Pleistocene megafaunal extinctions – were humans to blame?

Australia and the Americas had an extremely diverse fauna of large beasts (giant wombats and kangeroos in Australia; elephants, bears, big cats, camelids, ground sloths etc in the Americas) until the last glaciation and the warming period that led into the Holocene interglacial. The majority of these megafauna species vanished suddenly during that recent period. To a lesser extent something similar happened in Eurasia, but nothing significant in Africa. Because the last glacial cycle also saw migration of efficient human hunter-gatherers to every other continent except Antarctica, many ecologists, palaeontologists and anthropologists saw a direct link between human predation and the mass extinction (see Earth-Pages of April 2012. Earlier humans had indeed spread far and wide in Eurasia before, and the crude hypothesis that the last arrivals in Australasia and the Americas devoured all the meatiest prey in three continents had some traction as a result: predation in Eurasia and Africa by earlier hominids would have made surviving prey congenitally wary of bipeds with spears. In Australia and the Americas the megafauna species would have been naive and confident in their sheer bulk, numbers, speed and, in some cases, ferocity. Other possibilities emerged, such as the introduction of viruses to which faunas had no immunity or as a result of climate change, but none of the three possibilities has gained incontrovertible proof. But the most popular, human connection has had severe knocks in the last couple of years. A fourth, that the extinctions stemmed from a comet impact proved to have little traction.

English: s were driven to extinction by and hu...
Megafauna in a late-Pleistocene landscape including woolly mammoths and rhinoceroses, horses, and cave lions with a carcass. (credit: Wikipedia)

Since the amazing success of analysing the bulk DNA debris in sea water – environmental DNA or eDNA – to look at the local diversity of marine animals, the analytical and computing techniques that made it possible have been turned to ancient terrestrial materials: soils, permafrost and glacial ice. One of the first attempts revealed mammoth and pre-Columbian horse DNA surviving in Alaskan permafrost, thanks to the herds’ copious urination and dung spreading. Several articles in the 24 July 2015 issue of Science review ancient DNA advances, including eDNA from soils that chart changes in both fauna and flora over the last glacial cycle (Pennisi, E. 2015. Lost worlds found. Science, v. 349, p. 367-369). Combined with a variety of means of dating the material that yield the ancient eDNA, an interesting picture is emerging. The soil and permafrost samples potentially express ancient ecosystems in far more detail than would fossil animals or pollens, many of which are too similar to look at the species level and in any case are dominated by the most abundant plants rather than showing those critical in the food chain.

Nunavut tundra
Plants of the Arctic tundra in Nunavut, Canada (Photo credit: Wikipedia)

The first major success in palaeoecology of this kind came with a 50-author paper using eDNA ‘bar-coding’ of permafrost from 242 sites in Siberia and Alaska IWillerslev, E. and 49 others 2014. Fifty thousand years of Arctic vegetation and megafaunal diet. Nature, v. 506, p. 47-51. doi:10.1038/nature12921). Dividing the samples into 3 time spans – 50-25, 25-15 (last glacial maximum) and younger than 15 ka – the team found these major stages in the last glacial cycle mapped an ecological change from a dry tundra dominated by abundant herbaceous plants (forbs including abundant anemones and forget-me-not), to a markedly depleted Arctic steppe ecosystem then moist tundra with woody plants and grasses dominating. They also analysed the eDNA of dung and gut contents from ice-age megafauna, such as mammoths, bison and woolly rhinos, where these were found, which showed that forbs were the mainstay of their diet. Using bones of large mammals 6 member of the team also established the timing of extinctions in the last 56 ka (Cooper, A. et al. 2015. Abrupt warming events drove Late Pleistocene Holarctic megafaunal turnover. Science, DOI: 10.1126/science.aac4315), showing 31 regional extinction pulses linked to the rapid ups and downs of climate during Dansgaard-Oeschger cycles in the run-up to the last glacial maximum. By the end of the last glacial maximum, the megafauna were highly stressed by purely climatic and ecological factors. Human predation probably finished them off.

Large-animal extinction in Australia linked to human hunters

Diprotodon optatum, Pleistocene of Australia. ...
Artist's impression of a giant Australian wombat (Diprotodon) (credit: Wikipedia)

In North America, between 13 and 11.5 ka, around 30 species of large herbivorous mammals became extinct. Much the same occurred in Australia around 45 ka. Both cases roughly coincided with the entry of anatomically modern humans, where neither they nor earlier hominids had lived earlier. Such extinctions are not apparent in the Pleistocene records of Africa or Eurasia. An obvious implication is that initial human colonisation and a collapse of local megafaunas are somehow connected, perhaps even that highly efficient early hunting bands slaughtered and ate their way through both continents. But other possibilities can not be ruled out, including coincidences between colonisation and climate or ecosystem change. As many as thirteen different hypotheses await resolution, one that inevitably makes headline news repeatedly: that both the early Clovis culture and North American megafaunas met their end around the same time as the start of the Younger Dryas millennial cold snap because a meteorite exploded above North America ( One problem in assessing the various ideas is accurately dating the actual extinctions, partly because terrestrial environments rarely undergo the continual sedimentation that builds up easily interpreted stratigraphic sequences. Another is that it is not easy to prove, say, that all giant kangaroos died in a short period of time because of the poor record of preservation of skeletons on land. A cautionary take concerns the demise of the woolly mammoth that roamed the frigid deserts of northern Eurasia and definitely was hunted by both modern humans and Neanderthals. It was eventually discovered that herds still survived on Wrangell Island until the second millennium BC. There is a need for a proxy that charts indirectly the fate of megafaunas plus accurate estimates of the timing of human colonisation. In North America there is a candidate for the first criterion: traces of a fungus (Sporormiella – see Fungal clue to fate of North American megafauna in EPN of January 2010) that exclusively lives in the dung of large herbivores. Fungal spores get everywhere, being wind-dispersed, and in NE US lake cores they fell abruptly at about 13.7 ka. Sporormiella needs to pass through the gut of herbivores to complete its life cycle.

Aboriginal Rock Art, Anbangbang Rock Shelter, ...
Aboriginal Rock Art, Kakadu National Park, Australia (Photo credit: Wikipedia)

The same genus of fungus breaks down dung in Australia. Measuring spore content in sediment on the floor of a Queensland lake shows the same abrupt decline in abundance at between 43 to 39 ka before present (Rule, S. et al. 2012. The aftermath of a megafaunal extinction: ecosystem transformation in Pleistocene Australia. Science, v. 335, p. 1483-1486). Moreover, the fungal collapse is accompanied by a marked increase in fine-grained charcoal – a sign of widespread fires – and is followed by a steady increase in pollen of scrub vegetation at the expense of that of tropical rain forest trees. The shifts do not correlate with any Southern Hemisphere climatic proxy for cooling and drying that might have caused ecosystem collapse. That still does not mark out newly arrived humans as the culprits, as the early archaeological record of Australia, as in North America, is sparse and only estimated to have started at around 45 ka. Yet this is quite strong circumstantial evidence. The 20 or more animals – marsupials, birds and reptiles – with a mass more than 40 kg that formerly inhabited the continent would probably have been ‘naive’ as regards newly arrived, organised, well-armed and clever new predators, as would those of North America and much later in New Zealand, and would have been ‘easy prey’. Incidentally, faunas of both Africa and Eurasia are extremely wary of humans, possibly as a result of a far longer period of encounters with human hunter-gatherers.  In Australia’s case, the use of deliberate fire clearing to improve visibility of game may have had a major role, although it is equally likely that the demise of large herbivores would have left large amounts of leaf litter and dry grasses to combust naturally. Yet the Earth as a whole around 40 ka was slowly cooling and drying towards the last glacial maximum around 20 ka, so human influence may merely have pushed the megafauna towards extinction, such is the fragility of Australia’s ecosystems.

Some megafaunas of the recent past

Harvey was an imaginary, 2 m tall rabbit which befriended Elwood P. Dowd in Mary Chase’s 1944 comedy of errors named after the said rabbit, filmed in 1950 and starring James Stewart as the affable though deranged Dowd. Though not so tall, a giant fossil rabbit (relative to modern rabbits) weighing it at 12 kg has emerged from the prolific Late Neogene cave deposits of Minorca (Quintana, J. Et al. 2011. Nuralagus rex, gen. et sp. nov., an endemic insular giant rabbit from the Neogene of Minorca (Balearic Islands, Spain). Journal of Vertebrate Paleontology, v. 31, p. 231-240). At about 3 times heavier than Barrington my lagomorphophagic (rabbit-eating to the uninitiated) cat, this would have been, to him, a beast best avoided, as the name N. rex might suggest. So unexpected was a gigantic rabbit that, interestingly, it was first mistaken for a fossil tortoise, albeit one lacking a carapace.

Island faunas have long been recognized as havens for peculiar trends in evolutionary successions, either towards dwarfism as in the case of the tiny elephants on which H. floresiensis preyed until quite recently on the Indonesian island of Flores or gigantism as in this remarkable case. As the authors infer, on account of the creature’s ‘…(short manus and pes with splayed phalanges, short and stiff vertebral column with reduced extension/flexion capabilities), and the relatively small size of sense-related areas of the skull (tympanic bullae, orbits, braincase, and choanae)…’ this was a rabbit which sadly could not hop. This un-rabbit-like locomotion may well have been a result of it not having needed to hop, being so large as to challenge seriously the largest Neogene predators on the island – lizards – and thereby saving energy. For much the same evolutionary logic, neither did N. rex have long ears, having less need to detect a stealthy nemesis.

The demise of Late Neogene megafaunas in general has often been ascribed to human intervention. Though N. rex became extinct at around 3 Ma and avoided human predation, later giants did not fare so well. A case in point is the celebrated wooly mammoth, the last of the steppe mammoths, that first appeared in the fossil record of Siberia around 750 ka ago (Nicholls H. 2011. Last days of the mammoth. New Scientist, v. 209 (26 March 2011), p. 54-57). DNA evidence from hairs preserved in permafrost suggests that ancestors of the steppe mammoth line diverged with that of Asian elephants from African elephant ancestors around 5 Ma. Interestingly, ancestral steppe mammoths – without shaggy coats but having the archetypical curved tusks – roamed Africa until 3 Ma when they disappear to reappear in Europe and Asia, yet without adaptation to cold until the onset of northern glaciations around 2.5 Ma. At that point the true steppe mammoths evolved increased tooth enamel needed for a diet of mainly silica-rich grasses to resist wear. The family spread to North America when sea-level fell to expose the sea floor of the Bering Straits. The woolly mammoth is the star partly because specimens periodically turn up almost perfectly preserved in permafrost. This has allowed almost half of a full DNA sequence to be restored. Preserved haemoglobin from a woolly mammoth shares with that from modern musk oxen an ability to release oxygen at temperatures well below zero so that they could function even if their extremities became chilled.

The Woolly Mammoth at the Royal BC Museum, Vic...
Reconstructed woolly mammoth at the Royal BC Museum, Victoria, British Columbia (Image via Wikipedia)

Astonishingly, all elephants urinate so copiously that they soak their range lands in DNA, though it only lingers in ultra cold climes. This bizarre fact encouraged a large team of palaeobiologists to comb frozen soils in an alluvium section in Arctic Alaska for mammoth DNA (Haile, J and 17 others, 2009. Ancient DNA reveals late survival of mammoth and horse in interior Alaska.  Proceedings of the National Academy of Sciences of the USA, v. 106, p. 22352–22357). Mammoth DNA turned up in soils as young as 10.5 ka. Moreover mammoth overlapped with human occupation for several millennia, casting doubt on theories that mammoth extinction resulted either from human predation or the introduction of epidemic disease that might have felled mammoths quickly: they declined gradually. Yet the empirical fact that steppe mammoths in general and the woolly mammoth in particular survived through at least 8 major glacial-interglacial transitions only to become extinct at the start of the current Holocene interglacial period at the same time as humans recolonised the frigid desert of Arctic latitudes, where woolly mammoths could survive except at the last glacial maximum surely points to some influence that arose from human activity.