Professional Documents
Culture Documents
Restriction analysis of mitochondrial DNA one unique mutation, or, in the case of
(mtDNA) has demonstrated that most Na- haplogroup D, by a typical combination of
tive Americans exhibit one of four mutations
identifiable by the gain or loss of a restric-
Grant sponsor: National Institutes of Health; Grant numbers:
tion site (haplogroups A, C, and D) or, in the RR00169, RR05090; Grant sponsor: Regents of the University of
California; Grant sponsor: National Science Foundation; Grant
case of haplogroup B, by a 9-bp deletion number: SBR-9630926.
(Schurr et al., 1990). Subsequent sequenc- *Correspondence to: David Glenn Smith, Department of An-
thropology, University of California, Davis, CA 95616. E-mail:
ing analysis revealed that each of the four dgsmith@ucdavis.edu
haplogroups is also characterized by at least Received 25 November 1998; accepted 24 June 1999.
mutations, within the mtDNA (noncoding) ably unmixed) Native Americans (Bailliet et
control region (CR) (Torroni et al., 1993). al., 1994) and in prehistoric individuals
These four haplogroups have been argued to (Stone and Stoneking, 1993; 1998; Ribeiro-
represent founding lineages because all four dos-Santos et al., 1996; Kaestle, 1997), sug-
(1) are also found in Asia, presumed to be the gesting that founding haplogroups other
homeland of all Native Americans, (2) oc- than A, B, C, and D were once present in the
cupy central nodes in a phylogeny of New New World, whether or not they have sur-
World mtDNA haplotypes, and (3) are wide- vived.
spread throughout North, Central and South Bailliet et al. (1994) argued that the Hae
America. III site at np 16517 (cited above) character-
While widespread, the geographic distribu- izes two subtypes of each of three of the four
tion of the four haplogroups is markedly major haplogroups (A, C, and D), each sub-
nonrandom (Lorenz and Smith, 1996). For type representing two different founding
example, haplogroup A is extremely common lineages associated with each of the three
among Eskimo/Aleut and Northern Athapas- haplogroups. Alternatively, Torroni et al.
kan tribes, but extremely rare in non- (1993) have argued that the np 16517 site is
Athapaskan speakers of the Southwest hypervariable and, therefore, that the emer-
United States. Haplogroup B is extremely gence of parallel subtypes of founding hap-
common in the American Southwest but logroups in the New World post-dates the
absent, or rare, in the Arctic, Subarctic, and settlement of the New World. Bailliet et al.
Northwest Coastal regions. Haplogroup D, (1994) rejected this argument because it
while present throughout the New World, is could not explain the apparent absence of
the least common of the four haplogroups any members of haplogroup B who also
everywhere except in certain Western tribal exhibit a Hae III site loss at np 16517.
groups, notably speakers of Penutian lan- However, Easton et al. (1996) later reported
guages in whom it represents the most that this mutation characterizes two sub-
common haplogroup. The mtDNA hap- types of haplogroup B in the Yanomami.
logroup distributions of modern Native Easton et al. (1996) reported that hap-
American groups are quite uniform among logroups they called X6 and X7 comprised
many tribal groups known to be closely about 12% of the Yanomami samples they
related (e.g., see Lorenz and Smith, 1996), studied. These are found widespread in the
suggesting that genetic drift has not influ- New World and have been reported in Asia
enced their distributions sufficiently to pre- as well (Merriwether and Ferrell, 1996).
clude their use for assessing ancestor/ Members of these haplogroups share most of
descendant relationships between living the CR mutations, but lack the restriction
populations and prehistoric skeletal mate- sites (e.g., gains of a Dde I and an Alu I
rial (e.g., Kaestle, 1997; Malhi, 1998; Parr et restriction site at np 10394 and np 10397,
al., 1996; Stone and Stoneking, 1998; Car- respectively), characteristic of haplogroups
lyle et al., 1999) in nearby regions. C and D and differ from each other by the
presence or absence of the Hae III restric-
Other mtDNA haplogroups
tion site at np 16517. However, there are
The remaining few Native Americans that numerous apparent errors in those pub-
do not exhibit one of these four haplogroups lished sequences (e.g., half of all mutations
have been termed others (Torroni et al., reported were transversions) and the CR
1993) or members of haplogroup E (Bailliet sequences of members of haplogroups X6
et al., 1994) and could represent recent and X7 always cluster with members of
non-Native American admixture and/or mu- haplogroup C and/or haplogroup D of which
tations at the diagnostic restriction site for they are probably reversions (Stone and
one of the four common haplogroups (Tor- Stoneking, 1998). Since the Hae III restric-
roni and Wallace, 1995). Haplotypes that are tion site at np 16517 appears to be a hyper-
not members of haplogroups A, B, C, or D variable site that might have experienced
and that are characterized by specific muta- numerous transitions followed by rever-
tions have been reported in modern (presum- sions, it is not clear whether or not X6 and
HAPLOGROUP X IN NATIVE AMERICANS 273
Fig. 1. Distribution of Native American tribes assigned to haplogroup X. The language group of each
tribe is coded as follows: Algonquian, ; Kiowa-Tanoan, ; Wakashan, ; Penutian, ; Northern Hokan,
; Siouan, .
X7 are founding haplogroups. That site np unrelated groups. Thus, none of the above-
16517 is hypervariable is more plausible cited candidates for an additional (i.e., fifth)
than is the random survival of exactly two Native American founding mtDNA lineage
representatives that are mutually distin- is supported by compelling evidence or satis-
guishable by the same mutation in each of fies the three criteria Torroni et al. (1993)
four (or five) already rare Asian lineages. cited as being sufficient and necessary to
Forster et al. (1996) have argued that two constitute evidence for a founding lineage.
variants of haplogroup A (differentiated by
Is there a fifth mtDNA
the presence or absence of a C = T transi-
lineage in North America?
tion at np 16111) comprise two separate
founding lineages. Given recent evidence for Bailliet et al. (1994) suggested the possibil-
a single New World migration (e.g., see ity of a fifth haplogroup, defined by a C = T
Merriwether et al., 1995; Lorenz and Smith, transition at np 16278 and the absence of
1997; Stone and Stoneking, 1998; Bonatto the mutations that characterize haplogroups
and Salzano, 1997a,b), the presence of both A, B, C or D. This haplogroup is the same as
variants in Amerind, Eskimo-Aleut and Na- that comprising the Nuu-Chah-Nulth (i.e.
Dene populations (Shields et al., 1993), Nootka) lineage cluster I in the maximum
whom many once regarded as descendants likelihood phylogeny of Ward et al. (1991;
of at least two, if not three, separate migra- figure 2). Forster et al. (1996) considered
tions to the New World, does not conflict this haplogroup to be identical to haplo-
with this hypothesis. However, the absence group X, described by Torroni et al. (1996),
of this mutation is fairly rare, is concen- which is not to be confused with haplogroups
trated in the Northwest of the continent, X6 and X7 proposed by Easton et al. (1996).
and might represent a reversion of New Haplogroup X is also characterized by Dde I
World origin followed by admixture among site losses at np 1715 and np 10394, muta-
274 D.G. SMITH ET AL.
tions that are otherwise rare in North tion has been found in two of 16 prehistoric
America (Brown et al., 1997). The hap- skeletons from Windover Pond, Florida dated
logroup occupied a nodal position in a pheno- to between 7,000 and 8,000 years BP
gram, one of three criteria cited by Torroni et (Hauswirth et al., 1994) and in an approxi-
al. (1993) as necessary and sufficient to mately 600-year-old Oneota sample from
constitute a founding haplogroup of the New the Norris Farms site (Stone and Stoneking,
World. The haplogroup is also one of at least 1998) and, therefore, is unlikely to represent
ten found in Europe (at a frequency of about post-Columbian European admixture. Re-
3%) (Torroni et al., 1996), but not East Asia, cently, the CR mutations (C = T transition
and was once, but no longer, assumed to at np 16223 and C = T transition at np
represent European admixture in the New 16278) were reported in an individual from
World (Torroni et al., 1993; Scozzari et al., the Los Vaqueros site in Central California,
1997). The C = T transition at np 16278 is dating to approximately 7,400 ybp (Eshle-
also found in Africa, where it typically accom- man, unpublished data). Although the C =
panies a Dde I site gain at np 10394 which, T transition at np 16278 is typically associ-
in Europe, is found only in members of ated with the Dde I site loss at np 1715 in
haplogroups I, J, and K. It is also common in both North America and Europe, but not
Asian haplotypes, including a minority of elsewhere, restriction analysis of the prehis-
those in haplogroup F defined by Kolman et toric samples from the four archaeological
al. (1996), but it is not known whether or not sites described above has not been done, so it
such individuals exhibit the Dde I restric- is not certain that they are members of the
tion site loss at np 1715. In contrast, the Dde same haplogroup as that reported by For-
I site loss at np 1715, but not the C = T ster et al. (1996).
transition at np 16278, is characteristic of Forster et al. (1996) reported that seven of
Europeans of haplogroup I (and a small 62 Nuu-Chah-Nulth (11%) [with haplotypes
subset of members of haplogroup H) but similar to those comprising lineage cluster I
co-occurs with the Dde I site gain at np in Figure 2 of Ward et al. (1991)] and two of
10394 (Torroni et al., 1994). In the New 40 Yakima (5%), whose haplotypes occupy
World, the C = T transition at np 16278 (but central nodes in their phylogeny, exhibited
no other CR mutations characteristic of hap- both the C = T transition at np 16278 and
logroup X) has been reported in a Pima the Dde I site loss at np 1715 and, therefore,
Indian belonging to haplogroup B, a Kraho are members of haplogroup X. Although,
tribesman (from Brazil) belonging to hap- Stone and Stoneking (1998; Fig. 2) have
logroup A (Torroni et al., 1993), three Nuu- shown that four of these seven Nuu-Chah-
Chah-Nulth belonging to haplogroup A Nulth, as well as their prehistoric Oneota
(Lorenz and Smith, 1997), and a Washo sample that was assigned to the other
member of haplogroup C (Kaestle, 1998), haplogroup, show some similarities to mem-
but is otherwise rare or absent in Native bers of the Asian haplogroup F [described by
Americans of haplogroups A, B, C, and D in Kolman et al. (1996)], the Oneota sample
whom it probably represents a mutation was not tested for the Dde I site loss at np
independent of that at np 16278 in Native 1715. The hypothesized fifth haplogroup,
Americans with the Dde I site loss at np defined by a Dde site loss at np 1715, also
1715. This same C = T transition has also represented about one-quarter of the hap-
been characterized in three of 18 pre- logroups assigned by Scozzari et al. (1997) to
Columbian Amerindian skeletons from South a group of 63 Native Americans of two
America two of which are about 4,000 years different Chippewa (known as Ojibwa in
old (Ribeiro-dos-santos et al., 1996); while Canada) tribes. Since haplogroup X appears
this mutation alone is not diagnostic of to be absent in both the Dogrib and Tanana
membership in haplogroup X, the authors, (Brown et al., 1998) its presence in the
like Bailliet et al. (1994), consider it charac- Navaho might result from their admixture
teristic of a fifth haplogroup, which they call with unrelated Pueblo groups (Brown et al.,
V, that was not among those defined by 1958; Lorenz and Smith, 1996) after their
Horai et al. (1993). In addition, this muta- arrival in the Southwest approximately half
HAPLOGROUP X IN NATIVE AMERICANS 275
a millenium ago (Brugge, 1983). Although gued. Since the five New World haplogroups
haplogroup X has not been found in East represent a minority of those surviving in
Asia, the prehistoric linkage between the C the Old World (Bailliet et al., 1994), it is of
= T transition at np 16278 and the Dde I interest to determine whether or not addi-
restriction site loss at np 1715 in Native tional founding haplogroups were once pres-
America suggests that haplogroup X might ent, but have since become extinct, in the
be a founding lineage, as Forster et al. New World. Further studies of ancient DNA,
(1996) and Brown et al. (1998) have argued, such as those conducted by Stone and
rather than representative of recent Euro- Stoneking (1993; 1998), Parr et al. (1996),
pean admixture. However, notwithstanding Kaestle (1997; 1998), Malhi (1998), and Car-
the speculation of Brown et al. (1998), that lyle et al. (1999), would then be required to
haplogroup X in North America is limited to determine whether a restricted number of
Northern Amerindian groups, it is not known haplogroups in the New World reflects a
whether or not haplogroup X is widely dis- bottleneck or a founder effect in the evolu-
tributed throughout the New World, as would tionary history of Native America. In either
be expected of a founding Native American event, knowledge of the frequency distribu-
matriline. tion of this and other less common founding
lineages will increase the utility of hap-
Is haplogroup X a fifth founding lineage? logroup distributions as signatures of an-
To assess the likelihood that haplogroup X cestor/descendant relationships between liv-
is a fifth founding lineage, the proportions of ing and prehistoric (skeletal) populations.
haplogroups previously described as other, MATERIALS AND METHODS
which themselves are known to be widely
The samples
distributed in Native America (Lorenz and
Smith, 1996), that actually belong to hap- We have previously characterized the geo-
logroup X must be determined and the distri- graphic and ethnic distributions of hap-
bution of this haplogroup in North America logroups A, B, C, and D in a sample of 829
characterized, as we have previously done Native North Americans of alleged unmixed
for the other four major founding hap- ancestry (Lorenz and Smith, 1996). In that
logroups (Lorenz and Smith, 1996). If those study, we determined that 50 of the 829
other haplogroups that prove to be mem- individuals (or 6%) lacked the restriction
bers of haplogroup X are very restricted sites, or the 9-bp deletion, that characterize
geographically, occurring, for example, only any of these four haplogroups, and were,
in the three tribes cited above, they might therefore, described as other. These 50
share, through admixture, a recent muta- samples were geographically dispersed
tion that destroyed the characteristic marker among 9 of the 12 culture areas defined by
of haplogroup A, B, C, or D, and therefore Driver and Massey (1957) and among 12 of
constitute a subtype of one of these four the 26 language groups defined by Campbell
founding lineages. If very few of the other and Mithun (1979) that were represented by
haplogroups (while widely dispersed geo- tribes included in this study. Of these 50,
graphically), are identified as X, then an sufficient sample for further study was avail-
even larger number of founding lineages able for only 25 (data for most of the other 25
might exist, and the first immigrants to the samples had been taken from published
New World might have brought a far more reports). Eleven of the 25 samples unavail-
representative sample of Asian (or Eur- able for further study were Chippewa, and
asian) lineages to the New World than has at least one each of the remainder repre-
been alleged (Wallace et al., 1985). If, on the sented six different unrelated tribal groups.
other hand, most, or all, of the members of For all but four of the ethnic groups exhibit-
this diverse group of other haplogroups ing at least one other haplogroup among
are identified as members of haplogroup X, the 25 (Cora, Maya, Navaho, and Nuu-Chah-
these five haplogroups might represent all Nulth), at least one additional sample re-
the surviving founding New World hap- mained for study. As described above, the
logroups, as Forster et al. (1996) have ar- presence of haplogroup X in one of these four
276 D.G. SMITH ET AL.
groups unrepresented in the samples avail- haplotyped by restriction analysis and se-
able for this study, the Nuu-Chah-Nulth, quence analysis of the HVSI region and gave
has already been documented (Forster et al., the same results. Therefore, the disagree-
1996). The presence of haplogroup X in ment between restriction analysis and se-
several Navaho (southern Athapaskan) quence analysis of the HVSI region in hap-
samples has been reported (haplotype AM logroup assignment of this one sample was
29 described in Torroni et al., 1992 and in not due to mistyping. We estimate a 1.7%
Brown et al., 1998) but might result from rate of inconsistency between haplogroup
extensive and recent admixture of Navaho assignments based on restriction analysis
with Pueblo tribes living in the Southwest and sequence analysis of HVSI. Since no
(Brugge, 1983) that correspondingly exhibit mistyping of the 58 samples assigned to
traits common in Athapascans but rare or haplogroups A, B or C occurred, the higher
absent in the Southwest (Brown et al., 1958; percentage of mistypes in the others cat-
Lorenz and Smith, 1996). When the 25 egory probably occurred because the au-
samples were rescreened for the characteris- thors were conservative in assigning samples
tic markers for haplogroups A, B, C and D, to each of the four common haplogroups to
seven were discovered to have been mis- avoid observer errors. Thus, only 34% of
typed; two each belonged to haplogroups A the Native American samples originally
and C and three belonged to haplogroup D, screened were not members of one of the
respectively. four major Native American matrilines.
Due to the discovery of seven mistyped Twenty (of 70) additional samples of sera
samples, we randomly selected 58 of the from Chippewa from the Lac Courte Oreilles
original 512 samples for DNA sequencing of community in Hayward, Wisconsin (or
the HVSI region (Lorenz and Smith, 1996) Southwestern Ojibwa) that were deter-
in order to estimate an error rate in identify- mined to be members of other haplogroups
ing haplogroups through restriction analy- as parts of unrelated studies, and one sam-
sis. Using the presence of diagnostic mark- ple each representing the Pomo, Blackfoot,
ers in the CR, as described in Lorenz and and Sioux (Kaestle, 1997; Malhi, 1998; Smith
Smith (1997), samples were identified as et al., 1999) were also included in the pres-
members of one of three haplogroups (A, B, ent study. A total of 30 individuals from a
or C). Additionally, members of haplogroup sample of 189 Salteaux (or Northwestern)
A were also identified by the presence of at Chippewa, not previously screened for hap-
least 2 of 3 markers in the HVSII portion of logroups A, B, C or D, were screened only for
the control region: a T = C transition at np the presence of the Dde I site losses at np
146, an A = G transition at np 153, and an 1715 and 10394 that characterize hap-
A = G transition at np 235 (Malhi, unpub- logroup X. Since samples from other Chip-
lished data). Of the 58 samples sequenced, pewa groups had been found to exhibit this
17 were assigned to haplogroup A, 20 to mutation (Scozzari et al., 1997), we studied
haplogroup B, and 21 to haplogroup C 30 additional samples from a Salteaux
through restriction analysis. All 17 samples Chippewa group merely to confirm its pres-
assigned to haplogroup A and all 20 samples ence. Finally, our screening of a sample of
assigned to haplogroup B contained HVSI or 118 additional Northern Paiute samples
HVSII CR diagnostic markers that agreed yielded no members of haplogroups other
with their original haplogroup assignments. than A, B, C, or D.
Of the 21 samples assigned to haplogroup C
Laboratory analyses
by restriction analysis, 20 contained CR
markers that agreed with their original Each of these 70 samples, whose ethnic
haplogroup assignment. The remaining sam- identities and geographical origins are sum-
ple assigned to haplogroup C had no CR marized in Table 1, was extracted using
markers for the any of the five founding methods described in Lorenz and Smith
haplogroups. The single sample that did not (1996), then amplified using the primer coor-
have any diagnostic CR markers was re- dinants L16311651 and H17931776 and
HAPLOGROUP X IN NATIVE AMERICANS 277
TABLE 1. Samples screened for haplogroup X Dde I restriction enzyme (Boehringer Mann-
in this study
heim). Portions (510 l) of the restriction
Number digests were electrophoresed on a 6% poly-
assigned to
haplogroup acrylamide gel, stained with ethidium bro-
Language Number X (% of all mide, and photographed over a UV transillu-
Ethnic group group tested samples1 )
minator, using the ISO 2000 imaging system
Southwestern Algonquian 222,7 18 (25.7) (Alpha Innotech, San Leandro, CA) to deter-
(Lac Courte
Oreilles) mine whether or not the amplified frag-
Chippewa ments had been cut by the restriction en-
Northwestern Algonquian 303 2 (6.7) zyme. Samples both of whose PCR fragments
(Salteaux)
Chippewa were not successfully digested into two seg-
Cheyenne/ Algonquian 2 2 (11.1) ments of the appropriate size, because they
Arapaho
Micmac Algonquian 2 2 (40) lacked a restriction site at both np 1715 and
Narragansett Algonquian 14 0 (0) np 10394, were regarded as members of
Blackfoot Algonquian 1 1 (100) haplogroup X.
Cherokee Iroquoian 42,6 0 (0)
Kiowa Kiowa-Tanoan 2 2 (40) Both strands of the HVSI of the mtDNA
Jemez Pueblo Kiowa-Tanoan 3 3 (8.1) CR of a representative subset (n 11) of the
Creek Muskogean 15 0 (0)
Pomo Northern Hokan 1 1 (25) other samples with the Dde I restriction
Sioux Siouan 1 1 (100) site losses at np 1715 and np 10394 were
1 Estimate of the frequency of haplogroup X in all samples tested amplified using PCR primers with coordi-
of the ethnic group indicated. nants L 1602116038 and H 1637516356
2 Two of these samples are members of haplogroup H which is of
Campbell and Mithun, 1979; Sapir, 1929). from some site in East Asia. This explana-
Since the Kiowa and Jemez are geographi- tion is consistent with the distribution of
cally separated, they might seem to be more Y-chromosome haplotype 1C from which
likely to share haplogroup X due to common Karafet et al. (1999) hypothesized a bifur-
ancestry than to admixture. The presence of cated migration from the Lake Baikal region
haplogroup X in Navaho, but not Apache or of western Siberia, one branch leading to
any Northern Athapascans that have been Europe and another to the New World.
studied (e.g., Dogrib or Tanana), is consis- Haplogroup X, with a continent-wide fre-
tent with ethnohistorical reports that at quency of about 3%, is the least common of
least ten clans claiming Anasazi ancestry, at the five predominant haplogroups of Native
least one of which (the Coyote Pass People) America, but it, like the other four hap-
originated in Jemez Pueblo, joined and inter- logroups, exhibits a nonrandom distribution
married with the Navaho (Brugge, 1983). in North America characterized by very high
However, one of the Kiowa haplotypes is frequencies in some modern (e.g., Chippewa,
identical to one of the Algonquian haplo- Micmac, Kiowa) populations. As such, its
types, and the Navaho haplotype is more distribution among modern Native Ameri-
similar to the Algonquian haplotype than to can groups might provide clues to common
any of the Jemez haplotypes. Further stud- ancestral relationships among modern tribal
ies are needed to identify the source of the groups, and comparisons with ancient
haplogroup X haplotypes in the American mtDNA might be useful for assessing ances-
Southwest. The apparent antiquity of hap- tor/descendant relationships in North
logroup X in such geographically and/or America.
linguistically dispersed tribal groups is con- In addition to confirming its legitimacy as
sistent with its status as a founding Native a founding haplogroup, the widespread pres-
American matriline. ence of haplogroup X in prehistoric samples
Curiously, haplogroup X is found in mod- will be particularly useful for assessing the
ern populations of Europe and Southwest antiquity of Algonquian-speaking peoples in
Asia but not in those of Central Siberia, the Subarctic/Great Lakes region and evalu-
which is now regarded as the source for the ating the hypothesis that the Proto-Algonqui-
peopling of the New World (e.g., see Karafet ans originated on the Columbia Plateau,
et al., 1999; Kolman et al., 1996; Merri- then migrated eastward to their present
wether et al., 1996; Santos et al., 1999). homes between three and four millennia ago
Several CR mutations that were found to be (Denny, 1991; Malhi, 1998). The distribution
characteristic of haplogroup X in this study of haplogroup X in prehistoric skeletal popu-
(e.g., the T = C transition at np 16189 and lations in the Southwest United States asso-
C = T transitions at np 16223 and np 16278) ciated with Anasazi, Hohokam, Fremont, and
are found in some members of the Asian Hakataya cultural manifestations might
haplogroup F, but they are infrequent, not help sort out the relationships between each
found together, or with the relatively rare of them and Tanoan-speaking Pueblo groups
A = C transversion at 16183, and lack the living there today. Two samples each from
Dde I site loss at np 1715 (e.g., see Mongo- the prehistoric Anasazi and Fremont popula-
lian sequences in Table 2). Thus, it is un- tions were found not to exhibit the restric-
likely that the haplogroups F and X are tion sites characteristic of haplogroup A, B,
closely related. Alternatively, one of the C, or D (Parr et al., 1996; Carlyle, 1999) and
founding haplogroups of Native America might actually be haplogroup X, which we
might be of European origin, in which case have shown to persist in Pueblo groups in
the provocative issues of its time and route that area. If samples from the prehistoric
of entry into the New World and its incorpo- Windover, Florida site (Doran et al., 1985)
ration into Asian-derived tribal groups exhibiting the C = T transition at np 16278
emerge. A more likely scenario is that hap- but lacking the restriction sites characteris-
logroup X was once found in both Europe tic of any of the four common New World
and Asia but became extinct in Asia after haplogroups (Hauswirth et al., 1994) are
Native Americans peopled the New World members of haplogroup X, its absence in
282 D.G. SMITH ET AL.
Smith DG, Lorenz J, Rolfs BK, Bettinger RL. 1999. founding Native American mtDNAs. Am J Hum Genet
Implications of the distribution of Albumin Naskapi 53:563590.
and Albumin Mexico for New World prehistory. Am J Torroni A, Lott MT, Cabell MF, Chen Y-S, Lavergne L,
Phys Anthropol (in press). Wallace DC. 1994. mtDNA and the origin of Cauca-
Stone AC, Stoneking M. 1993. Ancient DNA from a sians: Identification of ancient Caucasian-specific hap-
pre-Columbian Amerindian population. Am J Phys logroups, one of which is prone to a recurrent somatic
Anthropol 92:463471. duplication in the D-Loop region. Am J Hum Genet
Stone AC, Stoneking M. 1998. mtDNA analysis of a 55:760776.
prehistoric Oneota population: Implications for the Torroni A, Huoponen K, Francalacci P, Petrozzi M,
peopling of the New World. Am J Hum Genet 62:1153 Morelli L, Scozzari R, Obinu D, Savontaus M-L,
1170. Wallace DC. 1996. Classification of European mtD-
Torroni A, Wallace DC. 1995. Mitochondrial DNA hap- NAs from an analysis of three European populations.
logroups in Native Americans. Am J Hum Genet Genetics 144:18351850.
56:12341236. Vigilant L, Stoneking M, Harpending H, Hawkes K,
Torroni A, Schurr TG, Yang C-C, Szathmary EJE, Wil- Wilson AC. 1991. African populations and the evolu-
liams RC, Schanfield MS, Troup GA, Knowler WC, tion of human mitochondrial DNA. Science 253:1503
Lawrence DN, Weiss KM, Wallace DC. 1992. Native 1507.
American mitochondrial DNA analysis indicates that Wallace DC, Garrison K, Knowler WC. 1985. Dramatic
the Amerind and the Nadene populations were founded founder effects in Amerindian mitochondrial DNAs.
by two independent migrations. Genetics 130:153 Am J Phys Anthropol 68:149155.
162. Ward RH, Frazier BL, Dew-Jager K, Paabo S. 1991.
Torroni A, Schurr TG, Cabell MF, Brown MD, Neel JV, Extensive mitochondrial diversity within a single
Larsen M, Smith DG, Vullo CM, Wallace DC. 1993. Amerindian tribe. Proc Natl Acad Sci USA 88:8720
Asian affinities and continental radiation of the four 8724.