By Ewen Callaway of Nature magazine

Woolly mammoths, woolly rhinos and other large animals driven to extinction since the last ice age each succumbed to a different lethal mix of circumstances. This conclusion--the result of a huge analysis of fossils, climate records and DNA--hints that it could be more difficult than thought to identify the species at greatest risk of disappearing today.

Researchers who studied the fate of six species of 'megafauna' over the past 50,000 years found that climate change and habitat loss were involved in many of the extinctions, with humans playing a part in some cases but not others. But there was no clear pattern to explain why the animals died off, and it proved impossible to predict from habitat or genetic diversity which species would go extinct and which would survive.

"It almost seems like it's a random process," says Eske Willerslev, a palaeo-geneticist at the University of Copenhagen who led the study published online today in Nature. "If you ran the whole experiment again, we would have woolly mammoths and no reindeer, so Santa would drag his sleigh with woolly mammoths."

Fall of the megafauna

Fifty thousand years ago, no fewer than 150 genera of large animals roamed the planet, including woolly mammoths, giant sloths and cave bears. Within 40,000 years two-thirds of them were gone. Some scientists, noting that modern humans were spreading throughout the world around this time, envisaged a blitzkrieg in which technologically savvy people hunted these animals to extinction. The end of an ice age and the habitat changes it wrought led other researchers to lay the blame on climate.

Willerslev says that some of the confusion arose because researchers often used different methods and different animals spread around the world to draw general conclusions about megafaunal extinctions. For a more consistent picture, he and his colleagues charted the population dynamics of woolly mammoths, woolly rhinos, wild horses, reindeer, steppe bison and musk ox.

The researchers created a series of snapshots of the European, Asian and North American ranges of these animals (drawn from climate records and hundreds of fossils) and a rough approximation of their population size (based on ancient mitochondrial DNA sequences) between 42,000 and 6,000 thousand years ago.

Most of these species boomed during the last ice age, which expanded their favoured steppe-tundra habitats, says palaeogeneticist Eline Lorenzen, a member of the team at the University of Copenhagen. As the climate subsequently warmed, woolly rhinos, woolly mammoths and the Eurasian populations of musk oxen went extinct as populations became more and more isolated from one another. But these extinctions happened thousands of years apart, and the animals' ranges changed in different ways. For instance, woolly rhinos roamed much of Europe and Asia until their extinction around 14,000 years ago, whereas the mammoths' range inched northward until they disappeared around 4,000 years ago.

Not all our fault

Humans are off the hook for some extirpations, but still suspected in others. Musk-ox remains are rarely found at sites where humans lived, for example, and the species' range has little overlap with that of humans. Its dwindling range after the ice age suggests that climate change alone probably did for it in Eurasia. Wild horses, on the other hand, lived across Europe and Asia until very recently, and two-thirds of European and Siberian archaeological sites contain their bones, hinting at a role for humans in their disappearance.

Biology offers one explanation for the differences, says Lorenzen. Musk oxen cope poorly with high summer temperatures, for example, and are now found only in the North American Arctic and Greenland. And woolly mammoths reproduced slowly, whereas reindeer are more fecund, "almost like a rodent", she says.

Prediction failure

The team found no way to predict the future extinction of a species, based on either an animal's genetic diversity or the size of its range. "Had I been around 20,000 to 35,000 years ago I would have predicted that reindeer would go extinct while Eurasian musk ox would do well," Willerslev says. This could mean that it will be difficult to determine which modern species are at greatest risk of extinction, he says.

Anthony Barnosky, a palaeobiologist at the University of California, Berkeley, agrees that ice-age extinctions might offer a preview of extinctions driven by a combination of climate change and human expansion today. But he says that a closer look at the biology and ecology of individual species should offer clues as to which are under greatest threat.

Others, such as Hendrik Poinar, a palaeogeneticist at McMaster University in West Hamilton, Canada, caution that the plight of megafauna could be misleading when applied to modern extinctions of much smaller animals, and even plants. "It's interesting what happened to the mammoth," he says. "But when we think about species today, megafauna represent a minute fraction of the fauna we have."

This article is reproduced with permission from the magazine Nature. The article was first published on November 2, 2011.