The Timing, Nature, and Aftershock of Pleistocene Extinctions in Australia
Tim F. Flannery, Principal Research Scientist, Australian Museum
Yesterday we had a broad overview which looked at the causes of extinction and the nature of extinctions worldwide, and I tried to outline the Australian case for when these extinctions happened, which we think is beyond 35,000 years ago -- before the last glacial maximum. And they carried off about 55 species of large mammals, leaving nothing in Australia -- at least on the land -- which was bigger than a person. No animals -- everything that was bigger than about, well, 80 kilograms or so, adult body weight, became extinct.
Now, this, I think, had very profound effects for the Australian ecosystem, and it's that ground which I'd like to cover today. I suppose the starting point for thinking about this is to imagine another, a complete megafaunal ecosystem -- and then, in our mind, do to it what had happened to Australia. So, just as a thought experiment, if you all imagine, say, the Serengeti Plain in Africa today, or the Masai Mara, or Krueger National Park -- where there are tremendous amounts of game -- and just get rid of everything off that landscape except a few middle-sized antelope that weigh perhaps 40 to 60 kilogram. And even they'll be rare, I think. So that's the system that we're looking at the day after megafaunal extinction has run its course in Australia.
Now, that is happening in the context of a very, very strange environment. I tried to convince you yesterday, that Australian ecosystems were different from any others.There's extremely low soil fertility, which has given rise to broad areas of plant communities that just have no mammalian herbivores at all; very few insect herbivores. They're a very low productivity ecosystem, where nutrients are recycled through fire, by and large.
Here is a typical Kwongan heathland flora in Western Australia. The soils here are basically nutrient-free. They're 100% silica. Grain size varies a little bit. Water-retention capacity varies slightly. And there are really microtraps for nutrients, I suppose, here and there in just tiny holes in the landscape. But this is a flat landscape. It goes on like that for many hundreds and thousands of kilometers. It's extremely diverse in terms of plant communities, but large mammals are absent. And I think that they've always been absent from these communities, these places. All of these plants defend their leaves with tannins and phenolics, and other chemicals that make them virtually inedible to anything. A lot of these chemicals work on microbes -- so if an animal ingests a leaf from this plant, what the leaf ends up doing is interfering with the microbes which are in the animal's gut, and the microbes' ability to survive. It binds to the coating, the body of the animal. So you can imagine leaves falling to the forest floor here. Even soil microbes have great difficulty breaking them down, and you end up with these areas. Typically, fungi are very important in terms of breaking down leaf litter and whatever else, but even that is so slow that materials sit around for decades waiting for a fire to come along to burn it.
...Well, this is just a kind of a theoretical model of Australian ecosystems. Along this axis you've got decreasing soil fertility and here decreasing rainfall, running down that way. And in the bottom corner, where you have very, very poor soils, and over a somewhat larger range of rainfall types, you get these plant communities which are all dominated by fire. Fire is the main means of nutrient recycling, and fires occur over a periodicity of, you know, five, ten years, through to perhaps 30 years in natural systems. So you think -- the equivalent really is being in an economy where you wait 20 years to spend your dollar, and then for someone else to spend it back for you, if you imagine dollars being equal to nutrients.
They represent about 10% of Australia's ecosystems, or we think they did in the past. There's another section up here which is high fertility, high soil fertility, high rainfall areas -- and these are the rainforests -- the Australians rainforests. Again, they're very, very low in megaherbivores -- there's very few of them -- and the fossil record suggests there were never very many. The reason for that is that, in this corner, microbial decomposition is very, very efficient. Plants lose their leaves, they drop to the forest floor, and within a few weeks the microbes of the forest floor break them down and return the nutrients to the plants.
So if you're a megaherbivore in this area, you're really running under difficulties, because you have to maintain your own body -- it's a homeothermic body -- and you maintain your gut flora, and you're competing against a gut flora which is otherwise widespread and doesn't have to bear the cost of running a homeothermic body. So large homeotherms are rare up in this part of the triangle, as well. And in Australia now, these environments represent about 1% or 2%, of the total area of the continent. In the past, they were more widespread.
In the middle here, though, there was a vast number of plant communities which weren't dominated by fire, and certainly weren't dominated by microbial decomposition. We know about this because we have some magnificent pollen records coming from very long cores in Australia....
Now, we have retrieved a number of offshore cores here, which have pollen in them, along this coast. They're taking pollen from a wide catchment -- so wind is blowing pollen into these sites, into the sea, where the pollen sinks over a very wide area. The sediment pile goes back about 10 million years, and what we see in this sediment pile is just stupendous stability for most of the period. For most of the period, pollen is raining into these sites. The pollen is coming from what's called "dry rainforest communities." These communities are almost extinct in Australia today, but in the past they were clearly very, very important. They include things like araucarias, the plants that are preserved in your petrified forests, in, I guess, the Midwest.
Vine thickets. The plants typically have lots and lots of spines on them, which I'll come back to later. They are very, very fire-sensitive. Their leaves generally have relatively high levels of nitrates and phosphates, and they're not particularly chemically heavily defended. Now, these communities -- we also find evidence for them in sites down here and sites up here in the rainforest. They're widespread. There's also some evidence -- some pockets, even, up in the northern part of Australia, and these are broad-leafed forests. They're a rainforest type, they're called "dry rainforest." They're very different structurally from the eucalypt forests that dominate those areas today.
Now, what these sites show is that, all of a sudden, after 10 million years of stability, at some time between 38,000 years ago and 120,000 years ago -- and we can't really zero in on exactly when -- there is a fundamental and instantaneous change in these plant communities. The dry rainforests disappear virtually overnight, like that, and what takes over is the scleromorph communities, which typify these areas today -- eucalypt forests, acacias, all of the scleromorph plants with their heavily chemically defended leaves, all of which are tied intimately to fire as a means of recycling nutrients.
The other thing that you see is that mangrove pollen becomes important for the first time. Presumably, that's because you're suffering soil erosion, flushing soil down into the river catchments, into the estuaries, and there's the first chance for mangroves to become important in these coastal communities, on the large scale.
Now, how do we explain that change? No climatic model that anyone has conceived of can actually adequately explain that kind of change. But if we go back to what I first said about megafauna -- imagine that situation in Africa, where you've taken everything off the Serengeti, but you've left just a couple of antelope. What you would get is tremendous buildup of vegetation. Now, in the Serengeti it may burn, it may get eaten by termites -- we don't know what the fate of it would be. But here in Australia, what you have is hundreds of communities, plant communities, growing on the very worst soils throughout the continent, which are already fire-promoting. And I suspect it's them that are waiting in the wings to take advantage of this increased fuel load.
Now, one thing that you have to understand about Australia to understand this hypothesis is that in Australia there are more than enough natural ignitions to burn the fuel most of the time. So if you want to change fire in Australia, you must change fuel loads -- you can't simply change ignitions. If you change ignitions -- in other words, you burn the fuel more frequently -- what you get is cooler, more small-scale burns, which are less catastrophic. If you just let the fuel grow, there will always be enough sources of natural ignition to burn it. It's very different to New Zealand, where human sources of ignition are critical in terms of starting fires; here in Australia they're not.
So I think the situation arises where the megafauna disappear, fuel builds up, and then there is this instantaneous transition away from dry rainforest communities, which are palatable megaherbivore herbage. And they're replaced by these scleromorph communities, which are inedible largely to large mammals. They're hostile places. Their leaves are chemically defended. But, worst of all, they're fire-promoting, so they recycle using fire.
Now, you can imagine that this has a profound effect on soil. The soil under rainforests -- rainforests are interesting places, even dry rainforests, in that most of the nutrients are held in a standing biomass of the forest itself. And they're very competent. Very little in the way of nutrients is let out of that living system. These fire-promoting communities are all nutrient-flushing communities. Because they use fire to recycle nutrients, every time there's a fire, nitrates are actually volatilized, lost to the air. Soil sulphurs are lost. The soil becomes ever more impoverished, and that, of course, promotes ever more scleromorphic plants and ever entrenches the fire regime. So I think extinction in Australia drove this cycle, drove a further impoverishment of Australian soils, and the development of the eucalypts and the acacias at the expense of the dry forest communities.
Now, we think this had a particularly profound effect in Northern Australia, somewhat more speculative ... But people who've been studying Lake Eyre -- down here, which is a vast inland sea -- normally dry, except when the northern monsoon is particularly strong. When the northern monsoon is strong, you get sufficient rain falling inwards. There's sea breezes blowing in over here, and you get rain falling in this area. And after several months, that rain works its way down through the channels and fills the lake.
The people who've been studying this model say that it works perfectly up to 50,000 years ago. Every time the lake should have been full, looking at the world climate monsoonal patterns, it was filled. The one exception happens at 11,000 years ago, which was one of the best times for filling the lake. The lake should have been at an all-time high; it remained absolutely dry.
Now, this was quite mysterious, until we started thinking about it in terms of vegetation change in Northern Australia. This area here presently has one of the steepest rainfall gradients from coast to inland anywhere that you'll find. The rain is dropped pretty much as soon as it goes across the coast. There's constant onshore winds. But what happens today is that the rain falls over a scleromorph plant community, with a hot soil with low water retention capacity, and all of that water runs off directly into the sea. Very little of it is actually transpired, because the scleromorph plants just haven't got the ability to do that. They don't have the spongelike root masses under them to even retain the water in the soil. It's bare, cracked soil that just lets the water run off.
But you can imagine 50,000 or more years ago, perhaps, when this was all covered in broad-leafed dry rainforest plant communities. You would get rain falling on these communities and being retained in them, much like a sponge. You'd get lots of transpiration, and these constant onshore winds blow the moisture further inland. What you end up with is an effective 60% increase in precipitation over this area, if you do the figures, and that's more than sufficient to fill the lake at times like 11,000 years ago.
So what we're really saying here is there's a chain of events: megafaunal extinction brought about by overhunting; dramatic change in plant communities throughout Australia, leading to global, or at least continental-wide, climate change. These are very profound influences, I think, that follow on from this initial extinction event.
One of the things that I've been curious about was, following extinction, why didn't the few moderately large species that remained -- like the kangaroos -- kind of expand to fill the available niches? Because there was a vast amount of food out there that just wasn't being eaten. I think the answer was that they were just being hunted very heavily. The evidence for that comes, in part, at least, from what I call "dwarfing" -- Pleistocene dwarfing -- and we have excellent examples in Australia in sediments which go back 35,000 years, where we can trace the average size of kangaroos through these sediments with a great deal of accuracy. And we find that, after extinction, there's no megafauna in the bottom of these sites. The kangaroos are just as big as they were then. But successively, over the next 35,000 years, they lose size, lose mean size, until they reach the modern proportions. And in the case of kangaroos, they weigh about half now what they did beyond 40,000 years ago. So they've lost a lot of body size.
There's one very interesting exception -- an island, a large island, off the coast of South Australia, where humans became extinct 3,800 years ago, and there kangaroos have increased in average body size by about 3.8% -- no, 8% -- over those 3,800 years. Kangaroo are very sexually dimorphic. The biggest males monopolize breeding. They're also the most vulnerable animals to be hit by hunters, because the males will always stand up and challenge the hunter; the females will run away. And you can imagine hunting pressure, which is just selecting over thousands of years for the large-size individuals -- what, in effect, it's selecting for is early-maturing dwarfs. It's like having a fishing net with a certain diameter in it. Only the small fish go through. If they're early-maturing dwarfs, then you drive your stock downwards in body size.
This seems to have been the case of what's happened in Australia, and everything that weighs more than 5 kilograms -- except wombats, curiously -- seem to have suffered this same phenomenon to a lesser or greater extent. The big things have dwarfed more than the small things, but everything has suffered. And there's some very powerful arguments that aboriginals were, in fact, extremely effective at suppressing numbers of animal, even in the 2 to 3 kilogram range, in parts of Australia.
So the big difference, really -- if this natural system had been allowed to go along unmanaged -- Australia wouldn't be like it is today. We wouldn't have relicts of dry rainforest; they would have been eliminated. I suspect we wouldn't have a lot of animals that we did have. The reason why the country resembled this when it was discovered in 1788 is that these people here had been managing fire in Australian landscapes for 40,000 years. Though managing it in the only way which was possible, which was to increase frequency of ignitions, and therefore make cooler, more small-scale burns.
Now, this created, throughout the length and breadth of Australia, a tight vegetational mosaic, where you have mature, or overmature, plant communities right side-by-side with burnt areas -- which are pretty much bare or regenerating. And this kind of spatial heterogeneity in Australian communities is particularly important for medium-sized mammals. Not many of them burrow; not many of them can find adequate refuge. They need to be able to have dense vegetation, where they can hide by day, adjacent to areas where they can feed at night. And this fire-stick farming was used very judiciously, many people think, to create this environmental mosaic; and to create these open parkland situations that were described by the early Australian explorers as being like a gentleman's park. And, of course, gave great hope to people who wanted to come in and graze sheep and cattle in these areas. What they didn't realize is that this park is the result of very infertile soils and management with the fire stick, rather than the surviving megafauna of Europe, in the way of deer, horse, cattle and sheep working on some of the richer soils on the planet, and with a lot of human effort thrown in. So the appearance was just superficial.
So, anyway, Europeans arrive in places like this, see that they're liking these gentlemen's parks, think that the aborigines are all dying of diseases, the kangaroos seem inferior to their own introduced animals. It seems as if God has ordained that this will be their heritage -- the Australian Europeans. So they walk in, take control of the fire stick -- or, most often, just throw it away; don't think about stocking rates, or anything else; and these communities change dramatically. These early drawings are so alien to most modern Australians that, for many years, it was thought that they resulted from the fact that the English artists hadn't actually quite been able to transfer to Australia and draw Australian environments properly. I now think that's not true. I think that a lot of Australia really did look like this in the past.
So you can imagine 40,000 of fire-stick farming all of a sudden stopped. You get the fuel buildup again that you would have had 40,000 years ago, and you have very dramatic wildfire events -- particularly in Central Australia. We've had fires in the 1970s which have burnt through three states. They've burned a million hectares at a time and burned for months. And if you're a middle-sized mammal that weighs a kilogram or so, and you wake up one morning and all you see, in all directions, is ash, you're in serious trouble. And we think that one of the most important reasons that 23 species of middle-sized Australians mammals became extinct in the center is simply the removal of this spatial heterogeneity. Initially, the heterogeneity would have been maintained by the large megaherbivore assemblages of the Pleistocene, then by fire-stick farming -- but today it's all gone, as these pyrophilic plant communities have taken over in Australia and changed things very dramatically.
There's an aboriginal man doing fire-stick farming. But for most of his life, he didn't do fire-stick farming. He grew up as a traditional aboriginal person, west of Alice Springs; spent most of his life as a stockman, not managing the land. And only now, after a 30- or 40-year break, has come back to start managing. And when you talk to people like this about their wildlife, if they haven't been back to their own land yet, if you meet them on a station and say: Look, what about all of these animals? -- and bring study skins along to show them -- they'll point to animals that have been extinct for 60 years, or we think, anyway -- pig-footed bandicoots and battongs and bilbies -- and saw plenty of them, plenty of them. We used to eat them all the time, you know? You go back to their country, and they're still full of confidence they're out there.
And when they get there they say: No one's been looking after the country. It's all overgrown. And the first thing they do is start burning. They just start throwing matches out of the back of trucks. But by then it's too late. The country literally is empty, and 30 or 40 years have gone by; enough time has passed for these huge fires to go through, and you really find that the middle-sized, all 23 species of middle-sized mammals -- everything between a rat and a red kangaroo -- have gone from most of these areas. They do survive in the coastal parts of the continent, because there there is more topographic variety, so fires never burn everything. There are always fire-shadow areas, where these things can survive. But in the center, which is basically flat, these extinctions have happened very dramatically.
People think that the introduction of the fox and the rabbit have also been important. I'd say they may be, but to a lesser extent. I mean, one of the problems, if you're a medium-sized mammal that wakes up to a field of ashes in the morning is to escape the predators -- or the evening, rather. And if you have the efficient things, like foxes there, of course, they're likely to pick you off more rapidly. But what we have found is that these extinctions, on the full scale, have happened in areas where, in fact, foxes have never penetrated into Northern Australia. So even in the absence of foxes, and just in the presence of the native carnivores, these fires are quite capable of driving to extinction this whole fauna, this whole middle-sized mammal fauna.
Some of the animals that were lost include very cute things like this -- one of the nailtail wallabies. And these were the mainstay, I should say, of aboriginal economy in Central Australia. The mammals were the things that were eaten, and they were very important to them.
These areas, by the way, where these animals survived, had cats in them. By the 1880s, cats had spread very quickly; foxes by perhaps 1920 or 1930; rabbits earlier. And the spread of these introduced species was always put down to this kind of European view that things from Europe must somehow be superior. They're placental mammals, after all -- they have to be superior. But, just to take you back to the Serengeti again, imagine taking a rabbit and putting it onto the Serengeti today. What do you think would happen? It would be eaten by something, or outcompeted, or trodden on by something, I'm sure, before it could establish. Because you're getting a coherent, very packed, tightly packed mammal community in that area.
But take everything out of the Serengeti but a few antelope, and then put the rabbit back in, and then you've got a pest species. It's not the inherent superiority of the rabbit that's important -- what's important is the Pleistocene history of the continent and this style of extinction sort of occurred on that continent. That's why Australia is so apparently plague-prone, simply because our extinctions carried off 95% of what is considered large mammals -- those species over 40 kilograms in weight. And that percentage, I think, is the highest for the planet. Even small things like the mice, hopping mice, which are placental mammals -- the only terrestrial group present in Australia -- were pretty profoundly affected by these extinctions.
Now, these fires continue to plague Australia. They burn into our capital cities. Every fire that's occurred in Sydney over the last century has been more serious than the one preceding it, because the buildup of fuel has been that much greater. We haven't had a way yet to deal with any of these fire-related problems, and, in fact, we continue to lose, along this margin here -- which is an extraordinary place. Here you'll get rainforest with microbial decomposition, and I can literally, in one step, step over, in most of Australia, into the fire-dominated scleromorph communities. There's usually strong edaphic control; there'll be a change in rock type, or whatever, that produces that boundary. But it is incredibly sharp -- it's razor-sharp across the continent. And these two are obviously incompatible systems. But during bad years, particularly in the northern territory, where there's uncontrolled fire, you'll get this community edging even further away into this one here.
I'd like to just finish up, I guess, by saying that the world we live in is not a competent, complete, fully functioning ecosystem world. In places like Australia and North America, the place is full of ecological ghosts. The time since these extinctions happened is so short that no new species have evolved to fill their niches. There are lots of empty niches out there, lots of loose connections, lots of opportunity for pest species. When we think about managing our national parks and whatever, perhaps we need to think more innovatively. These aren't wilderness areas that can just go along by themselves -- they're crippled ecosystems which need active management. And when we look at the distribution of animals, I don't think we can be too picky about where we take animals from, if we want to do translocations. We'll find that these events have so disrupted distributions of animals -- particularly in Australia -- that if we want to introduce things to western New South Wales, say, we may have to go 3,000 or 4,000 kilometers away for the nearest surviving population.
The same with plant communities. I mean, we have had this revolution in plant communities, where the dry rainforest communities have been all but extirpated on the continent, and the scleromorph communities have come out of the areas of poorer soils and claimed most of the continent as their own. In terms of managing that whole system, you can imagine the difficulties involved. And I should say that this change was so profound in Australia that we actually lost several plant genera during the Pleistocene. Species like dacricirpus and lagarostrobus, which are -- these are southern pines, very fire-sensitive -- were lost at the time of this event. So we're looking at very, very profound reorganizations following faunal and megafaunal extinction. And I think that anyone who's interested in land management, or the problems of introduced species -- or whatever else -- really needs to understand the nature of this event.
Now, I'm not saying that you will see things on this scale in North America. I think Australia was preadapted to have a particularly severe aftershock from extinction, and that was because of the large numbers of pyrophilic plant communities which had been evolving there for many millions of years. They were the jack that got let out of the box when extinction occurred. And maybe that didn't happen here in North America, but, still, I'd be very interested to look at any kind of information relating to megafaunal extinction aftershock that may have come out of the Northern Hemisphere.
We have time for questions. Please use the microphone in
the center aisle here. Anybody?
What do -- you mentioned that so little did eat these arid-adaptive,
these plants that occur on poor soil. What do red kangaroos eat?
I mean, they're not really megafauna, but they're certainly pretty
Yes. Red kangaroos are grazers. They eat short green grass,
which is an unusual thing to eat when you're out in Central Australia,
but they do manage to find it. I mean, as a general rule -- and
there's always exceptions to these things -- but the chemically
defended plants are generally fire-promoting, and they generally
have very, very low levels of browsing pressure on them. So all
...because red kangaroos are still quite common in Central Australia,
Oh, they are -- and increasingly so, since aboriginal pressure
was taken off them. And that's the other thing I should have
mentioned -- they've bounded back tremendously. All of the large
kangaroos were exceedingly rare on first contact, presumably because
of aboriginal predation. They've now become very, very common
-- and we think that's just an artifact of release from hunting
I just had a quick question about translocations. You mentioned
that you guys are beginning to do some stuff for restoration in
some of these areas. What's going on?
We are trying to -- we're experimenting at the moment with
translocations in a number of different ways. There's a guy called
John Walmsley who is creating really open-range zoos, where he's
fencing large areas, excluding foxes and rabbits, and then putting
lots of middle-sized mammals in -- but in the absence of carnivores.
Now, I don't know whether he's actually had to deal with fire
events yet. I doubt it. Some of the early work we did in the
north, trying to preserve the last remnants of some of these highly
endangered species, failed when wildfires burnt through the enclosures
they were kept in, and caused 100% mortality. So we still haven't
got it right yet. We're in the early stages of experimentation.
Perhaps the grandest scale experiment is happening in Western
Australia, where there's a large peninsula called the Peron Peninsula,
which is 150 square kilometers that has been fenced
off, and there'll be reintroductions taken there. We shall see what
What's being done about camels? What's your recommendation
Yes -- camels in Australia. Australia has
the only wild-living camel populations left on the planet, and
they're widespread in the arid regions. But they remain relatively
poorly studied. Their full ecological impact isn't known as yet.
I certainly think it's worthwhile running the experiment of keeping
them out there. I mean, they are, in some sense, an ecological
vicar for these extinct species, one hopes.
Our big problem in Australia is that most of the megafauna were
browsers. I think they were browsers of the dry rainforest communities,
which are now vanished. There's very few grazers among them --
and, surprisingly, the grazers like red kangaroos and wombats
survived -- they survived through the transition. And it's pretty
hard to find mammalian browsers of the right sort to replace these
things. Rhinoceroses is one of the things that's occurred to
me, but you can imagine people are a bit wary about that. I've
also suggested that Komodo dragons mightn't be a bad thing to
bring back in as vicars for these great big extinct goannas,
but we're yet to be convinced about that, as well.