Tight Genes

Suppose I told you that Gerry Shields spent his sabbatical in Utah working on family trees; you'd probably think he was studying the genealogy records kept by the Church of Latter Day Saints. You'd be wrong. Then if I commented that Professor Shields recently upset some folks because of what he said in Washington, D.C., you might suspect he was testifying at one of those federal hearings in which Alaskans always seem to be on both sides. You'd be wrong again.

Shields is a faculty member of the Institute of Arctic Biology at the University of Alaska Fairbanks. For the past several years he has been studying animal populations by means of mitochondrial deoxyribonucleic acid---that is, the DNA that does not get shuffled during sexual reproduction. Nuclear DNA comes half from the father, half from the mother, but only mothers can contribute mitochondrial DNA. This unshuffled DNA changes only by mutation, and the rate at which mutation occurs is fairly regular. (That is, given ideal circumstances, an expert looking at your mitochondrial DNA and that of your distant cousin could inform you that your kinship is, say, about seventh cousins---the mutational clock tells how far back your common ancestor lies.)

Thanks to advances in amplification techniques making it possible to recover enough DNA to study from very small amounts of tissue, Shields and his colleagues from Utah and Siberia have been able to do some sophisticated analyses of relationships among peoples of the north.

The researchers concentrated on a segment of the DNA molecule known as the control region. It evolves fairly rapidly, and thus makes a useful genetic marker among closely related groups. The Shields team obtained samples from volunteers in an array of people whose ancestors were in place before Columbus sailed, and their places ranged from Siberia to Greenland and down the Pacific Northwest coast.

Before the molecular clock in the genetic code was understood, the chief way anthropologists distinguished relationships among the First Nations was by language association and evolution. Elaborate linguistic family trees have been devised for the populations of the Western Hemisphere, according to which the Athabaskan-speaking peoples of interior Alaska and the Yukon are more closely related to Amerindian groups, such as the Plains Indians of the northern U.S. and southern Canada, than they are to the Aleut or Yupik people whose traditional territories border their own.

The DNA markers argue against the linguists' interpretations, and that argument caused some academic flurry at the recent meeting of anthropologists held in Washington, D.C. For example, according to Shields, populations speaking Na-Dene languages (Haida and Athabaskans among the studied peoples) genetically appear in the midst of groups speaking Eskimo-Aleut. Genetic diversity within and among the Siberian groups, including Chukchi, Yukaghir, and Siberian Eskimos, is about twice as large as that within and between New World Eskimos and Na-Dene speakers. This suggests that populations of New World Eskimos and Na-Dene speakers are biologically younger than their Siberian counterparts.

Putting his work together with that of his colleagues studying the DNA of different groups of people of the New World, Shields' team finds that there are four clades---truly distinct branches of the human family tree-represented, and all Eskimo-Aleut and Na-Dene groups fall into only one of these clades.

The people long lumped uncomfortably together as Native Alaskans thus appear to be pretty close cousins, who shared a common ancestry perhaps 5,000 years ago---and it also appears that language evolves far faster than DNA, and perhaps faster than anyone thought.