Death in the slow lane

Were the Late Pleistocene extinctions of large mammals the result of climate change or big-game hunting by humans? Reconstructing the biology of extinct species provides clues to the answer.

What caused the Late Pleistocene 'megafaunal' extinctions — the episode between about 50,000 and 10,000 years ago when mammoths, giant ground sloths, giant kangaroos (Fig. 1) and dozens of other large vertebrate species became extinct? The 'overkill' theory holds that human hunters drove the megafauna to extinction. An extreme form of the overkill theory is the 'blitzkrieg' model, in which the humans of Pleistocene times were big-game hunters, selectively and rapidly hunting the largest species to extinction as they swept through newly colonized continents. Others argue that the megafauna were killed off by climate and vegetation change at the end of the last ice age.

Figure 1: Prehistoric victim.

MICHAEL LONG/NHMPL

A reconstruction of Procoptodon, a giant short-faced kangaroo from the Late Pleistocene of Australia, which stood about 3 m high.

Usually, these theories are assessed by using the fossil record to compare the timing of megafaunal extinctions with human arrival on continents and climate change. But as he describes in a paper published in Proceedings of the Royal Society¹, Chris Johnson has taken a fresh approach by systematically comparing the traits of extinct mammal species with those that survived. He finds that it was not large size that predisposed species to extinction, but low reproductive rate. This does not fit in with the blitzkrieg view of events.

Johnson models the close relationship between body size and reproductive rate within taxonomic families of living mammals, and uses this to infer reproductive rates for extinct members of the same families. He cleverly uses a between-family comparison to sidestep the potential circularity of this method, and then shows that the likelihood of extinction was higher for groups with lower reproductive rates, regardless of their body size. Even relatively small mammals became extinct if their fecundity was low enough. Strikingly, the threshold reproductive rate at which the chance of extinction exceeds 50% is roughly the same — about one offspring per female per year — for all groups examined, from 700-gram lemurs to one-tonne cattle. It seems that what was lost during the Pleistocene was not so much the megafauna as the 'bradyfauna' — a whole way of life based on slow life-history.

These results run counter to the idea of extinction by rapid blitzkrieg because, if larger mammals were being selectively hunted, body size as well as reproductive rate should be an important determinant of extinction. However, Johnson noticed another interesting pattern in his data. Those mammal species that bucked the trend, surviving to the present day despite low reproductive rates, tend to be arboreal, nocturnal or inhabitants of dense forests, high latitudes or high altitudes — all of which ought to protect them from human hunters. Johnson interprets this as evidence in favour of more general overkill, where species of all body sizes were harvested.

Johnson's findings complement those of Alroy², who modelled the effects of human hunting on Late Pleistocene extinctions in North America. Without assuming human preference for big game, Alroy's model matched the observed pattern of large-mammal extinction. The two studies agree that even low levels of hunting could have led to extinction: Johnson suggests that, because of the slow reproductive rates of the victims, extinctions need not have occurred rapidly, while Alroy gives a median extinction date in North America of about 900 years after the Clovis hunter–gatherers migrated into the continent from Asia about 13,400 years ago. Whether or not this can be considered rapid blitzkrieg is a matter of perspective — ecologists and palaeontologists are used to working at very different timescales.

Johnson's result leaves the effects of climate change an open question. Species with slow reproductive rates would have been the most vulnerable to environmental degradation as well as to hunting. For Australia, it has been suggested that during the Last Glacial Maximum (around 20,000 years ago), expansion of the arid interior at the expense of the lush coastal zone was a causal factor in Pleistocene extinctions. The finding that, even in Australia, survivors tended to live in forest or other cryptic habitats, could be interpreted as evidence against this climate model. Moreover, some dates for Australian fossil sites^3,4 place megafaunal extinctions well before the Last Glacial Maximum, at around the time of the first evidence of humans.

Perhaps a similar approach to Johnson's could be used to test predictions of the climate hypothesis — comparing, say, extinction patterns in Australian taxa such as reptiles and mammals that differ in their ability to cope with extreme drought. Further, the climatic models for other continents are very different and need to be tested separately. The disappearance of the steppe–tundra and other mixed vegetational environments in the northern continents, to be replaced by zoned forest, has been implicated in extinctions of mammoth, woolly rhino and other species there⁵. Here, the patterns of survival may fit climate/vegetational models at least as well as those invoking overkill.

The Pleistocene extinctions did not touch all regions of the world. In southern Asia, and especially Africa, the megafauna survived largely intact through the Pleistocene to the present. Africa was excluded from Johnson's study, but why did it suffer so few extinctions? Overkill theorists claim that, because humans originated there, African mammals coevolved with people and were thus less 'naive' to human predation. Climate theorists, on the other hand, propose that the array of African vegetation types was modified relatively little by Pleistocene climate change, allowing the survival of even the largest mammals (elephants, rhinos, and so on). Johnson's approach, centred on taxonomic groups, points to a different line of enquiry. Could it be that African mammals, several groups of which are believed to be related and which are collectively known as the Afrotheria⁶, tend to have faster life-history relative to body size than mammals of other continents?

We can draw parallels with the current extinction crisis from Johnson's findings. Alroy points out that the modelled rates of Late Pleistocene extinction were too slow to be perceived by the humans of the time. Today's extinctions have accelerated to an observable pace. Moreover, slow life-history is a strong predictor of current extinction risk in living mammals⁷. Perhaps in another 50,000 years — or even sooner — we will be left only with those that live life in the fast lane.