The extended twilight of the mammoths

One of Charles R. Knight’s wonderful paintings of woolly mammoths walking through the snow of ancient Europe. On display at the Field Museum in Chicago.

When did the last woolly mammoths die?

There is no easy answer to the question. In its heyday the woolly mammoth (Mammuthus primigenius) was distributed across much of the northern hemisphere, from southern Spain to the eastern United States, and the entire species did not simply lay down and die at one particular moment. Some populations (such as the “dwarf” mammoths of Wrangel Island) survived until about 4,000 years ago, but most of the populations that lived on the mainland seem to have disappeared just under 13,000 years ago as part of an extinction event that also wiped out giant ground sloths, saber-toothed cats, enormous bison, and other megafauna.

Climate change, an asteroid impact, a superdisease, and hungry hungry humans have all been considered as triggers for the Pleistocene mass extinction, and determining the pattern of extinction has been important to determining what roles (if any) these mechanisms might have played. But even this is oversimplified. The Pleistocene mass extinction was not a uniform event. Some animals survived in one place while close relatives perished in others, and so a key part of figuring out what mechanism(s) caused the extinction is finding out when certain species became extinct.

The trouble is that the first and last representatives of any species are unlikely to be preserved in the fossil record. This means that even though there are fossil remains of mammoths from Europe and Asia dated to a little less than 10,000 years ago it is still difficult to know whether they were truly the last of the mammoths, the fossil record petered out, or we just have not yet found the bones that would extend the temporal range of the mammoth nearer in time. Such are the frustrations with dealing with fossil remains. Only a tiny fraction of all the organisms that ever lived became preserved as fossils, and an even tinier fraction of those have been discovered, hence Charles Darwin’s lament that the fossil record is “a history of the world imperfectly kept.”

But mammoths were not just shuffling collections of teeth and bones. They were living creatures that bled, defecated, urinated, shed hair, and eventually decomposed, spreading their genetic material all over the land they occupied. This means that there is a second sort of record for the mammoths that, under the right conditions, might be able to provide us with a better idea of where they lived and when they disappeared, and an attempt to mine this rich source of fossil data has just been published in the journal PNAS.

The new study, which was conducted by an international team of 18 scientists, focused on the extinction of mammoths and horses in the interior of Alaska. This has been an important area for research as populations of humans would have reached the interior of North America by going through Alaska first, and so what happened to the megafauna of Alaska might provide important clues about what caused the Pleistocene mass extinction. To find out more, though, the scientists did not go hunting for bones but for locations that have been relatively undisturbed for the past 15,000 years or so. Such locations might be good places to look for “sedimentary ancient DNA” (sedaDNA), genetic traces of ancient life that were frozen into ancient soil layers and remain intact.

The Alaskan permafrost is what made the investigation possible. In warm, wet conditions the genetic material of organisms might leach from one layer of soil to another, but the same quirks of preservation that entombed the bones of mammals also locked up some of their genetic traces. So long as an area of permafrost has remained undisturbed since the time it froze it has the potential to contain genetic traces of extinct animals. The researchers believed that they found such a site on the Yukon Flats near Stevens Village, Alaska, and they looked at a segment of 11,000 to 7,000 year old deposits to see if they could find any remains of ancient life.

What the scientists were after were bits of mitochondrial DNA preserved in the soil. This is the genetic material that is unique to the organelles called mitochonria inside animal cells, and sequences of mitochondrial DNA have been recovered from both extinct horses and mammoths (so the scientists would be able to recognize when they found what they were looking for). What they found was DNA from snowshoe hares, bison, moose, horse, and woolly mammoths, but not all the animals were present at each level. The DNA of different species were extracted from different levels of permafrost, meaning that different animals lived in the area at different times. And, significantly, the DNA retrieved from the mammoths matched other mitochondrial DNA extracted from mammoth remains from North America. This enhanced the likelihood that the mammoth DNA the team found actually came from animals that lived in the area and had not been somehow transported from elsewhere.

A stratigraphic profile of the permafrost layers the team investigated (lower=older, higher=younger). The traces of mammoths and horses were found only at one layer near the top of the column. From the PNAS paper.

So, if the team’s analysis is correct, both mammoths and horses lived in the interior of Alaska between about 11,000 and 7,000 years ago. This is significantly more recent than the youngest fossil remains of horses and mammoths, dated between 15,000 and 13,000 years ago. There are at least two factors that might contribute to this disparity. The first is that fossils from this more recent time were preserved but have not yet been found. More likely, though, is that the populations of both mammoths and horses had dwindled to the point where fossil preservation was becoming increasingly unlikely. There were so few of them that the death of an individual in circumstances amenable to preservation was becoming rarer and rarer.

Either way, this discovery has important implications for the extinction of horses and mammoths in North America. Based upon the fossil data alone it had been hypothesized that both disappeared around the time that humans became established in North America.* Some have taken this association to suggest that humans engaged in a blitzkrieg in which naive New World megamammals were quickly dispatched by the human hunters. If the new evidence is correct, though, humans did not wipe out horses and mammoths overnight. Instead humans lived alongside dwindling populations in Alaska for thousands of years. Likewise, these new findings also contradict the favored hypothesis of one of the study’s authors, Ross MacPhee, who previously proposed that some kind of “hyperdisease” carried by humans (or animals that traveled with humans) quickly wiped out these animals. The pattern of extinction was obviously more protracted.

*[The date of the first arrival of humans in the “New World” is in dispute, and the first records of people in North American might represent a wave of dispersal rather than humans becoming established on the continent. As far as this post is concerned, when humans became established in North America is more important to the question of extinction than “first contact.”]

A map of the world highlighting places in which sedimentary ancient DNA might be likely to be found. Areas of permafrost (grey) might be best, followed by warm and dry areas (yellow), with the liklihood of preservation declining in warm and wet areas (greens). From the PNAS paper.

Granted, this is only evidence from one site, but it is still an intriguing finding. Alaska was the gateway to the rest of North America, and if the “quick kill” models of extinction were correct those animals should have been the first to go. Instead they persisted for thousands of years after humans established themselves in North America, and a map included in the paper suggests that there are probably other similar sites out there. Areas of permafrost that have been undisturbed by flooding or other events since the time of the Pleistocene could provide other windows into the past through the methods the scientists employed in this study.

But one of the things I appreciated most about the paper was that the authors proposed a synthesis of “traditional” paleontology and genetic data to help resolve the questions still surrounding the Pleistocene extinction. The new techniques are not meant to replace what paleontologists have been doing with fossils but to complement those investigations, thus providing intertwining lines of evidence as well as spurring new hypotheses. That is why paleontology is such a vibrant field right now; the traditional methods of fossil hunters are being combined with lab techniques to derive a more comprehensive understanding of the history of life on earth.

No doubt the methods and findings presented in the new paper will be debated, but that’s what science is all about. At the very least, though, I think it is wonderful that the authors of the new paper have taken research in a new direction that might provide better resolution on the timing of the Pleistocene mass extinction. The new paper does not so much solve a problem as open up a number of new questions, and I look forward to future studies of this kind.

[For more, check out John McKay’s post on the same research.]

Haile, J., Froese, D., MacPhee, R., Roberts, R., Arnold, L., Reyes, A., Rasmussen, M., Nielsen, R., Brook, B., Robinson, S., Demuro, M., Gilbert, M., Munch, K., Austin, J., Cooper, A., Barnes, I., Moller, P., & Willerslev, E. (2009). Ancient DNA reveals late survival of mammoth and horse in interior Alaska Proceedings of the National Academy of Sciences DOI: 10.1073/pnas.0912510106