DNA from animals and plants that populated Siberia from 395,000 to 10,000 years ago has been recovered from permafrost. The discovery of the genetic fragments 鈥� by far the oldest DNA sequences yet authenticated 鈥� will give scientists unprecedented power to reconstruct ancient ecosystems and track them over time.
Eske Willerslev and his team at the University of Copenhagen in Denmark sampled permafrost along 1200 kilometres of Siberia鈥檚 Arctic coast. They drilled to depths of 31 metres, removed cores of soil, and then extracted any DNA preserved in the frozen earth. Using the PCR technique, the team amplified segments of a plant chloroplast gene and several mitochondrial genes from vertebrates.
What they found greatly exceeded their expectations. The permafrost contained DNA from 28 families of trees, shrubs, mosses and herbs some 300,000 to 400,000 years old, plus eight species of mammals including woolly mammoth, steppe bison and musk ox dating back 30,000 years.
Advertisement
鈥淟ooking for plants was a long shot. We were surprised by the diversity in such small samples,鈥� says Willerslev. 鈥淭hen it was a surprise to find signs of both small and big animals.鈥� The team also studied sediments from a temperate New Zealand cave. They found DNA from 29 kinds of plants and two species of the huge, flightless moa bird.
By drilling at widely spaced locations, and at different depths, the team could recreate ancient landscapes and watch them evolve. 鈥淔rom just 2 grams of soil, you can obtain a meaningful sample of the ecosystem,鈥� says Willerslev. That鈥檚 in sharp contrast to studies based on fossils, which reveal information about a particular petrified animal or plant. The scientists hope the method will let them reconstruct ecosystems up to a million years old.
Willerslev was struck by the dramatic shift in vegetation during the past 400,000 years. Herbs dominated the oldest landscapes while shrubs and mosses played secondary roles. The proportion of herbs declined steadily until shrubs came to the fore at the start of the Holocene, 10,000 years ago. Willerslev thinks this shift helps explain the mysterious extinction of the region鈥檚 large mammals about the same time.
Another controversy the soil core technique may settle is whether or not megafauna such as mammoth, steppe bison and horses roamed the region during the last ice age, 22,000 to 16,000 years ago. As well as DNA from the animals themselves, Willerslev鈥檚 team found DNA from 鈥渄iverse and abundant鈥� herb species that would have supported such large grazers. 鈥淥ur data show the major megafauna were definitely in the area,鈥� he says.
The team were determined to avoid the problems of contamination by modern DNA that have bedevilled earlier claims of ancient DNA (快猫短视频, 11 February 1995, p 5). Critics such as Hendrik Poinar and Svante P盲盲bo at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, say that independent researchers have failed to replicate claims of multimillion-year-old DNA from insects in amber, plant fossils or dinosaur bones.
One source of contamination is the drilling process itself. Willerslev鈥檚 team tackled that by spiking their drills with a strain of bacteria. Since none of the bacterial DNA showed up in their samples, they could be sure that they were not injecting modern DNA into ancient deposits, they will report in Science (DOI: 10.1126/science.1084114).
They also sent samples for testing in Britain at the Wellcome Ancient Biomolecules Centre in Oxford. 鈥淲e extracted the DNA ourselves, did the PCRs, and got the same results,鈥� says Thomas Gilbert, a molecular biologist at the centre. 鈥淲e both found mammoth DNA, for example. Neither lab has worked with mammoth DNA before. I think it鈥檚 clear that it鈥檚 real.鈥�
It seems that Willerslev鈥檚 team have got things right, and their claims will stand up to scrutiny. Even Poinar, a vocal critic of previous ancient-DNA claims, applauds the new findings 鈥� with some reservations about dating issues. Robert Wayne, an evolutionary biologist at UCLA, wants to see the approach replicated in different areas and calibrated against known stratigraphic records. Still, he says, 鈥渋t鈥檚 an outstanding result鈥�.