Professional Documents
Culture Documents
ABSTRACT The Evenks and Evens, who speak closely bors rather than with linguistically related peoples. In
related languages belonging to the Northern Tungusic this study, we assess the correlation between linguistic
branch of the Tungusic family, are nomadic reindeer and genetic relationship in three different subgroups of
herders and hunters. They are spread over an immense Evenks and Evens, respectively, via mtDNA and Y-chro-
territory in northeastern Siberia, and consequently dif- mosomal analyses. The results show that there is some
ferent subgroups are in contact with diverse peoples evidence of a common origin based on shared mtDNA
speaking Samoyedic, Turkic, Mongolic, Chukotka-Kam- lineages and relatively similar Y-haplogroup frequencies
chatkan, and Yukaghir languages. Nevertheless, the amongst most of the Evenk and Even subgroups.
languages and culture of the Evenks and Evens are sim- However, there is little sharing of Y-chromosomal STR
ilar enough for them to have been classified as a single haplotypes, indicating that males within Evenk and
ethnic group in the past. This linguistic and cultural Even subgroups have remained relatively isolated. There
similarity indicates that they may have spread over is further evidence of some female admixture in different
their current area of habitation relatively recently, and Even subgroups with their respective geographic neigh-
thus may be closely related genetically. On the other bors. However, the Tungusic groups, and especially the
hand, the great distances that separate individual Evenks, show signs of genetic drift, making inferences
groups of Evens and Evenks from each other might about their prehistory difficult. Am J Phys Anthropol
have led to preferential mating with geographic neigh- 133:1013–1027, 2007. V 2007 Wiley-Liss, Inc.
C
The peoples of Central and Northeastern Siberia fering mainly in phonological and lexical features (Kon-
belong to different language families and practice differ- stantinova, 1964). The Even dialects, on the other hand,
ent modes of subsistence: in southern Siberia, they are are judged to be more divergent, especially the western
mainly pastoralists speaking Turkic or Mongolic lan- dialects. These are not understood by speakers of east-
guages, while the populations to the north are mainly ern dialects, nor do Western Evens understand Eastern
fully nomadic reindeer herders and hunters, such as the Evens. The central Even dialects, on the other hand,
Tungusic-speaking Evenks and Evens and the Yuka- are mutually comprehensible with the eastern dialects
ghirs, as well as a Turkic-speaking cattle- and horse- (Novikova, 1960). Both groups practice hunting and
breeding enclave, the Yakuts. Amongst all the Siberian reindeer herding, harnessing their reindeer to sleds in
peoples, the Evenks and Evens stand out through their
widely dispersed area of settlement: although according Grant sponsor: Max Planck Society; Grant sponsor: Wenner-Gren
to the 2002 census there were only 35,000 Evenks and Foundation for Anthropological Research; Grant number: 6828;
19,000 Evens in the Russian Federation (Federal’naja Grant sponsor: Russian Foundation for Basic Research (RFBR);
Sluzhba Gosudarstvennoj Statistiki, 2004), the Evenks Grant number: 03-06-96033p2003ap a_a.
are spread from the Taymyr Peninsula and Yenissey
*Correspondence to: Brigitte Pakendorf, Max Planck Institute for
river in the northwest to the Amur river and Sea of
Evolutionary Anthropology, Deutscher Platz 6, D-04103 Leipzig,
Okhotsk in the southeast (a distance of *3,200 km), and Germany. E-mail: pakendorf@eva.mpg.de
the Evens are settled in an area from the upper reaches
of the Yana river in the west to the Sea of Okhotsk
Present address of Innokentij N. Novgorodov: Department of For-
and Kamchatka in the east, a distance of *1,900 km eign Languages, Faculty of Information Technology, The Yakut State
(Evstigneev, 2003). The closely related Evenk and Even Engineering and Technical Institute, Yakutsk, Republic Sakha
languages both belong to the Northern Tungusic branch (Yakutia), Russian Federation.
of the family (Comrie, 1981; Atknine, 1997); each is split
into a large number of dialects, reflecting the wide Received 28 September 2006; accepted 24 January 2007
dispersal of the population (Novikova, 1960; Nedjalkov,
1997). Notwithstanding the enormous geographical dis- DOI 10.1002/ajpa.20590
tances separating the Evenk dialects, they are claimed Published online 9 May 2007 in Wiley InterScience
to be close enough to enable mutual comprehension, dif- (www.interscience.wiley.com).
C 2007
V WILEY-LISS, INC.
1014 B. PAKENDORF ET AL.
winter and riding them in summer. Because of their of how closely related different Tungusic-speaking
close linguistic and cultural similarity, in the past, the groups are by studying mtDNA and Y-chromosomal vari-
two groups were often classified as one ethnic group ation in three subgroups of Evenks from the Evenk
called Tungus (Evstigneev, 2003). There are two differing National okrug, southern Yakutia, and the northern
hypotheses on the origins of the Northern Tungusic Amur oblast (both okrug and oblast are administrative
groups. The first, predominantly supported by Okladni- divisions in the Russian Federation) and in three sub-
kov (cf. references in Tugolukov, 1980) and Vasilevich groups of Evens from central and eastern Yakutia and
(1969), holds that the Evenks and Evens are the de- the Magadan oblast, and comparing them with neighbor-
scendants of neolithic populations that were settled in ing groups: Yakut-speaking Evenks from northwestern
South Siberia *3,000 to 4,000 years ago. According to Yakutia, Yakuts, Yukaghirs, Koryaks, and Tuvans. In
Okladnikov (cf. references in Tugolukov, 1980), these ne- particular, we consider to what extent Evenks and Evens
olithic ancestors of the Evenks and Evens were reindeer- intermarried with their geographic neighbors, rather
herders; the modern-day Tungusic populations of the than choosing a mate from a more distantly located, but
Amur region and Manchuria lost their reindeer-herding linguistically and culturally closely affiliated group, and
culture after migrating to the southeast. In the middle of whether there are differences with respect to female and
the first millenium AD, the arrival of Turkic groups on male admixture.
the shores of Lake Baykal is assumed to have split the
ancestors of the northern Tungus (Evenks and Evens) MATERIALS AND METHODS
into a western and eastern group; this led to their
migration to the north and initiated the formation of the Samples and DNA extraction
Evenks and Evens as separate peoples without contact In December 2003, 46 cheek swab samples were
with the Tungusic-speaking groups from the Lower collected from healthy Even men in the Eveno-Bytantay
Amur (Vasilevich, 1969). The second hypothesis takes district and the village of Topolinoe, Tompo district,
the opposite view: according to this, the Evenks and Republic Sakha (Yakutia). These were augmented by
Evens are believed to have originally been horse-herders four Even maternal lineages from the Yukaghir village
settled in Manchuria, possibly migrating to the Middle of Andryushkino (Nizhnekolyma district, Republic Sakha
Amur or Transbaykal region, where they adopted (Yakutia)), which were collected in 1995. The Even sam-
domesticated reindeer, under pressure of immigrating ples were divided into Western and Central Even sub-
Dagurs (Tugolukov, 1980; Janhunen, 1996). Their migra- groups on the basis of dialectal distinctions (Novikova,
tion further to the north is assumed to have taken place 1960); in five cases, the maternal and paternal lineages
relatively recently, in the 12th and early 13th century came from different subgroups, so that the number of
AD, while their spread over the vast area they inhabit Western and Central Evens differed for the mtDNA
today is believed to have been initiated by the Yakuts, and Y-chromosome analyses (cf. Tables 1 and 5). A fur-
who migrated to the Middle Lena from near Lake ther sample of 65 published Even sequences from the
Baykal in the 13th and 14th century (Tugolukov, 1980; Magadan oblast, who belong to the Eastern dialect
Janhunen, 1996). The Russian colonization of Siberia, group, was taken from the literature (Derenko, 1997).
which began in the 17th century, led to large-scale popu- For the Y-chromosomal analyses, we included published
lation movements in the area, and it is during this time Y-chromosomal SNP data from a mixed sample of 31
that some Evenk and Even groups spread further to Evens from the Kolyma River (Yakutia) and the Maga-
the northwest, southeast, and northeast, partly under dan oblast (Karafet et al., 2002, Karafet, personal com-
the pressure of the spreading Yakuts (Dolgikh, 1960; munication). Evens in both areas speak dialects belong-
Vasilevich, 1969). The latest migration was that of the ing to the Eastern dialectal group (Novikova, 1960). We
Evens to the Kamchatka Peninsula in the middle of the were thus able to compare three subgroups of Evens
19th century (Spevakovsky, 1994). belonging to the three major Even dialectal divisions
If the hypothesis of Vasilevich (1969) is correct, then (Western, Central, and Eastern) in both mtDNA and Y-
Evenks and Evens have constituted separate ethnic chromosomal analyses.
groups for *1,500 years, whereas the hypothesis pro- Twenty-four Evenk DNA samples collected in 2001 in
posed by Tugolukov (1980) and Janhunen (1996) suggests Iengra village in the administrative territory of Neryun-
a much more recent shared ancestry of the Northern gri, Republic Sakha (Yakutia) were obtained from the
Tungusic groups. The two hypotheses predict differing DNA sample collection of the Institute of Health,
degrees of genetic relatedness of Evenks and Evens, Yakutsk. These were augmented by one Evenk maternal
with Vasilevich’s hypothesis predicting a deeper split lineage from the Megino-Kangalas district of the Repub-
between the groups, while the recent spread hypothesis lic Sakha (Yakutia) and two Evenk paternal lineages
predicts genetic unity within and between these ethnic from the Olenëk and Verkhoyansk districts, respectively.
groups. However, since they are spread over such a vast Since the majority of these Evenk samples came from
territory, it is possible that they have admixed with their women, only nine unrelated Iengra Evenks were avail-
immediate neighbors rather than finding ethnically and able for Y-chromosomal analyses. In addition, mtDNA
linguistically related marriage partners in a distant com- and Y-chromosomal data for a sample of 40 Evenks from
munity. Although Evenk and Even samples have been several villages in the Evenk National okrug (mainly
included in previous broad areal studies of genetic diver- along the Stony Tunguska River) were included (Paken-
sity and linkage disequilibrium (Posukh et al., 1990; Tor- dorf et al., 2006). These are labeled as Stony Tunguska
roni et al., 1993; McComb et al., 1995; Kaessmann et al., Evenks (abbreviated as STE) in this article. Published
2002; Karafet et al., 2002; Starikovskaya et al., 2005; mtDNA sequence data for 47 Evenks from the northern
Derenko et al., 2006), the correlation of genetic affilia- Amur oblast (Kaessmann et al., 2002) were included in
tion with linguistic relationship within the Northern the mtDNA comparisons, and published Y-chromosomal
Tungusic groups has not yet been the focus of a molecu- SNP data from 78 ‘‘Eastern Evenks’’ (Karafet et al.,
lar anthropological study. Here, we address the question 2002) were included in the Y-chromosomal analyses.
STE, Stony Tunguska Evenks; Iengra, Iengra Evenks; Nyukzha, Nyukzha Evenks; YSE, Yakut-speaking Evenks. Data for STE, YSE, Yakuts, Yukaghirs and Tuvans from Paken-
dorf et al. (2006); for Koryaks from Schurr et al. (1999).
a
Sequences with mutations at 16189-16223-16266-16362 were defined as belonging to subhaplogroup D5a; sequences lacking the mutation at 16266 were assigned to subha-
plogroup D5.
TABLE 3. Shared mtDNA haplotypes and pairwise Fst values between Tungusic-speaking groups and their neighbors
Iengra Nyukzha
W. Evens C. Evens E. Evens STE Evenks Evenks YSE Yakuts Yukaghirs Koryaks Tuvans
W. Evens 15 0.02 0.06 0.08 0.07 0.05 0.00 0.00 0.03 0.07 0.02
C. Evens 6 13 0.05 0.02 0.11 0.01 0.04 0.03 0.02 0.06 0.00
E. Evens 6 7 29 0.13 0.12 0.07 0.07 0.06 0.05 0.03 0.06
STE 4 5 3 16 0.21 0.06 0.12 0.08 0.04 0.12 0.02
Iengra Evenks 5 3 3 3 12 0.08 0.01 0.03 0.11 0.15 0.12
Nyukzha Evenks 4 4 4 4 8 16 0.04 0.04 0.04 0.10 0.03
YSE 5 6 4 5 4 5 18 0.00 0.04 0.08 0.05
Yakuts 13 8 11 9 7 8 12 62 0.03 0.06 0.03
Yukaghirs 5 6 4 4 4 5 6 8 13 0.04 0.01
Koryaks 3 2 14 1 2 2 1 3 6 37 0.06
Tuvans 4 6 5 6 4 5 5 12 5 3 34
Diagonal values (indicated by bold values) show the number of haplotypes within populations.
Values above the diagonal values are the pairwise Fst values between populations; bold italic values are values that are not signifi-
cantly different (5% level) without Bonferroni correction.
Values below the diagonal values are the number of shared haplotypes between populations.
Linguistic affiliation: Turkic ¼ Yakuts, YSE, Tuvans, Tofa, Todzha, Soyots, Khakas, Altai, Kazakh, Kirghiz; Tungusic ¼ STE, Iengra,
Nyukzha, C. Evens, W. Evens, E. Evens (subgroups ¼ Evens vs. Evenks); Mongolic ¼ Buryats, Mongols; isolates ¼ (1) Koryaks. (2) Yukaghirs.
Geography: N/NE ¼ Yakuts, YSE, Yukaghirs, STE, C. Evens, W. Evens; E ¼ Koryaks, E. Evens; SE ¼ Iengra, Nyukzha; S ¼ Bury-
ats, Mongols, Soyots; SW ¼ Tuvans, Khakas, Tofa, Todzha, Altai; Central Asia ¼ Kazakh, Kirghiz.
MDS Tungusic: Group 1 ¼ W. Evens, Yakuts, YSE; Group 2 ¼ Koryaks, E. Evens; Group 3 ¼ Yukaghirs, C. Evens, Tuvans; Group 4
¼ STE; Group 5 ¼ Iengra; Group 6 ¼ Nyukzha.
Evens and Evenks grouped: Group 1 ¼ W. Evens, E. Evens, C. Evens; Group 2 ¼ STE, Iengra, Nyukzha; Group 3 ¼ Yakuts, YSE;
Group 4 ¼ Koryaks; Group 5 ¼ Yukaghirs.
a
P < 0.05.
b
P < 0.01.
c
P > 0.2.
Since these samples all came from villages along the They were unrelated up to the grandparental generation
Nyukzha River or nearby (Wiebe, personal communica- as far as could be established, and were asked to name
tion, Karafet, personal communication), they are labeled the birthplace and ethnicity of their parents and grand-
as Nyukzha Evenks in this article. We were thus able to parents to exclude known admixture from other ethnic-
compare three geographical subgroups of Evenks belong- ities.
ing to two different dialectal groups: southern (Stony DNA was extracted from the blood samples collected
Tunguska Evenks) and eastern (Iengra and Nyukzha from the Iengra Evenks following a standard phenol-chlo-
Evenks). roform protocol (Maniatis et al., 1982), while DNA from
All donors were healthy individuals who voluntarily the cheek swab samples collected from the Evens was
gave their sample after informed consent was obtained. extracted using a salting-out method (Miller et al., 1988).
Referencesd
ber of populations from Siberia and Central Asia, the
1
1
2
3
1
2
3
3
3
2
3
data for which were taken from the literature: for both
1 ¼ This study; 2 ¼ Karafet et al. (2002), the Eastern Evens are there called ‘‘Evens,’’ the Nyukzha Evenks are there called ‘‘Eastern Evenks’’; 3 ¼ Pakendorf et al. (2006).
the mtDNA analyses and Y-chromosomal SNP and STR
analyses, they were compared to northeastern Siberian
Yakuts, Yakut-speaking Evenks, Yukaghirs, and South
4 (12.9)
Other
5 (6.4)
1 (8.3)
Siberian Tuvans (Pakendorf et al., 2006). In addition, for
the mtDNA analyses, they were compared with north-
eastern Siberian Koryaks (Schurr et al., 1999), with
12 (21.8)
R-M17
NA
NA
South Siberian Turkic Khakas, Tofalar, Todzhins, Soyots,
and Altaians (Derenko et al., 2003), as well as with
TABLE 5. Y-chromosomal haplogroup frequencies in several Siberian populations (absolute numbers, % in parentheses)
Karafet et al. (2002) analyzed M178 in addition TatC and gave only numbers for M178, however, all individuals with TatC also had M178.
pared with the populations analyzed by Karafet et al.
(2002), since these offered the highest resolution. Figure 1
(30.8)
(16.7)
(16.4)
Q-P36
(0.5)
1 (8.3)
1 (1.8)
Karafet et al. (2002) analyzed P14, not M89, this appears to be a parallel mutation to M89 (Jobling and Tyler-Smith, 2003).
mtDNA analyses
Both strands of the mtDNA HVR1 were sequenced
9 (37.5)
11 (27.5)
(9.1) 15 (27.3)
(72.7) 2 (6.1)
1 (0.5)
N-P43
4 (12.9)
(22.2)
(20.5)
(30.8)
(33.3)
N-TatCc
(94)
4
4
5
hvrbase.org/).
F-M89b J-M172
Y-chromosomal analyses
1 (1.3)
1 (0.5)
1 (1.3)
1 (7.7)
19 (61.3)
42 (53.8)
4 (12.1)
2 (15.4)
1 (0.5)
5 (9.1)
28 (70)
1 (7.7)
NA
NA
12 4 (33.3)
C-M217
24 1 (4.2)
184 3 (1.6)
13 1 (7.7)
55 3 (5.5)
40
33
N
Statistical analyses
Yukaghirs
Koryaks
Tuvans
Fig. 1. Map showing the approximate location of the Tungusic-speaking populations and their closest neighbors.
Y-chromosomal SNPs and STRs, as well as pairwise and D in the Tungusic-speaking groups and their neigh-
Fst values between populations and analyses of molecu- bors were constructed using the program Network ver.
lar variance (AMOVA) from the mtDNA sequence data 4.1 (www.fluxus-engineering.com) with the following
were calculated using the program Arlequin ver. 2.000 sites downweighted to 2: 16093, 16129, 16172, 16187,
(Schneider et al., 2000). The significance of Fst values 16189, 16209, 16223, 16234, 16256, 16278, 16290, 16291,
was tested with 10,000 permutations. The program pack- 16293, 16294, 16304, 16311, 16320, 16355, 16362. In the
age Statistica ver. 7.1 (StatSoft, Inc.) was used to per- network consisting of haplogroup D sequences, np 16093
form multidimensional scaling analyses (MDS) of the was downweighted to 1, to reduce excessive reticulation.
pairwise Fst values for both the mtDNA and Y-chromo- The Network program was also used to assess the qual-
somal data. Mantel tests to assess the correlation ity of the sequences generated for this study as described
between mtDNA and Y-chromosomal variation, as well by Bandelt et al. (2002). Furthermore, MJ-networks
as between mtDNA variation and geography, and Y-chro- were constructed from Y-STR haplotypes with individual
mosomal variation and geography, were performed in STRs weighted relative to their mutation rate (Kayser
Arlequin ver. 3.01 (Excoffier et al., 2005). For this, we et al., 2000) as described elsewhere (Kasperaviciute
analyzed the correlation between Fst matrices calculated et al., 2004). For the network consisting of Y-haplogroup
from mtDNA sequences and Y-SNP frequency (in 11 pop- C3 haplotypes, DYS389II was downweighted to 1 and
ulations), from mtDNA sequences and Y-STRs (in eight DYS385II was downweighted to 2, to resolve excess
populations), from mtDNA haplogroup frequency and Y- reticulation.
SNP frequency (in 11 populations), from mtDNA sequen-
ces and mtDNA haplogroup frequency and geographical
great circle distances, respectively, as well as from Y- RESULTS
SNP frequency and Y-STRs and geographical distances, mtDNA analyses
respectively. A SAMOVA (Spatial Analysis of Molecular
Variance; Dupanloup et al., 2002) analysis with the HVR The accuracy of the sequences was evaluated following
1 sequence data was performed with the program the method of Bandelt et al. (2002), in which the sites
SAMOVA 1.0 (http://web.unife.it/progetti/genetica/Isabelle/ they determined as ‘‘speedy’’ were removed from the
samova.html). Median-joining networks (Bandelt et al., data and a RM-network (Bandelt et al., 1995) was con-
1999) of HVR1 sequences belonging to haplogroups C structed from the remaining slowly-evolving sites and
Italic values are values that are not significant at the 5% level without Bonferroni correction.
nor geographic proximity are good predictors of genetic C-M86 also had 13 and 14 repeats at this locus. Duplica-
relationships (Table 7). The analyses further confirm tions at DYS393 have been found in four individuals
the close affinities of the Western Evens with Yakut- from the Czech Republic, Turkey, and Poland, however
speaking Evenks and Yakuts, and of the Dolgans (a with very different haplotypes than the ones found in
group who speak a language very closely related to the Evens (http://www.yhrd.org/index.html, release 20,
Yakut) with the remaining Evens and Evenks: in this January 10, 2007). Also, a duplication of DYS19 involv-
grouping, 45% of the variation is due to differences ing alleles 16 and 17 was found in four Tuvan individu-
between groups, and only 5% of the variation is due to als in the sample analyzed in Pakendorf et al. (2006).
differences between the populations included in the indi- These four Tuvans belong to haplogroup C-M86; this
vidual groups. In contrast, when these populations are same duplication has been reported in Kalmyks (Nasidze
grouped strictly according to their linguistic affiliation, et al., 2005).
approximately the same amount of the variation is The Western Evens show a very low haplogroup diver-
accounted for by differences between populations within sity, much lower than that found in either the other
the groups as by differences between the groups. Even subgroups or the Evenks, and almost as low as
Typing nine Y-chromosomal STRs in the Western and that of Yakuts (Table 8). In their Y-STR diversity, how-
Central Even subgroups as well as in the Stony Tung- ever, they do not differ significantly from the Central
uska and Iengra Evenks revealed a duplication of the Evens. The Stony Tunguska Evenks, too, show reduced
DYS393 locus in four Evens: two Western Evens belong- haplogroup diversity and MPD values, as well as signifi-
ing to haplogroup N-TatC had 13 and 14 repeats at this cantly lower Y-STR haplotype diversity than that found
locus, and two Central Evens belonging to haplogroup in any of the other populations compared (Table 8).
As can be seen in Table 9, there is not much sharing
of Y-STR haplotypes amongst the Siberian populations.
As shown by pairwise Rst values, the Western Evens are
closer to the Yakut-speaking Evenks from northwestern
Yakutia than they are to the Central Evens, while the
Central Evens are closer to the Stony Tunguska and
Iengra Evenks than they are to the Western Evens. The
Iengra Evenks are not significantly different from the
Central Evens, Stony Tunguska Evenks, and Yukaghirs;
however, this may be an artefact of the small sample size.
A MJ-network of Y-STRs within haplogroup N-TatC
(Fig. 7) demonstrates that the Western Evens belonging
to this haplogroup are divided into two groups. Eight of
the Western Evens share their STR haplotypes with
Yakuts and Yakut-speaking Evenks, and two have a sin-
gle haplotype that is part of the Yakut-dominated half of
the network. The two individuals with the 13/14 duplica-
tion in DYS393 are indicated by asterisks; the location
of their haplotypes (identical to the two most frequent
haplotypes in Yakuts) in the figure comes from coding
Fig. 6. MDS plot based on pairwise Fst values from Y-chro- them as having 14 repeats (which we reconstruct as hav-
mosomal SNPs between Tungusic-speaking populations and ing been the ‘‘ancestral’’ state before the duplication);
their immediate neighbors. with 13 repeats at DYS393, they would be identical to
Linguistic affiliation: Uralic ¼ Komi, Khanty, Selkup, Forest Nenets, Tundra Nenets, Nganasan; Turkic ¼ Yakuts, YSE, Tuvans,
Altai, Kazakh, Kirghiz, Dolgans, Uyghurs, Uzbeks; Tungusic ¼ STE, Iengra, Nyukzha, Chinese Evenks, W Evenks, Oroqen, Man-
chu, C. Evens, W. Evens, E. Evens (subgroups ¼ Evens vs. Evenks incl. Oroqen; Manchu excluded); Mongolic ¼ Buryats, Mongols;
isolates ¼ (1) Koryaks, (2) Yukaghirs, (3) Eskimos, (4) N. Han, (5) Russians, (6) Kets.
Geography: NW ¼ Komi, Khanty, Selkup, ForestNenets, TundraNenets, Ket; N/NE ¼ Nganasan, W. Evenks, Yakuts, YSE, Yukaghirs,
STE, C. Even, W. Even, Dolgans; E ¼ Koryaks, E. Evens, Eskimos; SE ¼ Iengra, Nyukzha, Buryats, Mongols, Chinese Evenks, Oroqen,
Manchu, N Han; SW ¼ Tuvans, Altai; Central Asia ¼ Kazakh, Kirghiz, Uyghur, Uzbeks; W ¼ Russians.
Tungusic vs. Yakut-speaking: Tungusic ¼ W. Evenks, STE, Iengra, Nyukzha, Chinese Evenks, E. Evens, W. Evens, C. Evens;
Yakut-speaking ¼ Yakuts, Dolgans, YSE.
W. Evens with Yakuts: Tungusic ¼ W. Evenks, STE, Iengra, Nyukzha, Chinese Evenks, E. Evens, C. Evens, Dolgans; Yakut-speak-
ing ¼ Yakuts, YSE, W. Evens.
a
P < 0.01.
b
P > 0.05.
c
P < 0.05.
NA ¼ not applicable.
a
The sample size was 184 for the haplogroup diversity calculations and 183 for the STR diversity calculations.
TABLE 9. Shared number of Y-STR haplotypes (below diagonal) and pairwise Rst values (above diagonal) between
Tungusic-speaking groups and their neighbors
Western Evens* Central Evens STE Iengra Evenks YSE Yakuts Yukaghirs Tuvans
Western Evens 8 0.18 0.47 0.22 0.08 0.27 0.14 0.22
Central Evens 2 8 0.08 0.02 0.29 0.65 0.14 0.13
STE 1 1 7 0.09 0.55 0.78 0.32 0.23
Iengra Evenks 1 0 2 6 0.28 0.68 0.06 0.09
YSE 1 0 1 3 13 0.13 0.17 0.22
Yakuts 3 3 1 2 5 44 0.54 0.53
Yukaghirs 1 1 0 1 0 1 7 0.04
Tuvans 1 1 1 0 0 3 0 29
the two haplotypes marked by arrows in Figure 7. The et al., 2000), this event would have taken place 4,630
remaining ten Western Evens have an identical haplo- (2,780–7,350) years ago.
type (Type 1 in Fig. 7) that is in the part of the network We performed Mantel tests to evaluate the correlation
that includes Tuvans and Yukaghirs. These haplotypes between mtDNA and Y-chromosomal variation as well as
are related to haplotypes based on 6 Y-STRs from differ- between mtDNA variation and geography and Y-chromo-
ent Finno-Ugric populations, Russians, and Buryats somal variation and geography. None of the analyses
(data not shown; taken from Zerjal et al., 1997; Lahermo showed a significant correlation (data not shown).
et al., 1999). This indicates a dual source for haplogroup
N-TatC in the Western Evens, of which about half origi- DISCUSSION
nated through admixture with Yakuts. The two Central
Evens belonging to haplogroup N-TatC are to be found By analyzing both M48 and M86, which are generally
in the Yakut half of the network, indicating an admixed assumed to be equivalent mutations defining the C3c
origin for these haplotypes as well. This also holds for haplogroup, we found that the mutations could be sepa-
the two Iengra Evenks with Tat C: their haplotypes are rated and that the M86 mutation occurred on the back-
the most common ones in Yakuts, making an admixed ground of derived M48 chromosomes. Since both M48
source very likely (types marked by an asterisk in Fig. 7). and M86 are generally not typed in the same study,
A MJ-network of Y-STRs on the background of hap- there is a lack of data on the distribution of the C-M48*
logroup C3 shows that the Evens (mainly Central Evens) subhaplogroup, making any inferences about the history
and Evenks have separate C-M86 haplotypes that differ of this subhaplogroup speculative. Judging from the
from each other by at least two mutational steps (Fig. 8). number of mutations that separate the C-M48* sub-
The two Central Evens with the 13/14 duplication at haplogroup from the C-M86 haplogroup, the separation
DYS393 are shown with 13 repeats (arrows in Fig. 8), time falls in the Neolithic, around 4,600 years ago. The
which is the most plausible ‘‘ancestral’’ state for the Evenks and Yakuts are widely assumed to have migrated
duplication. With 14 repeats, they would each be at a to Yakutia relatively late, in the first half of the second
distance of one mutational step and constitute a separate millennium AD (Tugolukov, 1980; Gogolev, 1993; Janhu-
haplotype (data not shown). Furthermore, the four indi- nen, 1996), but the Yukaghirs are believed to be the de-
viduals with the derived state at M48 and the ancestral scendants of the Neolithic inhabitants of Central and
state at M86 (C-M48* in Fig. 8) are separated from the Northeastern Siberia (Gurvich and Simchenko, 1980;
C-M86 haplotypes by at least five mutations, indicating Gurvich, 1994). It is therefore possible that the under-
a long time period since the M86 mutation arose on the ived C-M48* subhaplogroup was carried to Northeastern
background of M48. Assuming an average mutation rate Siberia by the ancestors of the Yukaghirs, where it was
of 3 3 103 (95% CI 1.89 3 103 to 4.94 3 103) (Kayser passed into the Iengra Evenks and Yakut-speaking