South from Alaska: a pilot aDNA study of genetic history on the Alaska Peninsula and the Eastern Aleutians

Abstract The Aleutian Islands were colonized, perhaps several times, from the Alaskan mainland. Earlier work documented transitions in the relative frequencies of mtDNA haplogroups over time, but little is known about potential source populations for prehistoric Aleut migrants. As part of a pilot investigation, we sequenced the mtDNA first hypervariable region (HVRI) in samples from two archaeological sites on the Alaska Peninsula (the Hot Springs site near Port Moller, Alaska; and samples from a cluster of sites in the Brooks River area near Katmai National Park and Preserve) and one site from Prince William Sound (Mink Island). The sequences revealed not only the mtDNA haplogroups typically found in both ancient and modern Aleut populations (A2 and D2) but also haplogroups B2 and D1 in the Brooks River samples and haplogroup D3 in one Mink Islander. These preliminary results suggest greater mtDNA diversity in prehistoric populations than previously observed and facilitate reconstruction of migration scenarios from the peninsula into the Aleutian archipelago in the past.

Project MUSE Today s Research. Tomorrow 51nspiration. South from Alaska: A Pilot aDNA Study of Genetic History on the Alaska Peninsula and the Eastern Aleutians Jennifer Raff Justin Tackney Dennis H O'Rourke Human Biology, Volume 82, Numbers 5-6, October-December 2010, pp. 677-693 (Article) Published by Wayne State University Press For additional information about this article http://muse.jhu.edu/journals/hub/summary/v082/82.5-6.raff.html Human Biohg$! IU bln Jiii, mI hT r.rJ.i..! i'--. --1 - ■-1.4 Access Provided by University of Utah at 02/14/11 7:54PM GMTSouth from Alaska: A Pilot aDNA Study of Genetic History on the Alaska Peninsula and the Eastern Aleutians JF.NNIFF.R RAFF,1 JUSTIN TACKNF.Y,1 AND DF.NNIS H. O'ROURKF.1 Abstract The Aleutian Islands were colonized, perhaps several times, from the Alaskan mainland. Earlier work documented transitions in the relative fre­quencies of mtDNA haplogroups over time, but little is known about potential source populations for prehistoric Aleut migrants. As part of a pilot investiga­tion, we sequenced the mtDNA first hypervariable region (HVRI) in samples from two archaeological sites on the Alaska Peninsula (the Hot Springs site near Port Moller, Alaska; and samples from a cluster of sites in the Brooks River area near Katmai National Park and Preserve) and one site from Prince William Sound (Mink Island). The sequences revealed not only the mtDNA haplogroups typically found in both ancient and modern Aleut populations (A2 and D2) but also haplogroups B2 and D1 in the Brooks River samples and haplogroup D3 in one Mink Islander. These preliminary results suggest greater mtDNA diversity in prehistoric populations than previously observed and facilitate reconstruction of migration scenarios from the peninsula into the Aleutian archipelago in the past. Hrdlicka's (1945) inference that the modern population of the Aleutian Islands de­rived from a migration of people from mainland Alaska through the Alaska Penin­sula who replaced the original inhabitants of the island chain approximately 1,000 years ago was predicated on his assessment of change in cranial form over time. Hrdlicka considered the earliest population in the island chain, represented at the time by the Chaluka midden burials [Aigner 1978; Aigner et al. 1976; Laughlin 1963; McCartney and Turner 1966; Veltre and Smith 2010 (this issue)], to be Pre-Aleut, whereas the later population was indistinguishable from the modern Aleut population and hence was termed simply Aleut. Subsequently, Laughlin and Marsh (1951) used the terms Paleo-Aleut and Neo-Aleut to refer to Hrdlicka's two prehistoric morphological groups. We adopted Laughlin's terminology (Smith et al. 2009) to unambiguously distinguish the prehistoric population(s) of the Aleutian archipelago (i.e., Paleo- and Neo-Aleuts) from the modern popula­tion of the region (Aleuts). department of Anthropology, University of Utah, Salt Lake City, UT. Human Biology, October-December 2010, v. 82, nos. 5-6, pp. 677-693. Copyright © 201.0 Wayne State University Press, Detroit, Michigan 48201-1309 KEY words: ALEUTIAN ISLANDS, ADNA, HVRI, MTDNA HAPLOGROUPS, ALEU­TIAN PREHISTORY.678 / RAFFF.TAL. Hrdlicka's interpretation of Aleutian prehistory has been challenged on both archaeological (Laughlin 1980; Laughlin and Marsh 1951) and linguistic grounds [Woodbury 1984; see also Berge 2010 (this issue)] but has been tested only rarely with the type of data that led to its formulation, that is, biological data (e.g., Moor- rees 1957; Ousley 1995; Turner 1967, 1974, 1991). Although Laughlin collected some data on biological variation among Aleut communities in the 1950s (Laugh­lin and Marsh 1951), no real effort to document the pattern of biological variation among Aleut populations, or their possible progenitors, was attempted until the initiation of recent genetic studies [Coltrain et al. 2006, 2010 (this issue); Craw­ford et al. 2010 (this issue); Hayes 2002; Hayes et al. 2003; O'Rourke et al. 2000; Rubicz et al. 2003; Smith et al. 2009; Zlojutro et al. 2006). With the advent of new molecular genetic methods in the 1980s, it became possible to access DNA in prehistoric samples and therefore to characterize pat­terns of genetic variation in prehistory. This innovation made possible the more direct assessment of ancestor-descendant relationships and the direct testing of archaeologically derived hypotheses regarding prehistoric population move­ments. In earlier work we documented the distribution of mtDNA haplogroups in prehistoric inhabitants of the eastern Aleutians (Hayes 2002; Hayes et al. 2003; Smith et al. 2009), compared them to the distribution of mtDNA variation in con­temporary Aleut communities (Rubicz et al. 2003; Zlojutro et al. 2006), and made inferences regarding the origin of Aleut populations and the nature of prehistoric colonizations to the area. In the current pilot study we report sequence variation in the first hypervariable region (HVR1) of the mitochondrial genome of prehistoric inhabitants from three different regions (two on the Alaska Peninsula and one further north in Prince William Sound) in an effort to refine our understanding of population movement from the mainland to the Aleutian Islands over time. Many of the samples examined molecularly here have also been directly dated and char­acterized for isotopic profile [Coltrain 2010 (this issue)]. Materials and Methods Samples. The samples studied here come from three regions and several archae­ological sites on or near the Alaska Peninsula (Figure 1). The oldest material comes from the Hot Springs site located on the Bering coast of the Alaska Peninsula near Port Moller, Alaska (Dumond et al. 1975; Okada and Okada 1974; Weyer 1930; Workman 1966). This large site is well known archaeologically and has been inten­sively studied in recent years to elucidate the prehistoric cultural boundary between Aleut and Yupik Eskimo peoples on the Alaska Peninsula (Dumond 1987a; Masch- ner 1999). Samples from Mink Island derive from the excavation of a single ar­chaeological site by die National Park Service in the 1980s (Jeanne Schapf, personal communication, 2004). Mink Island is a small island located in the North Pacific Ocean just off the Alaska mainland in Prince William Sound. It is included here so that we have a representative of a more northerly prehistoric population not on the peninsula proper. Finally, a group of samples from the interior of the northern regionaDNA in the Alaska Peninsula / 679 ALASKA Hot Springs Site Figure 1. Geographic location of Alaska Peninsula archaeological sites. of the peninsula come from excavations directed by Don Dumond of the University of Oregon in the Naknek/Brooks River area (Dumond 1987b; Harritl 1988; Hilton 1998). A number of samples derive from excavations at the Paugvik site on the Nak- nek River, a pail of the Brooks River drainage (Dumond 1981), and the remainder of this group comes from excavations of a cluster of sites along the Brooks River itself (Dumond 1987b; Harritl 1988). The genome of choice for ancient DNA analyses has traditionally been the mitochondrial genome, and it is also targeted in this pilot project. Unlike the bi- parentally inherited genome of the cell nucleus, the small (16,569 base pahs) circular genome of the mitochondria is exclusively maternally inherited. Lack­ing the molecular editing machinery of the nuclear genome, mtDNA accumulates substitutions at a rapid rate, permitting examination of a population's genetic his­tory in an archaeological time frame. Importantly, each mitochondrion contains potentially hundreds of copies of the mtDNA molecule, and with numerous mito­chondria per cell, thousands of mtDNA molecules may be present for every copy of the nuclear genome. This high copy number feature of mtDNA makes it espe­cially useful for aDNA studies because it maximizes the likelihood of amplifying a target sequence from even seriously degraded molecular material.680 / RAFF F.T AL. The large number of substitutions in mtDNA result in many single nucleo­tide polymorphisms (SNPs), which can be correlated into a few larger groupings called haplogroups. Although many haplogroups exist in populations through­out the world, only five major pan-American haplogroups are known to have been present in the founding population of the Americas, with potentially an additional six minor haplogroups (A2a, A2b, D2a, D3, C4c, D4h3a) playing a role during the initial colonization or subsequent post-Last Glacial Maximum expansions out of Beringia. Four of these five major pan-American mtDNA haplogroups (A2, B2, Cl, and D1) are widely distributed in indigenous Ameri­can populations, whereas haplogroup X2a has a limited geographic distribution, confined to North America, particularly the Great Lakes region (Brown et al. 1998; Perego et al. 2009; Smith et al. 1999). We use HVRI SNPs to define lin­eages of mtDNA haplogroups following O'Rourke and Raff (2010), van Oven and Kayser (2009), and Gilbert et al. (2008). DNA Extraction and Purification. Approximately 0.25-0.50 g of dry-weight bone was decontaminated with ultraviolet light and a 10-min bleach treatment and rinsed with 0.2 |im filtered water. The bone was decalcified and digested over­night with a 55°C incubation in 0.5 M EDTA (pH 8.5), and 1 mg/ml Proteinase K. The digested sample was extracted using the Geneclean for Ancient DNA Kit (MP Biochemicals, Solon, Ohio) following the manufacturer's protocol. For each extraction an extraction blank was run concurrently with the ancient samples. Reported sequencing results are only for those samples that had a corresponding nonamplifying extraction blank. Amplification and Sequencing. Samples were amplified using the poly­merase chain reaction (PCR) in 50 |iL reactions containing 5 |iL of lOx PCR buffer (ABI, Carlsbad, California), 3 |iL of 25 mM MgCl, (ABI), 4 |iL of 10 mM dNTPs (ABI), 0.1 |iL BSA, 0.5 |iL each of the forward and reverse primers (20 |iM; see Table 1 for primer sequences), 32.3 |iLH,0, 0.6 |iLAmpliTaq Gold DNA Polymerase (ABI), and 4 |iL template. Reactions were hot-started for 10 min at 95°C before the first amplification cycle, followed by 40 cycles of denaturation at 95°C for 30 s, annealing at 51 °C for 30 s, and extension at 72°C for 30 s. A final extension of 5 min at 72°C followed the 40-cycle amplification reaction. Ten microliters of the PCR product was visualized on 2.5% agarose gel adjacent to a size standard. For each PCR reaction a PCR blank was run concurrently with the ancient samples. Samples were processed for sequencing only if the cor­responding PCR blank did not amplify. Chosen amplified PCR products were cleaned using either the QIAquick PCR Purification Kit (Qiagen, Germantown, Maryland) or the UltraClean PCR Clean-Up kit (MoBio Laboratories, Carlsbad, California). Purified PCR products were sequenced on both strands at the Univer­sity of Utah HSC Core Facilities. Sequences were manually corrected; cases of heteroplasmy are reported as Y (pyrimidines) or R (purines) pending cloning to identify contributing sequences.aDNA in the Alaska Peninsula / 681 Table 1. Primer Sequences Used for HVRI Amplification and Sequencing Primer Designation Sequence 5'-3' H VR 1_P1 i! HVRI_PIRc HVRl_P21!b HVR1_P2R HVR1_P31! HVR1_P3R GTTCTTTCATGGGGAAGCAG TTGATGTGGATTGGGTTTTT AAAACCCAATCCACATCAAA GGGTGGGTAGGTTTGTTGG CCCACTAGGATACCAACAAACC ATTGATTTCACGGAGGATGG Contamination and Quality Control. Contamination from modern nucleic acids is a constant risk when working with aDNA (Cooper and Poinar 2000; Gil­bert et al. 2003,2005; Handt et al. 1994). We took several measures to both reduce the possibility of contamination and detect it when it did occur. These measures included the following: (1) Only bone samples without pathogenic lesions were chosen for extraction. This selection reduces the risk for the interior of the bone samples to become contaminated in situ and during routine handling of the samples by prior investigators. (2) All samples were pretreated with bleach to remove surface contamination before aDNA extraction. (3) Protocols used freshly bleached and/or ultraviolet-light cross-linked dedicated instruments, labware, and aerosol-resistant filtered pipettor tips. Reagents were certified DNA-free by the manufacturer and aliquoted into single-use volumes. (4) Pre- andpost-PCR activities were performed in separate laboratories located at opposite ends of the building, each fully dedi­cated to aDNA analysis. Amplifications of modern DNA positive controls are never performed in this building. (5) Laboratory personnel were masked, coated, sleeved, and gloved. All garments were bleached before entry into the pre-PCR laboratory, and gloves and sleeves were periodically bleached throughout the laboratory work. (6) Preamplification procedures were performed in a HEPA-filtered, positively pressured Airclean 600 sterile enclosure equipped with an internal ultraviolet light, housed in a HEPA-filtered, positively pressured, fully ultraviolet-light-equipped pre-PCR room. (7) The thermal cycler used for all amplifications is located in a separate room of the building, isolated from both the pre- and post-PCR laborato­ries. (8) As previously described, negative extraction and PCR controls were used to identify false-positives. (9) Sequences were confirmed with a replicate amplifica­tion for each amplicon from a second extraction. If the replicate was not identical in sequence to the original, a third extraction, amplification, and sequence reaction was performed to resolve the ambiguity. A fuller explanation of laboratory procedures and methods can be found in Hayes (2002) and Smith et al. (2009). We obtained full or partial HVSI sequences from eighteen samples: four from Port Moller, eight from Brooks River, and six from Mink Island (Table 2). Results682 / RAFF F.T AL. Table 2. Sequence Variants Observed in HVRI Sequences of Three South Alaskan Archaeological Samples Site Sample Coverage 16092 16093 16111 16129 16136 16189 REF 16012-16422 T T C G T T Port Moller HS 82-3 16012-16422 Port Moller HS 84-2 16012-16422 T C Port Moller HS 82-2* 16012-16296 Y A Port Moller HS 82-5 16012-16422 Brooks River NAK2B4* 16162-16415 - - Brooks River NAK2B3* 16055-16254, 16266-16422 Brooks River NAK2B1 16012-16422 T Brooks River NAK2B2* 16162-16425 T Brooks River BR1 Bi.i 2 • 16023-16253 C Brooks River BR1 B11961* 16023-16256, 16266-16422 T A Brooks River BR5 Bur2* 16012-16254, 16260-16425 C Brooks River BR9 Burl 16012-16422 T Mink Island XMK-030 Southwest slope 16012-16422 C Mink Island XMK-030 Unit 15A/Bur2 16012-16422 T Mink Island XMK030 misc. fragments 16012-16422 C Mink Island XMK030 Unit 10A/Bur2 16012-16422 C A Mink Island XMK030 Unit 15h/Bur4 16012-16422 A Mink Island XMK030 Bur3/IndA 16012-16422 c A *, Shortened or incomplete coverage. R, purine heteroplasmy. Y, pyrimidine heteroplasmy.aDNA in the Alaska Peninsula / 683 16217 16223 16224 16271 16290 16298 16311 16319 16325 16327 16362 Haplogroup "C Dale (BP) T C T T C T T G T C T rCRS T C C D2 N/A T T A Y C A2 2070 T C - - - - - - D2 C C Unknown N/A C C A C T C Dl? N/A T A C A2 N/A T R c A2 N/A T A c A2 N/A C B 1295 Y R A2 N/A C B 938 T A C A2 1318 T C C D2 1029 T T A C A2 1135 T A C D3 1110 T C D2 1078 T C D2 1215 T C C D2 1130684 / RAFF FT AL. Of these sequences, we could preliminarily assign haplogroups to seventeen in­dividuals: Eight belonged to haplogroup D (including six D2 individuals, one D3 individual, and one D1 individual), seven belonged to haplogroup A (all A2), and two individuals belonged to haplogroup B2. Haplogroups C and X were not observed. Sample sizes in this pilot study are too small to allow for meaningful sta­tistical comparisons, but some useful preliminary observations can be made. As expected, the most frequent haplogroups present are haplogroups A and D, as was also true in the earlier eastern Aleutian study (Smith et al. 2009). But two individuals belonging to haplogroup B2 were also present in the Brooks River col­lection. Although extreme caution must be exercised because of the small sample sizes (e.g.. Smith et al. 2009), the relative haplogroup frequencies appear to differ considerably between the three sites (Table 3). The haplotypes present showed considerable sequence diversity. Mitochon­drial DNA haplotypes A2a and A2b are frequent in Arctic populations, but we did not observe the diagnostic transitions at np 16192 and np 16265 for these two haplogroup A sublineages in any of our samples. Either these samples represent greater haplotype diversity than expected or key SNPs are missing because of a damaged template. We are aware that a damaged template is present in our samples (the D1 sample from the Brooks River area, for example, has ample evi­dence of damage) but detailing the extent of such damage to sequences requires analyses beyond the scope of this pilot project (Axelsson et al. 2008; Briggs et al. 2007; Ho et al. 2007). It is worth noting that the full HVRI sequence of one Mink Island sample (XMK030-B3) was previously sequenced by R. Malhi under a contract with the National Park Service, and our sequence of this sample is identical to the one reported by Malhi (n.d.). In addition, L. Marchani provided partial sequence data on three different Mink Island samples (XMK030 Unit 10a/Bur2, XMK030 Unit 15b/Bur4, and XMK030 Unit 15a/Bur2) as part of a feasibility study on these specimens three years ago in another laboratory, and these sequences were also replicated in the current study. Thus several sample sequences reported here have been independently replicated by other investigators in other laboratories, provid­ing evidence of authenticity of and confidence in the reported ancient sequences. Discussion The Hot Springs Site. The smallest set of remains examined derives from the Hot Springs site located near Port Moller, Alaska, on the Bering Sea coast near the midpoint of the peninsula (see Figure 1). As indicated by Coltrain [2010 (this issue)], these samples come from the most ancient of the three sites examined. The mean age of samples examined from this site is more than 2,000 years, but at least two samples are dated to approximately 3300 BP. Perhaps the age of the remains from this site helps to explain why it proved more difficult to obtainaDNA in the Alaska Peninsula / 685 Table 3. Haplogroup Frequencies Present in Samples from Peninsula Sites Site Haplogroup A (l'kI Haplogroup D f fA ) Haplogroup B ('/t I Mink Island Port Moiler Brooks River 16.7 33.3 62.5 83.3 66.6 12.5 0 0 25 nucleic acids for analysis from here than from samples from either Mink Island or the Brooks River area. Only three of four Hot Springs site samples examined resulted in sufficient sequence data to provisionally assign haplogroup lineages. As indicated in Table 3, two of these samples represent lineages from mtDNA haplogroup D, one was lineage A2, and the fourth sample contained no SNPs unambiguously associated with any of the mtDNA haplogroups. Both of the haplogroup D lineages belong to lineage D2, although one does not exhibit the transition at np 16129 typically associated with this lineage and the other is characterized by an incomplete HVRI sequence (terminating at np 16296) but is also clearly a representative of lineage D2. The single individual from the Hot Springs site representative of haplogroup A has SNPs indicative of lineage A2, a pan-American lineage widely distributed throughout both North and South America, including both ancient and modern Aleuts. Additional sequence data are required to determine whether this individ­ual is lineage A2a or A2b, because the SNP differentiating these sublineages is not present in this sample. As with diagnostic SNPs absent in other sequenced samples, this may be the result of template damage (Briggs et al. 2007; Ho et al. 2007). Finally, one Hot Springs site sample exhibited transitions at np 16224C. and np 16311C but was otherwise characterized by the Cambridge Reference Se­quence. These two SNPs are not diagnostic for any of the founding mtDNA hap­logroups in the Americas. Brooks River Area. The eight samples from the Brooks River area of the interior peninsula provided unexpected results. Five of the samples had HVRI sequences representative of mtDNA lineage A2. A single individual represented haplogroup D, and with a transition at np 16325C. appears to represent lineage D1 rather than the more commonly observed D2 in the Aleutian region (although evidence of postmortem sequence damage in this sample makes the inference of lineage D1 less secure than we would like). The high frequency of haplogroup A and the possible presence of haplogroup D1 here suggest less of an Aleut influ­ence and perhaps greater similarity to Yupik Eskimo populations of the region. A haplotype network for the haplogroup D lineages observed across these three sites is given in Figure 2. It is clear that the distribution of lineages is geographi­cally structured in these few samples, with the putative D1 lineage in Brooks River being most divergent from the lineages at the other two archaeological sites. Nevertheless, the other sites are also distinct in the network, with the reticulations6 8 6 / RAFF F.T AL. Figure 2, Median network of D haplogroups from three sites in southern and southwestern Alaska. indicative of small sample sizes that likely have not captured the full extent of the sequence diversity present in these lineages in prehistory. It is worth noting that the network forhaplogroup A does not exhibit the same geographic structure (data not shown). The most unexpected result from these analyses, however, is the pres­ence of two HVRT sequences consistent with mtDNA haplogroup B2. These two samples share transitions at np 16189C and np 16217C that are diagnostic for haplogroup B2 (lineages B2 and B4 cannot be distinguished with HVRT, but B2 is the expected American founding haplogroup and we infer it here). Given that neither sample exhibits any other SNPs observed in other peninsula samples or any SNPs that suggest any other haplogroup, the inference of haplogroup B2 for these two samples is most likely, if unexpected. Haplogroup B2 is common among populations farther south in the Americas and is present in coastal popu­lations of southwest British Columbia (e.g., Bella Coola and Nuu-Chal-Nuth; Ward et al. 1991) but has never been reported in precontact northern North American populations and has been found only rarely in contemporary popula­tions of the region (Helgason et al. 2006; Saillard et al. 2000). Haplogroup B2 is also apparently absent from northern populations in Siberia (Shields et al.aDNA in the Alaska Peninsula / 687 1993). However, Merriwether et al. (1995) did report three haplogroup B2 indi­viduals from Old Harbor on Kodiak Island, which lies adjacent to the northern peninsula region that contains the Brooks River area, just across Shelikof Strait. Both of the haplogroup B samples reported from the Brooks River area of the peninsula were directly dated by Coltrain [2010 (this issue)], with one dating to 799 BP and the other to 493 BP, thus precluding modern intrusive burials as a source of the unexpected haplogroup. Both sequences have been replicated, and none of the laboratory personnel possess this mtDNA lineage, suggesting that recent contamination is an unlikely explanation. Moreover, no ancient samples possessing this haplogroup have been analyzed in our laboratory, so some form of crossover contamination is even less likely. We acknowledge that this result is unexpected and has not been reported for other groups in northern North America. However, if it is a contaminant, it seems likely to have been introduced before collection for molecular analyses. The pilot analyses conducted to date on these samples are not sufficient to deter­mine whether this may be the case. Because we have no reason to believe that these results are spurious, they imply that the prehistoric population at the northern end of the Alaska Peninsula 800-1,000 years ago was characterized by a greater degree of genetic diversity than has typically been observed, even in many modern samples from the area. It also implies a greater degree of movement of populations and/or individuals, per­haps across substantial distances, and rates of admixture not generally associated with prehistoric populations. Although the presence of haplogroup B in modern Kodiak Island communities (Merriwether et al. 1995; Merriwether and Ferrell 1996) could have been the result of recent migration and admixture, the presence of the same rare haplogroup in the region in precontact times implies a modem link to a deeper history. Mink Island Site. The Mink Island samples are from the youngest site exam­ined, are not from the peninsula but rather from Prince William Sound just off the northeast coast of the Kenai Peninsula, and have an average age of only about 500 years. Like the Brooks River samples, all the Mink Island material yielded sufficient DNA for HVRI sequencing and mtDNA lineage assessment. Of the six samples examined, five represented haplogroup D and one was from lineage A2. Four of the haplogroup D individuals were lineage D2. However, all are missing at least one diagnostic SNP for this haplogroup (np 16129A or np 16271C) while exhibiting the other. Moreover, three of the four share a transition at np 16092C that is widely distributed in modern Arctic populations. The other haplogroup D individual in this collection possessed a transition at np 16319A that may indicate membership in lineage D3, although one informative SNP for this haplogroup (np 16173) was not observed because it lay too close to an amplification primer for a clear sequencing read. Regrettably, this SNP was not known when the primers were designed. Nevertheless, the overall sequence motif of this sample is more consistent with haplogroup D1 than D2.6 8 8 / RAFFF.TAL. Collectively, these late prehistoric sequences may provide additional insight into the previously observed transition in mtDNA frequencies in the eastern Aleu­tians after about 1000 BP. Using data from the Chaluka Midden site on Umnak Is­land, Smith et al. (2009) demonstrated that before 1000 BP haplogroup A was the predominant haplogroup in the eastern Aleutian region. Based on a small number of samples (n = 11) dating to between 1000 BP and 3500 BP, these investigators identified eight samples as haplogroup A (73%) and only three (27%) as hap­logroup D. This was a clear reversal of frequencies observed in post-AD 1000 samples and in modern Aleuts [Crawford et al. 2010 (this issue); Zlojutro et al. 2006]. The Hot Springs site at Port Moller yielded only three samples that could be identified with respect to lineage and two of these were lineage D2. Unfortu­nately, these two samples were not directly dated, but if they are of comparable age to the dated Hot Springs site samples [Coltrain 2010 (this issue)], they would be expected to be contemporaneous with the earlier material at Chaluka. This in­ference, if correct, suggests that individuals bearing haplogroup D lineages were present not far from the eastern Aleutian region but had yet to move further out in the archipelago in significant numbers, if the early Chaluka material is reflec­tive of the mtDNA variation in the region at the time. At the least it is logical to conclude that haplogroup A occurred at significantly higher frequency than hap­logroup D in the early eastern Aleutians, but the same appears not to be true only a few hundred kilometers eastward at Port Moller, although this tentative inference is tempered by the small number of samples examined from this site to date. The Mink Island site represents the most recent occupation examined here, and the distribution of mtDNA haplogroups mirrors that in the eastern Aleutians, with a predominance of haplogroup D, despite the great distance separating them. Thus by 500 BP or so, the Aleut pattern of a high frequency of haplogroup D and a low frequency of haplogroup A may have become established throughout the southern Alaska coastal region. Whether the expanded mtDNA variability exhib­ited by the few Brooks River samples is common to a broader geographic region in southern Alaska in prehistory awaits enhanced sampling from additional early sites in the region. A Temporal Perspective. The mtDNA haplogroup frequency shift docu­mented by Smith et al. (2009) in the Chaluka midden site series at about 1000 BP could be the result of a population bottleneck, that is, the result of genetic drift following a population crash. Indeed this is essentially the conclusion reached by Laughlin in his reexamination of Hrdlicka's replacement hypothesis. Laugh­lin and Marsh (1951) noted that the dolicho-/brachycranic dichotomy observed by Hrdlicka (1945) also characterized the geographic diversity of contemporary Aleut populations. They invoked selection in the larger eastern Aleutian region as the mechanism that gave rise to a change in cranial shape but small population size and genetic drift for the persistence of dolichocephaly in modern Aleuts of the region. Thus they attributed a considerable portion of the diversity observed in the Aleutians, both geographically and temporally, as the result of stochasticaDNA in the Alaska Peninsula / 689 forces in populations of small size and low density, at least in the central and west­ern Aleutians. The data reported by Crawford et al. [2010 (this issue)] seems to mirror Laughlin's view of contrasting variation between the western/central Aleu­tian islands and the eastern islands of the chain. In the absence of samples from the central and western Aleutians for aDNA study, it is not possible to directly test this model of Aleutian population history. However, it was possible to test for the effects of small size and drift in the mtDNA haplogroup frequency data available from the eastern Aleutians. Using Cabana et al.'s (2008) drift simulation program Nsitu, Smith et al. (2009) tested for the effects of drift on changes in haplogroup frequencies over time and statistically rejected drift as the sole mechanism structuring haplogroup frequencies in the eastern Aleutian data (based on published frequencies). Thus the most parsimonious explanation of the ancient genetic data from the eastern Aleutians appears to be a population expansion of (perhaps) culturally different but genetically related (i.e., possessing only mtDNA haplogroups A and D in any appreciable frequency) people into the eastern Aleutians about 1,000 years ago, followed by amalgamation of the two previously distinct gene pools. The most recent genetic data on contemporary Aleut communities suggests that this expan­sion and amalgamation has been overlain recently by admixture with non-Aleut populations [Crawford et al. 2010 (this issue); Zlojutro et al. 2006]. The pilot data presented here suggest that populations inhabiting areas of the Alaska Peninsula before or immediately after the documented transition in hap­logroup frequencies in the eastern Aleutians were characterized by a greater de­gree of genetic variation than previously assumed. They could easily have served as a platform for migration and change in the Aleutian region to the west. More­over, the early dates associated with the Hot Springs site coupled with the (pre­sumed) relatively high frequency of haplogroup D lineages shared with modern Aleuts suggest that the movement of peoples into the archipelago approximately 1.000 years ago, which appears to be associated with an increase in haplogroup D frequency, may have been a migration of coastal peoples derived from Bering coast populations on the peninsula. The presence of haplogroups B2 and D1 in the Brooks River samples sug­gests a resident population in the interior of the peninsula that likely had greater contact with populations farther east or elsewhere on the mainland-populations that were characterized by the full suite of haplogroups found in many Native American populations and that would have also served as a source population for new genetic diversity as peninsula populations moved westward toward the islands in the Aleutian chain late in prehistory. Finally, we want to emphasize the pilot nature of the data on prehistoric populations presented here. Collectively, they represent only 17 individuals who lived in three different geographic areas of south Alaska over a span of perhaps 3.000 years. We cannot assume that they are precisely representative of the populations of which they were members. As a group, they exhibit nearly equal frequencies of haplogroups A and D-not too dissimilar from pooled ancient690 / RAFF F.T AL. samples from the eastern Aleutians or from modern Aleut populations. However, as noted by Smith et al. (2009), increasing sample sizes per site revealed both site differences in haplogroup frequencies as well as temporal changes. Given the temporal and isotopic distinctions between these sites, described by Coltrain [2010 (this issue)], we think it best for the moment to consider them as distinct populations. The Brooks River population, in particular, appears distinctive with respect to the other peninsula and Aleut samples, both ancient and modern, exam­ined by us and others [Crawford et al. 2010 (this issue); Rubicz et al. 2003; Smith et al. 2009; Zlojutro et al. 2006]. Thus the interior of the peninsula may represent a greater linkage with the larger Native American gene pool of the continent than the smaller, marine-adapted populations living on both the Bering and Pacific coasts. As such, it may represent a conduit for genetic exchange from outside the region and a source of diversity carried by late prehistoric migrants to the islands. In this scenario the eastern reaches of the peninsula are important to our under­standing of the population dynamics of genetic change in the Aleutian system and to critical tests of Hrdlicka's craniometric-based hypothesis of population replace­ment and the distribution of Paleo- and Neo-Aleuts in the archaeological record. Further work to increase the samples sizes of the peninsula samples as well as the sequence coverage beyond HVRI will be critical to robust tests of population his­tory models of the region. Acknowledgments We offer our appreciation and thanks to the Council of Katmai De­scendants and the Native Village of Nelson Lagoon for providing permission to analyze the genetic variation in the Brooks River area and Hot Springs site, respectively. We are grateful to Jeanne Schapf of the National Park Service for providing access to the Mink Island material and to Debra Corbett of the U.S. Fish and Wildlife Service for facilitating communication and sampling of the material from the Hot Springs site. Herbert Maschner of Idaho State University also provided access to some Hot Springs site materials. We are deeply indebted to Don Dumond for aiding in access to the Brooks River area material curated at the University of Oregon Museum of Natural History and for stimulating our interest in the population history of the peninsula. We thank Elizabeth Marchani for early sequence work on the Mink Island samples. This work was supported by the National Science Foundation through grants OPP-0327641 and OPP-0732486 awarded to D. H. O'Rourke, by the University of Utah Research Committee, and by the College of Social and Behavioral Science. University of Utah. Received 16 June 2010; accepted for publication 29 June 2010. Literature Cited Aigner, J. S. 1978. The l.ithic Remains from Anangula, an 85GO-Year-Otd Aleut Coastal Village. Tubin­gen, Germany: Archaeologica Venatoria 1 nstitut fiir Urgeschichte der Universitat Tiibingen. Aigner, J. S., R. !!ullem, D. Veltre et al. 1976. Preliminary reports on remains from Sandy Beach Bay, a 4300-5600 BP Aleut village. Arctic Anthropol. 13:83-90.aDNA in the Alaska Peninsula / 691 Axelsson, E„ E. Willerslev, M. T. P. Gilbert et al. 2(X)8. The effect of ancient DNA damage on infer­ences of demographic histories. Mot. Biol Evol. 25:2181-2187. Berge. A. 2010. Origins of linguistic diversity in the Aleutian Islands. Hum. Biol. 82(5-6):557-581. Briggs. A. W„ U. Stenzel, P. L. F. Johnson et al. 2007. Patterns of damage in genomic DNA sequences from a Neanderthal. Prov. Natl. Acad. Sci. USA 104:14.616-14.621. Brown. M. D., S. H. Hosseini, A. Torroni et al. 1998. mtDNA haplogroup X: An ancient link between Europe/Western Asia and North America'.' Am. J. Hum. Genet. 63:1852-1861. Cabana. G. S.. K. Hunley, and F. A. Kaestle. 2008. Population continuity or replacement'.' A novel computer simulation approach and its application to the Numic expansion (western Great Basin. USA). Am. J. Phys. Anthropol. 135(4):438-447. Coltrain, J. B. 2010. Alaska Peninsula stable isotope and radioisotope chemistry: A study in temporal and adaptive diversity. Hum. Biol 82(5-6):613-627. Coltrain. J. B„ M. G. Hayes, and D. H. O'Rourke. 2006. Hrdlicka's Aleutian population-replacement hypothesis: A radiometric evaluation. Cun: Anthropol 47(3):537-548. Cooper. A., and H. N. Poinar. 2000. Ancient DNA: Do it right or not at all. Science 289:1139. Crawford. M. H„ R. C. Rubicz, and M. Zlojutro. 2010. Origins of Aleuts and the genetic structure of populations of the archipelago: Molecular and archaeological perspectives. Hum. Biol 82(5-6):695-717. Dumond, D. 1981. Archaeology on the Alaska Peninsula: The Naknek Region, 1960-1975. Univer­sity of Oregon Anthropological Papers 21. Eugene: Museum of Natural and Cultural History. University of Oregon. Dumond. D. E. 1987a. A reexamination of Eskimo-Aleut prehistory. Am. Anthropol 89:32-56. Dumond. D. E. 1987b. Prehistoric Human Occupation in Southwestern Alaska. University of Oregon Anthropological Papers 36. Eugene: Museum of Natural and Cultural History. University of Oregon. Dumond. D., L. Conton, and H. Shields. 1975. Eskimo and Aleuts on the Alaska Peninsula: A reap­praisal of Port Moller affinities. Arctic Anthropol 12:49-67. Gilbert. M. T. P.. H.-J. Bandelt, M. Hofreiter et al. 2005. Assessing ancient DNA studies. 1'r. Ecol. Evol. 20:541-544. Gilbert. M. T. P.. T. Kivisild, B. Gronnow et al. 2008. Paleo-Eskimo mtDNA genome reveals matrilin- eal discontinuity in Greenland. Science 320:1787-1789. Gilbert. M. T. P.. E. Willerslev. A. J. Hansen et al. 2003. Distribution patterns of postmortem damage in human mitochondrial DNA. Am. J. Hum. Genet. 72:32-4-7. Handt, O., M. Hoss, M. Krings et al. 1994. Ancient DNA: Methodological challenges. Experientia 50:524-529. Harritt, R. 1988. The Lute Prehistoric of Brooks River, Alaska. University of Oregon Anthropological Papers 38. Eugene: Museum of Natural and Cultural History. University of Oregon. Hayes. M. G. 2002. Paleogenetic Assessments of Human Migration and Population Replacement in North American Arctic Prehistory. Ph.D. dissertation. University of Utah. Hayes. M. G., J. B. Coltrain. and D. H. O'Rourke. 2003. Mitochondrial analyses of Dorset. Thule. Sadlermiut, and Aleut skeletal samples from the prehistoric NorthAmericanArctic. hi Mummies in a New Millennium: Proceedings of the 4th World Congress on Mummy Studies. N. Lynnerup, C. Andreasen. and J. Berglund, eds. Copenhagen: Danish Polar Center. 125-128. Helgason, A.. G. Palsson, H. S. Pedersen et al. 2006. mtDNA variation in Inuit populations of Greenland and Canada: Migration history and population structure. Am. J. Phys. Anthropol. 130:123-134. Hilton. M. 1998. Integrating natural and cultural resource management in Katmai National Park and Preserve. CRM [Cultural Resource Management] 9:17-19. Ho. S. Y. W„ T. H. Heupink, A. Rambaut et al. 2007. Bayesian estimation of sequence damage in ancient DNA. Mol. Biol. Evol. 24:1416-1422. Hrdlicka, A. 1945. The Aleutian and Commander Islands and Their Inhabitants. Philadelphia: Wistar Institute of Anatomy and Biology.692 / RAFF F.T AL. Laughlin, W. S. 1963. Eskimos and Aleuts: Their origins and evolution (physiological and cultural adaptation to facilitate the evolutionary success of Eskimo-Aleut stock). Science 142:633-645. Laughlin. W. S. 1980. Aleuts: Survivors of the Bering Lund Bridge, New York: Holt. Rinehart & Winston. Laughlin. W. S.. and G. H. Marsh. 1951. A new view of the history of the Aleutians. Arctic 4:75-88. Malhi, R. n.d. Report on the Extraction of DNA from the XMK-U3U-BJ. B2. B3u. und B3b Teeth, Final report to the U.S. National Park Service. Maschner, H. 1999. Prologue to the prehistory of the low'er Alaska Peninsula. Arctic Anthropol. 36:84­102. McCartney, A. P., and C. G. Turner II. 1966. Stratigraphy of the Anangula unifacial core and blade site. Arctic Anthropol. 3:28-40. Merriw'ether, D. A., and R. E. Ferrell. 1996. The fom founding lineage hypothesis: A critical re-evalu­ation. Mot. Phylogenet. Evot. 5:241-246. Merriw'ether, D. A., F. Rothhammer, and R. E. Ferrell. 1995. Distribution of the fom founding lineage haplotypes in Native Americans suggests a single w>ave of migration for the New' World. Am. J. Phys. Anthropol. 98:411-430. Moorrees, C. F. A. 1957. The Aleut Dentition: A Correlative Study of Dentul Characteristics in an Eskimoid People. Cambridge, MA: Harvard University Press. Okada, H„ and A. Okada. 1974. Preliminary report of the 1972 excavations at Port Moller, Alaska. Arctic Anthropol. 1 l(suppl.):l 12-124. O'Rourke, D. H„ M. G. Hayes, and S. W. Carlyle. 2000. Spatial and temporal stability of mtDNA haplogroup frequencies in Native North America. Hum. Biol 72(1): 15-34. O'Rourke, D. H„ and J. A. Raff. 2010. The human genetic history of the Americas: The tinal frontier. Cun: Biol. 20:R202-R207. Ousley, S. D. 1995. Relationships between Eskimos, Amerindians, and Aleuts: Old data, new' perspec­tives. Hum. Biol. 67:427-4.58. Perego, U. A., A. Achilli, N. Angerhofer et al. 2009. Distinctive but concomitant Paleo-Indian migra­tion routes from Beringia marked by tw'o rare mtDNA haplogroups. Cun: Biol 19:1-8. Rubicz, R. C„ T. G. Schurr, P. L. Babb et al. 2003. Mitochondrial DNA variation and the origins of the Aleuts. Hum. Biol 75(6):809-835. Saillard, J„ P. Forster, N. Lynnerap et al. 2000. mtDNA variation among Greenland Eskimos: The edge of the Beringian expansion. Am. J. Hum. Genet. 67:718-726. Shields, G. F„ A. M. Schmiechen, B. L. Frazier et al. 1993. mtDNA sequences suggest a recent evo­lutionary divergence for Beringian and northern North American populations. Am. J. Hum. Genet. 53:549-562. Smith, D. G„ R. S. Malhi, J. Eshleman et al. 1999. Distribution of mtDNA haplogroup X among native North Americans. Am. J. Phys. Anthropol. 110:271-284. Smith, S., M. G. Hayes, G. Cabana et al. 2009. Inferring population continuity versus replacement with aDNA: A cautionary tale from the Aleutian Islands. Hum. Biol. 81(4): 19-38. Turner, C. G„ II. 1967. The Dentition of Arctic Peoples. Ph.D. dissertation. University of Wisconsin, Madison. Turner. C. G., II. 1974. The use of prehistory for direct comparative baselines in the study of Aleut mi­croevolution and adaptation. In International Conference on the Prehistory and Puleoecology of the Western North American Arctic and Subarctic. S. Raymond and P. Schledermann, eds. Calgary, Canada: University of Calgary Archaeological Association, 205-215. Turner, C. G„ II. 1991. The Dentition of Arctic Peoples. New'York: Garland. Van Oven, M„ and M. Kayser. 2009. Updated comprehensive phylogenetic tree of global human mitochondrial DNA variation. Hum. Mutat. 30:E386-E394. http://w'W'W'.phylotree.org; doi: 10.1002/humu.20921. Veltre, D. W„ and M. A. Smith. 2010. Historical overview' of archaeological research in the Aleut region of Alaska. Hum. Biol. 82(5-6):487-506. Ward, R. H., B. L. Frazier, K. Dew'-Jager et al. 1991. Extensive mitochondrial diversity within a single Amerindian tribe. Proc. Natl Acad. Sci. USA 88:8720-8724.aDNA in the Alaska Peninsula / 693 Weyer, E„ Jr. 1930. Archaeological Materials from the Village Site at Hot Springs, Port Moller, Alaska. Anthropological Papers of the American Museum of Natural History, v, 21, pt. 4, New York: American Museum of Natural History, 237-279. Woodbury, A, C. 1984, Eskimo and Aleut languages. In Handbook of North American Indians, v, 5, Arctic, D, Damas, ed, Washington, DC: Smithsonian Institution, 49-63. Workman, W, 1966. Prehistory of Port Moller, Alaska Peninsula, in light of fieldwork in 1960. Arctic AnthropoL 3:132-153. Zlojutro, M„ R. Rubicz, E. J. Devor et al. 2006. Genetic structure of the Aleuts and circumpolar populations based on mitochondrial DNA sequences: A synthesis. Am. J. Phys. AnthropoL 129:446^-64.