Evolutionary Anthropology 7 ARTICLES Mitochondrial DNA Studies of Native Americans: Conceptions and Misconceptions of the Population Prehistory of the Americas JASON A. ESHLEMAN, RIPAN S. MALHI, AND DAVID GLENN SMITH A decade ago, the first reviews of the collective mitochondrial DNA (mtDNA) data from gional patterning among native popu- Native Americans concluded that the Americas were peopled through multiple migrations lations of North America. All four from different Asian populations beginning more than 30,000 years ago.1 These reports haplogroups are shared with Asian confirmed multiple-wave hypotheses suggested earlier by other sources and rejected the populations, confirming the conclu- dominant Clovis-first archeological paradigm. Consequently, it appeared that molecular sions of classical genetic studies that biology had made a significant contribution to the study of American prehistory. As Cann2 the first Americans migrated from comments, the Americas held the greatest promise for genetics to help solve some of the Asia across the Bering land bridge.4,5 mysteries of prehistoric populations. In particular, mtDNA appeared to offer real potential as Early analyses of restriction fragment a means of better understanding ancient population movements. A decade later, none of the length polymorphism in the entire mi- early conclusions remain unequivocal. Nevertheless, in its maturity, the study of Native tochondria genome showed that these American mtDNA has produced a volume of reports that still illuminate the nature and timing four major clades could be readily dis- of the first peopling and postcolonization population movements within the New World. tinguished by the gain or loss of one or more restriction sites or by the For several reasons, mtDNA has for studying prehistory. The human presence or absence of a 9 base-pair been regarded as particularly useful mitochondrion is an extra nuclear or- deletion in the COII-tRNAlys inter- ganelle having DNA that exists as a genic regions.5 Torroni and cowork- circular molecule 16,569 base pairs in ers7 found that diagnostic mutations length, in which all nucleotide posi- in the CR accompanied the restriction Jason A. Eshleman is a postdoctoral re- tions and coding loci are known.3 Be- markers and the fragment deletion searcher at the University of California, cause this DNA is uniquely maternally that characterize the four haplo- Davis, where he completed his doctoral research examining mtDNA extracted inherited and, unlike nuclear DNA, groups, as is expected of a nonrecom- from prehistoric California sites. He has does not recombine, all changes in bining DNA molecule. Each haplo- coauthored a number of publications ex- mtDNA sequence are the result of ac- amining mtDNA diversity in the New group could be further divided into World. E-mail: [email protected]. cumulated mutations inherited from subclades or discrete haplotypes Ripan S. Malhi is a postdoctoral researcher mother to daughter. In addition, based on additional restriction frag- in the Department of Human Genetics at University of Michigan in Ann Arbor. He has mtDNA mutates an order of magni- ment length polymorphisms or spe- conducted his doctoral research on Native tude faster than does nuclear DNA, cific CR mutations. American migrations in the northeastern with the control region mutating at an and southwestern United States with mod- Although corresponding haplo- ern mtDNA, as well as an analysis of an- even greater rate, making it particu- groups can be found in various Asian cient DNA from the Columbia Plateau. He larly useful for analyses at shallow populations, only founding haplo- has recently authored (with Drs. Eshleman time depths. Finally, mtDNA exists in and Smith) an analysis of mtDNA diversity types of the New World are shared across North America. high copy number in haploid condi- between the two continents, again David Glenn Smith is a professor of anthro- tion. Consequently, it is easily assayed pology at the University of California, Davis. confirming that the Americas were He directs research in Native American ge- in the laboratory and can be recovered initially settled by a limited number of netic variation ancient mtDNA analysis as from prehistoric biological material female immigrants from Asia whose well as non-human primate genetics. Un- in sufficient quantities for amplifica- der his direction, the laboratory of molecu- mtDNA underwent subsequent evolu- lar anthropology at UC-Davis has been in- tion and analysis using the polymer- tion independent of its ancestral form vestigating human migrations with ancient ase chain reaction. 4 DNA from the Great Basin, the Columbia in Asia. The fact that shared haplo- Plateau, the Ohio River Valley, lower Great types on both sides of the Pacific are Lakes, Central California and Central Mex- uncommon has generated consider- ico. HAPLOGROUPS AND able debate as to the size and source HAPLOTYPES of the ancestral population (or popu- Key words: mtDNA, Native Americans, migra- Early studies of Native American lations), as well as the number of tions, ancient DNA mtDNA revealed four major clades, or waves of migration that came out of haplogroups, of haplotypes.4,5 Al- Asia. However, some haplogroups Evolutionary Anthropology 12:7–18 (2003) though they are broadly distributed share more than one haplotype with DOI 10.1002/evan.10048 6 Published online in Wiley InterScience throughout the Americas, these four Asia, and it is not clear whether the (www.interscience.wiley.com). haplogroups exhibit significant re- divergence they represent occurred in 8 Evolutionary Anthropology ARTICLES Asia or the New World. This has made within-haplogroup diversity. Torroni high frequencies of haplogroup A but problematic the use of mtDNA diver- and colleagues7 noted that sequence lacked haplogroup B and exhibited sity to estimate the time of coloniza- diversity within haplogroups A, C, and only low frequencies of haplogroups C tion, the size and source of the ances- D was substantially greater than that and D, whereas Eskimo-Aleut popula- tral population, and the number of in haplogroup B for the populations tions appeared to have high frequen- waves of migration out of Asia. sampled. This, they argued, was evi- cies of haplogroups A and D but dence of a later migration of B matri- lacked haplogroups B and C. How- lines to the New World. The fact that ever, on closer inspection of a large WAVES OF MIGRATION haplogroups A, C, and D are found in number of samples, Merriwether, Although there has been little scien- Eastern Siberia, a likely staging point Rothhammer, and Ferrell16 demon- tific controversy about the Asian ori- for any trans-Beringian migration, strated that groups traditionally clas- gins of Native American populations, whereas haplogroup B is curiously ab- sified as Eskimo and Na-Dene had contention surrounds the question of sent from the region, is consistent measurable frequencies of all four the number of waves of migration. with this argument. These two early haplogroups when larger samples The Americas are home to approxi- migrations were argued to be inde- were assayed. Reasoning that it is un- mately half of the world’s language pendent of later migrations of Na- likely that separate migrations from stocks.8 This extraordinary linguistic Dene. However, this model did not ad- Asia would have introduced exactly diversity among the indigenous dress the relationship of Eskimos to the same four rare Asian types,20 Mer- groups of Native America suggests to other American populations. Because riwether, Rothhammer, and Ferrell16 many comparative linguists that there the Eskimos represent the most recent concluded that the Americas must was either a single colonization sev- arrivals in historical linguistic models have been peopled from a single eral tens of thousands of years ago or of the settlement of the New World, source. Postcolonization forces might that there were multiple colonizations the work of Torroni and coworkers7 subsequently have led to the regional by speakers of different unrelated lan- implies that there were as many as patterning in the Americas that ap- guage phyla.8 There is dispute, how- four independent migrations. peared to differentiate the three hy- ever, about whether or not linguistic Due to the pronounced regional pothesized linguistic phyla. Moreover, evidence supports an early8,9 or later patterning of mtDNA haplogroup and Lorenz and Smith17 showed that first occupation of the Americas.10 haplotype frequency distributions in when a larger, more regionally diverse The pattern of language diversity has the Americas, estimates of genetic di- sample of haplogroup B was analyzed, been used to support the tripartite di- versity are strongly influenced by within-haplogroup diversity was not vision of Native American groups sampling. Moreover, little is known less than that for haplogroups A, C, widely popularized by Greenberg, about the actual number of initial col- and D. Further consideration that Turner, and Zegura.11 Although this onists or the population dynamics in- more than a single founding haplo- division initially was suggested much volved in colonizing a continent free type of one or more haplogroups sur- earlier,12 Greenberg, Turner, and Ze- of other humans. Indeed, the degree vives in modern Native American pop- gura11 proposed that the Amerind, to which Native American popula- ulations14,21 renders comparisons of Na-Dene, and Eskimo-Aleut exhibit tions experienced an initial founder diversity among the four haplogroups parallel