In January, Proceedings of the National Academy of Sciences, USA published a paper by Gregory Adcock and colleagues, reporting that sequences of ancient mitochondrial DNA had been recovered from 10 Australian fossil humans. While 9 of them fitted within known human variation, it was claimed that the oldest sequence, from the 60,000 year old Mungo 3 fossil, was distinct from those of recent humans, and cast doubt on the theory that modern humans had originated in Africa. Further commentaries and media coverage claimed that the results in fact supported the alternative multiregional model of human origins.
The new Science letter, published June 1st 2001, points out that ancient DNA discoveries are easily contaminated and carry a considerable burden of proof, especially when they involve human sequences or surprising examples of preservation. Both concerns apply in the case of these ancient Australian remains, as DNA is not expected to survive for this length of time outside of cold environments, and similar remains elsewhere have not yielded ancient DNA. Unfortunately the Australian work did not use standard ancient DNA authentication tests, such as independent replication by other laboratories, biochemical studies of bone preservation, and cloning of DNA sequences (to reveal amplification artefacts). Without such data it is impossible to rule out the possibility that the ancient Australian sequences result from modern human contamination of the bone during handling over the years, complicated by DNA damage.
Claimed DNA sequences from dinosaur bones were found to result from this process and the Mungo sequence shows features common in damaged DNA.
Furthermore, reanalysis of the DNA data does not support the Australian claim that Mungo 3 represents a distinct outgroup to modern humans. The new phylogenetic analyses use the same model as the original paper but when additional modern Aboriginal and African sequences are added they show that all the ancient Australian sequences are well within what is expected for modern human variation. Lastly, even if the problems with both the data and the analysis are ignored, the trees published by Adcock's team do not support the multiregional model for modern human origins, as has been claimed, since all the modern human sequences are closely related to each other, while Neanderthal sequences form a distinct outgroup. This is exactly what would be expected from a recent African origin.
* Cooper, A., Rambaut, A., Macaulay, V., Willerslev, E., Hansen, A. & Stringer, C. 2001. Human origins and ancient human DNA. Science 292:1655-1656 (June 1, 2001)
Below are the scientists who have co-authored the letter in Science:
This page is part of the Fossil Hominids FAQ at the talk.origins Archive.
Home Page |
Illustrations | What's New | Feedback | Search | Links | Fiction
Copyright © Jim Foley || Email me