–30 kya) was unprecedented in the Late Pleistocene or aligns with global or hemispheric trends.

Recent research combining speleothem data with insights into moisture balance and CO2-plant dynamics shows that glacial periods in Australia were not invariably drier than interglacials but alternated between wet and dry phases, while interglacials were consistently dry²². Lower CO2 rather than higher aridity explains reduced woody cover in Australia during glacials. The implication is important: while there being fewer woody plants may have affected some browsing animals, cool conditions during the extinction window did not automatically mean reduced water sources, as confirmed by hydrological studies²⁰.

Yet, perhaps the most overlooked aspect of the climate debate is what we can glean from human history. We must keep in mind that humans were not passive bystanders to acute environmental change. They were affected just as animals were. This is especially the case when it comes to increased aridity, as humans are obviously highly reliant on water. If there was a surge in dryness coinciding with or post-dating human arrival that decimated most Australian megafauna, we would expect to see evidence that resident humans were negatively impacted too.

We do not see this. In fact, genetic testing indicates that the effective population of Aboriginal Australians remained remarkably stable and actually slowly increased after arrival, with no bottleneck anywhere in the past 50,000 years^{3 23}. We also see strong regional differentiation which indicates that Aboriginal groups largely descend from the same ancestors who migrated to their respective regions very long ago, in stark contrast to hunter-gatherers in Europe who faced repeated extinctions and population turnovers as a result of climatic turbulence²⁴.

All animals have their limits and even those Australian megafauna whose ancestors had evolved to endure aridity over millions of years were not immune from extreme hyper-aridity. However, attributing their demise to this cause raises the question: why did the same aridity fail to eliminate or even affect the newly arrived humans? The more plausible explanation is that early Australians simply faced low environmental stress, particularly compared to their Pleistocene European counterparts, which fits with Australia being a place of relative climatic stability for the past few million years²⁵(see graph below).

Moreover, while every continent, including Australia, experiences natural environmental change and extinction over time, this relative stability casts serious doubt on the hypothesis set forth by Wroe and others wherein heavy, staggered losses of Australian megafauna took place from the Middle Pleistocene onwards^{6 18}. Such high attrition of megafauna across the same period is not known from Europe and North America where climatic turbulence was far greater. It would imply the unlikely scenario that Australia followed a uniquely intense pattern of decline.

There are other flaws in the climate change narrative worth noting. For example, a vast continental shelf was exposed in northern Australia ~50 kya, which continued to expand as sea levels fell further. I always wondered why, if there was intense environmental change after 50 kya, it was not possible for these animals to shift their ranges away from the arid interior and onto the less-dry continental shelf. I later read a paper by Brook et al. making the exact same point²⁶. It is not logical that animals would make themselves sitting ducks to climate change by not exploiting the extra real estate available at the time.

Lastly, if aridity was to blame for the extinctions, why were Tasmania and New Guinea not spared the loss of megafauna? Tasmania and especially New Guinea are much wetter than Australia on average. Rain-drenched New Guinea continuously hosted dense rainforests and many rivers²⁷. Yet, its megafauna suffered the same fate as their mainland Australian counterparts following human arrival²⁶, as did Tasmania’s megafauna immediately following the formation of a land bridge that allowed humans to cross²⁸. To my knowledge, this inconvenient fact has never been addressed.

The various lines of evidence against intense, widespread drying on mainland Australia, the presence of continental shelves which could mitigate the worst effects of environmental change, and similar extinctions on nearby wet regions climate played at best a minor role in the disappearance of Australian megafauna, likely restricted to a handful of locations.

Sahulian Susceptibility

So now that we’ve exposed the many flaws in the climate change hypothesis, that leaves us with only humans as the culprit as they were the only other agent in the picture at the time. But the hunter gatherers of Sahul had a low population density and initially primitive(likely wooden) technology²⁹, so how did they wipe out all those megafauna? While much of this can be explained by Sahul’s megafauna being totally naive to hominins, I would argue other circumstances were also decisive.

Recall the high degree of isolation of Australasia and dominance of marsupials there. Isolation reduces selection pressure and may hinder competitiveness. Meanwhile, marsupial anatomy itself may preclude marsupial predators from being effective³⁰, and also results in a lack of cursoriality(running ability) among marsupials generally³¹. This could imply that large Sahulian herbivores lacked strong anti-predatory defenses, such as being able to escape quickly, compared to their Afro-Eurasian counterparts. Studies on Procoptodon goliah locomotion seem to validate this³².

This would not bode well for the megafauna who were about to encounter a super-predator in the form of Homo sapiens. The people who reached Sahul initially lacked the advanced stone tools of their counterparts in Europe and northern Asia², but this was offset(perhaps considerably so) by the fact that the prey they encountered were unusually vulnerable. Marsupial and reptilian predators, accustomed to this less challenging milieu, would have been easily outcompeted.

Models suggest a low-density Aboriginal population could quickly drive large mammals like Diprotodon to extinction²⁹. Nearby New Zealand provides a concrete example of tiny human populations causing the resident megafauna-the Moa-to perish rapidly³³. Lastly, humans would not have been randomly distributed across a landmass as dry as Australia but rather concentrated in areas where water was available, where they were most likely to cross paths with megafauna³⁴.

All of this combines to explain why the early settlers were capable of eliminating megafauna through hunting, but this piece would not be complete without mentioning the potentially significant role of Aboriginal fire-use in the extinctions. I consider this one of the most fascinating parts of the extinction debate.

Scorched Earth?

Around the time of human settlement, vegetation change occurred in some places-sclerophyllous(tough leaved) trees like Eucalyptus replaced non-sclerophyllous ones, while the proportion of herbaceous taxa increased. While some view it as proof of climate change, many scientists hypothesize that the shift in vegetation was instead brought on by an increase in fire due to humans. Anthropogenic firing may have replaced fire-sensitive plants with fire-tolerant ones and even partly contributed to the extinctions³⁵.

Indigenous Australians practiced what is known as fire-stick farming for thousands of years. In The Biggest Estate on Earth: How Aborigines Made Australia, Bill Gammage describes how the Australian landscape looked neat and park-like to early Europeans due to firestick farming, revealing how effectively a low-density human population could wield fire. It was used for clearing vegetation through burning to facilitate hunting by promoting open environments, as well as reducing fire intensity in natural fires by removing excess vegetation.

The extent of anthropogenic fire during the initial human settlement of Sahul, their impacts on its landscapes, and their role in the extinctions remains hotly contested. Some claim that peaks in fire activity after human arrival were natural³⁶ or that human fire had modest impacts on biota³⁷. While bushfires have always taken place in Sahul, not all fire may have been natural. Burning and canopy opening occurred even in the wet New Guinea region at times of cool temperatures and relatively high precipitation, when bushfires would have been quite rare, possibly indicating human agency³⁸.

A key site for studying fire and vegetation change is Lynch’s crater in northeast Australia. Initial studies suggested anthropogenic fire preceded the shift from mixed rainforest/sclerophyllous forest to homogenous sclerophyll forest¹⁵. However, later research argued that vegetation change and fire followed megafaunal elimination by humans, as the absence of herbivores allowed excessive plant growth, intensifying natural fires by providing extra fuel^{11 39}.

At Murray-Darling, a spike in charcoal levels slightly after 50 kya was likely due to humans. Animals went extinct soon after but vegetation did not change appreciably aside from a temporary decline in plants with large seeds and fruits thought to be dispersed by large herbivores¹³. The authors conclude that continued fires after extinction were due not to increased vegetative fuel but ongoing burning by people. The results and conclusion are quite different from the second study on Lynch’s crater, but it may be expected as the Murray-Darling basin contained dry, fire tolerant plants as opposed to the wetter, fire-sensitive rainforest taxa of the former.

This shows it is necessary to examine ecosystem-specific reactions to fire. The lead author of the Murray-Darling paper produced another great study which did just that by comparing two sites in Tasmania: one located in what was then dry shrubland and another in what was then rainforest⁴⁰. It found that the shrubland site saw no vegetative turnover at the outset of human burning around 41 kya, but the rainforest was transformed into a scrubby landscape. The role of herbivore collapse on vegetation was not explored, but human firing probably had the greater impact on flora as burning clearly preceded turnover of rainforest taxa.

If rainforests(and possibly other wet habitats) were destroyed by humans through burning, this could provide a useful mechanism for extinction. One may speculate that extinction via habitat destruction rather than direct hunting alone could explain why there are no “kill sites” showing evidence of humans hunting megafauna in Australia⁶. However, the sheer time gap to the present could be a sufficient explanation, as this makes the chances of such evidence being preserved astronomically low⁴¹.

Further, it is worth noting that most Australian megafauna lived in fire-tolerant habitats but extinctions occurred there too, such as in the Murray-Darling basin and Caledonia Fen^{13 14}. Moreover, New Guinea’s rainforest persisted through the end of the Pleistocene but its megafauna did not. This could imply hunting was a bigger factor than burning in extinction. More research from more areas, along with more precise chronology will help resolve the exact role of anthropogenic fire in Sahul.

Conclusion

It is not hard to demonstrate why climatic explanations for Late Pleistocene extinctions in Australia fall short, especially focusing on the lack of signs of aridification during the extinctions. It can even be argued that there is irony in searching for evidence of increased dryness as the cause, since any aridification severe enough to wipe out so many types of dry-adapted megafauna across the entirety of Australia would already be known as it would have left clear and unmistakable signs.

Humans, meanwhile, were more than capable of decimating the vulnerable dwellers of this isolated land despite primitive technology and low population density. They swept across Australia from the north in rapid fashion shortly after 50 kya, with extinction following soon after. The extent to which anthropogenic fire affected flora and contributed to extinctions is uncertain as of yet.

The case of Australia is central to the topic of Late Pleistocene extinctions, and can provide useful insights in the broader debate. Natural environmental change undoubtedly contributed to extinction in large areas of the planet, but Australia and Sahul in general may present the clearest example of prehistoric humans inflicting continent-wide losses of megafauna without much aid from climate change. Recognizing human potency and accounting for continent-specific circumstances will be key to properly assessing the Late Pleistocene extinctions globally.