Research Paper
1237
Melanesian origin of Polynesian Y chromosomes
Manfred Kayser*, Silke Brauer*, Gunter Weiss*, Peter A. Underhill†,
Lutz Roewer‡, Wulf Schiefenhövel§ and Mark Stoneking*
Background: Two competing hypotheses for the origins of Polynesians are the
‘express-train’ model, which supposes a recent and rapid expansion of
Polynesian ancestors from Asia/Taiwan via coastal and island Melanesia, and
the ‘entangled-bank’ model, which supposes a long history of cultural and
genetic interactions among Southeast Asians, Melanesians and Polynesians.
Most genetic data, especially analyses of mitochondrial DNA (mtDNA) variation,
support the express-train model, as does linguistic and archaeological
evidence. Here, we used Y-chromosome polymorphisms to investigate the
origins of Polynesians.
Results: We analysed eight single nucleotide polymorphisms (SNPs) and seven
short tandem repeat (STR) loci on the Y chromosome in 28 Cook Islanders from
Polynesia and 583 males from 17 Melanesian, Asian and Australian populations.
We found that all Polynesians belong to just three Y-chromosome haplotypes, as
defined by unique event polymorphisms. The major Y haplotype in Polynesians
(82% frequency) was restricted to Melanesia and eastern Indonesia and most
probably arose in Melanesia. Coalescence analysis of associated Y-STR
haplotypes showed evidence of a population expansion in Polynesians,
beginning about 2,200 years ago. The other two Polynesian Y haplotypes were
widespread in Asia but were also found in Melanesia.
Addresses: *Max Planck Institute for Evolutionary
Anthropology, Inselstraße 22, D-04103 Leipzig,
Germany. †Department of Genetics, Stanford
University, Stanford, California 94305, USA.
‡Institute of Legal Medicine, Humboldt University
Berlin, Hannoversche Straße 6, D-10115 Berlin,
Germany. §Max Planck Institute for Behavioural
Physiology, Von-der-Tann Straße 3, D-82346
Andechs, Germany.
Correspondence: Manfred Kayser
E-mail: kayser@eva.mpg.de
Received: 17 July 2000
Revised: 18 August 2000
Accepted: 18 August 2000
Published: 27 September 2000
Current Biology 2000, 10:1237–1246
0960-9822/00/$ – see front matter
© 2000 Elsevier Science Ltd. All rights reserved.
Conclusions: All Polynesian Y chromosomes can be traced back to Melanesia,
although some of these Y-chromosome types originated in Asia. Together with
other genetic and cultural evidence, we propose a new model of Polynesian
origins that we call the ‘slow-boat’ model: Polynesian ancestors did originate
from Asia/Taiwan but did not move rapidly through Melanesia; rather, they
interacted with and mixed extensively with Melanesians, leaving behind their
genes and incorporating many Melanesian genes before colonising the Pacific.
Background
The origin of the Polynesians — the people living in the
area of the Pacific bounded by Fiji to the west, Hawaii to
the north, Easter Island to the east and New Zealand to the
south — has long drawn the attention of researchers from
different fields. Linguists group all the languages spoken in
Polynesia, Micronesia, the main part of island Melanesia
(excluding the Papuan languages spoken in New Guinea
and a few adjacent islands of Melanesia), island Southeast
Asia, mainland Malaysia and Madagascar into one Austronesian language family that originated on or near
Taiwan [1–3]. Recently, a re-analysis of 1,200 Austronesian
languages revealed that nine out of the ten subgroups containing 26 languages are spoken exclusively by Taiwanese
aborigines, whereas the other 1,174 Austronesian languages
belonging to the tenth subgroup are also spoken outside of
Taiwan [2,3]. Archaeological evidence — mostly from
pottery remains of the Lapita culture — points to southern
China and Taiwan for the origin of Polynesians. Lapita
remains, which are widespread throughout Polynesia and
island Melanesia dating back to 3,600 to 2,500 years ago,
have not been found in New Guinea and Australia, and
have been interpreted to be derived from a similar culture
established about 6,000 years ago in Taiwan and southern
China [4–7].
The archaeological and linguistic evidence have given rise
to the ‘express-train to Polynesia’ hypothesis for the
colonisation of the Pacific [4,8,9], according to which Austronesian-speaking people migrated rapidly during the last
few thousand years from Asia/Taiwan into the Pacific.
Recent genetic data from mitochondrial DNA (mtDNA)
have been interpreted to support this hypothesis and a
Taiwanese origin of Polynesians [10–12]. Data from
human leukocyte antigen (HLA) genes also suggest Polynesian affinities with Asians [13–15].
An alternative ‘entangled-bank’ hypothesis for the colonisation of Polynesia assumes a long and more complex
history of interaction between Polynesia, Melanesia and
1238
Current Biology Vol 10 No 20
Southeast Asia and holds that there was no single discrete
migration event or ‘express-train’ to Polynesia [16]. In this
respect, Lapita has also been interpreted as having evolved
gradually in island Melanesia and introduced to Polynesia
from Melanesia rather than imported from Asia [17,18].
Whereas most of the mtDNA and HLA data have been
interpreted as indicating an Asian/Taiwan origin of Polynesians, the haemoglobin genes of Polynesians do indicate
affinities with Melanesians [19–22]. A third hypothesis,
that Polynesians came from South America [23], receives
no support from any genetic data [12,24] and is not generally accepted by either archaeologists or linguists [6].
To date, conclusions from genetic data regarding Polynesian population history are restricted to maternally inherited mtDNA and recombining autosomal DNA. Insights
from the male-inherited Y chromosome are not available or
are of limited information because of the lack of polymorphic markers in this geographic region [25] or the limited
numbers of Melanesian and Southeast Asian populations
analysed [26]. To investigate the origin of Polynesian
Y chromosomes, we have therefore undertaken an extensive study of eight Y-chromosomal single nucleotide polymorphisms (Y-SNPs) and seven Y-chromosomal short
tandem repeat (Y-STR) loci or microsatellites in 18 populations from Polynesia, Melanesia, Asia and Australia.
Results
We initially investigated 28 males from the Cook Islands
for eight Y-SNPs (M4, M5, M9, M16, M21, M119, M122
and RPS4Y711). This analysis revealed that every Cook
Islander belongs to one of only three Y-chromosome haplotypes (Table 1, Figure 1), defined by polymorphisms at
M9, M122 and RPS4Y711. To investigate the origin of
these Polynesian Y-chromosome haplotypes, we analysed
an additional 583 individuals from 17 populations from
Asia, Australia and Melanesia at the three Y-SNPs that
were found to be polymorphic in the Cook Islanders. We
also analysed seven Y-STRs (DYS19, DYS389I,
DYS389II, DYS390, DYS391, DYS392 and DYS393).
DYS390.3del/RPS4Y711T haplotype
At the Y-STR locus DYS390, alleles of short fragment
length (19–21 repeats) were observed in the Cook
Islanders, mainland and island Papua New Guineans,
eastern Indonesians and Australians, whereas other populations had exclusively alleles of 21–27 repeats in length.
To determine the molecular basis of these short alleles,
we sequenced DYS390 in all individuals with allele
lengths of 18–22 repeats. In addition, from every population in which short alleles were found, we also sequenced
DYS390 in a further three or four individuals that had
alleles longer than 22 repeats. This was so that any underlying sequence heterogeneity in alleles with the same
fragment length would be revealed. In total, we
sequenced DYS390 from 319 individuals. DYS390 is a
complex STR with four different repetitive segments,
designated 390.1–390.4 (Figure 2). All Cook Islanders
with alleles 19–21 had a deletion of the 390.3 segment,
and all of the length variation was in the combined
Table 1
Y-chromosomal haplotypes observed in Polynesia and their frequency distribution in other populations from Melanesia, Asia
and Australia.
Y-chromosomal haplotypes
Population*
DYS390.3del/RPS4Y711T
Cook Islands (Coo, 28)
Papua New Guinea
coast (PNC, 31)
Nusa Tenggara (Ten, 31)
Moluccas (Mol, 34)
Tolai New Britain (TNB, 16)
Trobriand Islands (Tro, 54)
Papua New Guinea
highlands (PNH, 31)
Southern Borneo (Bor, 40)
Korea (Kor, 25)
Han Southern China (Chi, 36)
Han Chinese Taiwan (TaC, 26)
Taiwan aborigines (Tai, 43)
Philippines (Phi, 39)
Vietnam (Vie, 11)
Malay (Mal, 18)
Java (Jav, 53)
Australia 1 (Aus1, 60)
Australia 2 (Aus2, 35)
M122C/M9G
M9G
Others
82.1
7.1
10.7
0
25.8
16.1
14.7
12.5
9.3
9.7
3.2
11.8
6.3
9.3
61.3
64.5
64.7
81.3
81.5
3.2
16.1
8.8
0
0
3.2
2.5
0
0
0
0
0
0
0
0
0
0
0
17.5
28.0
58.3
57.7
11.6
35.9
45.5
27.8
22.6
0
2.9
96.8
72.5
52.0
36.1
38.5
88.4
51.3
45.5
55.6
73.6
35.0
25.7
0
7.5
20.0
5.6
3.8
0
12.8
9.1
16.7
3.8
65.0
71.4
*The abbreviations used for each population and the number of individuals sampled are indicated within the brackets.
Research Paper Melanesian origin of Polynesian Y chromosomes Kayser et al.
1239
Figure 1
Y-chromosomal haplotypes observed in
Polynesia and their frequency distribution
in other populations from Melanesia, Asia
and Australia (for abbreviations, see Table 1).
Red, DYS390.3del/RPS4Y711T; blue,
M122C/M9G; yellow, M9G; green,
other haplotypes.
Kor
Chi
Tai
TaC
Pacific Ocean
Phi
Vie
Mal
Mol
PNH
Ten
TNB
Bor
Tro
Jav
PNC
Aus1
Coo
Indian Ocean
Aus2
Current Biology
390.2–390.4 segment. All other individuals in our study
with alleles 19 and 20 also had the 390.3 deletion, with the
exception of Australians who instead had the previously
described deletion in 390.1 ([27]; M.K. and M.S., unpublished work). Some individuals from mainland/island
Papua New Guinea and Indonesia with alleles 21–23 had
the 390.3 deletion and some did not, whereas none of the
individuals sequenced with alleles of at least 24 repeats
had the 390.3 deletion.
Hurles et al. [26] also reported DYS390 alleles 20 and 21 in
different samples of Cook Islanders and coastal Papua
New Guinea. They found that all chromosomes with
these alleles also carried a specific allele at the Y-chromosomal minisatellite locus MSY1. In their study, sequence
analysis was not performed to clearly identify the
DYS390.3 deletion, nor was the RPS4Y711 marker
analysed. Nevertheless, we can infer that the individuals
An analysis of eight Y-SNPs revealed that the DYS390.3
deletion is associated with a C→T mutation at position
711 of the RPS4Y gene [28]. The combined DYS390.3
deletion and RPS4Y711T haplotype (DYS390.3del/
RPS4Y711T) had a frequency of 82% in the Cook
Islands, 26% in coastal Papua New Guinea, 10–15% in
the Moluccas and Nusa Tenggara Islands of eastern
Indonesia, 9–12% in island Papua New Guinea (New
Britain, Trobriand Islands) and was observed once in the
Papua New Guinea highlands and southern Borneo, but
it was not found in any other Southeast Asian, East Asian
or Australian population studied here (Table 1,
Figure 1). In the Cook Islanders and mainland/island
Papua New Guineans, RPS4Y711T was completely associated with the 390.3 deletion; all individuals with the
390.3 deletion had this mutation, while all individuals
lacking the 390.3 deletion carried the ancestral
RPS4Y711C. Only in Indonesia, especially eastern
Indonesia, some individuals with the RPS4Y711T carried
the 390.3 deletion and some did not.
Figure 2
DYS390.1
DYS390.2
390.3DYS390.4
Putative pre-mutation
allele
(CTGT)1 deletion
DYS390.3 deletion sequence motif
Multiple random
deletions / insertions
Single copy
CTGT
CTAT
CAT
Current Biology
Schematic structure of the repetitive sequence at the Y-STR locus
DYS390 and proposed mutational event of the DYS390.3 deletion.
Each square represents one repeat unit.
1240
Current Biology Vol 10 No 20
Figure 3
Haplotype diversity (dark grey bars), their
standard deviation and the mean number of
pairwise differences (light grey bars) of Y-STR
haplotypes. (a) DYS390.3del/RPS4Y711T
haplotype. (b) M122C/M9G haplotype. In (a),
Indonesia includes five individuals from Mol,
five from Ten and one from Bor; Melanesia
includes one individual from PNH, eight from
PNC, two from Port Moresby, five from Tro,
and two from TNB; Polynesia includes
39 from Coo and seven Western Samoans. In
(b), East Asia includes 21 individuals from
Chi, 15 from TaC, seven from Kor, and five
from Vie; Southeast Asia includes
14 individuals from Phi, 12 from Jav, seven
from Bor, five Mal, five Tai, four from Mol and
one from Ten; and Oceania includes five
(a)
6.0
5.44
5.0
4.00
4.0
2.73
3.0
0.97
0.90
2.0 0.87
±0.03
±0.02
1.0 ±0.09
0.0
Indonesia Melanesia Polynesia
(11)
(18)
(46)
(b)
5.61
6.0
5.0
4.42
4.0
3.13
3.0
0.96
0.95
2.0 0.99
±0.01
±0.02
±0.07
1.0
0.0
East Asia Southeast Oceania
(48)
Asia
(11)
(48)
Current Biology
individuals from Tro, three from PNC, two
from Coo and one from TNB (a single
Australian was excluded). The units on the
they observed with DYS390 alleles 20–21 belong to the
DYS390.3del/RPS4Y711T haplotype that we observed,
because they have an A→G mutation at SRY1532 [26],
which is associated with the RPS4Y711T mutation [29,30].
In a previous study, the DYS390.3 deletion was also found
and characterised by sequence analysis in 7 out of 10 individuals from Western Samoa [27,31], although RPS4Y711
was not analysed.
Combining the present data with these published data
revealed a total of 75 individuals from 11 population
samples from Polynesia, Papua New Guinea (mainly
island and coastal Papua New Guinea), and Indonesia
(mainly Moluccas and Nusa Tenggara) that carry the
DYS390.3del/RPS4Y711T haplotype (Table 1, Figure 1);
39 seven-locus Y-STR haplotypes were observed within
these 75 individuals. Pairwise RST analysis based on
Y-STR haplotypes revealed statistically significant differences (p < 0.01) between the following three groups:
vertical axis are the mean number of pairwise
differences, for the light grey bars, and the
haplotype diversity, for the dark grey bars.
Melanesia (mainly coastal and island Papua New Guinea),
Polynesia (mainly Cook Islands), and Indonesia (mainly
Moluccas and Nusa Tenggara Islands). The highest
Y-STR haplotype diversity was found in Melanesia and
the lowest in Indonesia/Polynesia (Figure 3a). Also, the
highest mean number of pairwise differences between
haplotypes was found within Melanesia and the lowest
was in Polynesia (Figure 3a). Y-STR haplotype sharing
was only observed within groups, mainly within Polynesia,
but not between any groups.
A median-joining network connecting all 39 haplotypes
revealed that all Polynesian haplotypes form a tight cluster
and can be connected to each other mostly by single-step
mutations, with the exception of one of the West Samoan
haplotypes (Figure 4). In contrast, Melanesian and
Indonesian haplotypes appeared in different parts of the
network and were separated by a large number of mutations. The network shown in Figure 4 was calculated by
Figure 4
31 39
36 37
24
33
27
32
21
17
22
15
13
16
12
23 34 18
28
11
14
38
26
7
35
2
6
19
4
1
Median-joining network of 39 Y-STR
haplotypes from 75 individuals belonging to
the DYS390.3del/RPS4Y711T haplotype.
Circles, Y-STR haplotypes with the area
proportional to the number of individuals;
lines, mutation steps; parallel lines, identical
mutations. Red haplotypes are from
Polynesia, blue from Melanesia, and yellow
from Indonesia. The scale bar indicates one
mutation.
3
9
20
29
30
5
25
10
8
1
Current Biology
Research Paper Melanesian origin of Polynesian Y chromosomes Kayser et al.
1241
Table 2
Demographic inferences concerning the DYS390.3del/RPS4Y711T and the M122C/M9G haplotype based on associated variation
at seven Y-STR loci.
Posterior probabilities*
DYS390.3 deletion‡
M122C§
0.40 (0.06; 2.86)
0.33 (0.11; 1.02)
0.25 (0.07; 0.78)
Population growth rate per generation × 10–3
6.9 (0.3; 36.9)
6.4 (1.0; 22.5)
14.5 (5.9; 33.2)
Time of expansion in 1,000 years
4.9 (0.1; 64.6)
5.0 (1.2; 16.9)
6.0 (2.7; 14.5)
Time back to mrca in 1,000 years
17.1 (1.6; 100.6)
11.5 (5.0; 32.5)
11.1 (5.1; 28.3)
Parameter
Initial effective population size in 1000 individuals
Prior
probabilities*†
*Median (and 95% equal-tailed intervals). †Prior probabilities used for calculations (see Materials and methods and Figure 5). ‡Based on 75
individuals and 39 Y-STR haplotypes (diversity: 0.958 ± 0.011). §Based on 108 individuals and 81 Y-STR haplotypes (diversity: 0.984 ± 0.006).
weighting the loci according to mutation rate; calculating
the network with equal weights for all loci resulted in
increased reticulation within the Polynesian cluster while
the rest of the network remained virtually unchanged
(data not shown). A minimum-spanning network based on
the sum of size differences between the distinct haplotypes also gave a very similar picture of haplotype relationships (data not shown).
To infer demographic data from Y-STR haplotypes of
individuals carrying the DYS390.3del/RPS4Y711T haplotype, a Bayesian-based coalescence approach was used.
The time back to the most recent common ancestor
(mrca) of all 75 individuals carrying the DYS390.3 deletion
on the RPS4Y711T chromosome background was estimated to be 11,500 years (Table 2, Figure 5). A signal of
slight population growth dating back to the start of a population expansion ~5,000 years ago was detected. When the
analysis was restricted to Polynesians, a much stronger
signal of population growth was detected, indicating a
population expansion starting about 2,200 years ago.
These results did not change significantly when different
prior probability distributions were applied (data not
shown), indicating that these results do reflect the data
and not the prior distributions.
differences for RST based on Y-STR haplotypes for all
pairwise comparisons (p < 0.05) except Southeast Asia
with Oceania. Y-STR haplotype diversity and mean
number of pairwise differences were highest in East Asia,
lower in Southeast Asia and lowest in Oceania (Figure 3b).
Shared Y-STR haplotypes between groups were observed
for East versus Southeast Asia (four haplotypes) and
Southeast Asia versus Oceania (one haplotype). The time
back to the mrca of the 108 individuals carrying the
M122C mutation was estimated to be 11,100 years
(Table 2, Figure 5). For all individuals carrying the
M122C/M9G haplotype, a signal of moderate population
growth was detected, with the start of population expansion about 6,000 years ago. The M9G haplotype was
Figure 5
(a)
(b)
Pr
Pr
5 10 15 20
Time of expansion
in 1000 years
(d)
(c)
Additional major Y-chromosome haplotypes identified in
Polynesia
In addition to the DYS390.3del/RPS4Y711T haplotype
only two other Y-SNP haplotypes were observed in the
Cook Islanders (Figure 1). The M122C/M9G haplotype
occurred at a frequency of 7.1% whereas M9G had a frequency of 10.7% (Table 1). Outside Polynesia, the
M122C/M9G haplotype was frequent in East Asian and
Southeast Asian populations, but was also present in low
frequency in all Melanesian populations analysed; it was
completely absent in highland Papua New Guinea and
nearly absent (1 out of 95) in Australia. Classifying all individuals into three groups consisting of East Asia, Southeast Asia, and Oceania revealed statistically significant
5 10 15 20 25 30 35
Time back to mrca
in 1000 years
Pr
Pr
5 10 15 20
Time of expansion
in 1000 years
5 10 15 20 25 30 35
Time back to mrca
in 1000 years
Current Biology
Bayesian-based demographic data inferred from Y-STR variation
associated with the (a,b) DYS390.3del/RPS4Y711T (n = 75) and
(c,d) M122C/M9G (n = 108) haplotypes. Each panel shows the prior
(dashed lines) and posterior (continuous lines) probability distribution
of the (a,c) time of population expansion and (b,d) time back to the
mrca. Pr indicates probability.
1242
Current Biology Vol 10 No 20
present in all populations analysed and at higher frequency
than in the Cook Islanders (Table 1). As M9G also occurs
outside Asia, especially in Europe, it does not provide any
additional information concerning Polynesian origins.
Discussion
Three Y-chromosomal haplotypes, as defined by three
Y-SNPs and a specific deletion at the Y-STR locus DYS390,
were identified in a Polynesian population sample from the
Cook Islands. One of these, the DYS390.3del/RPS4Y711T
haplotype, was found at a markedly high frequency of 82%.
Hurles et al. [26] observed a lower frequency of this haplotype (48%) in an independent sample from the Cook
Islands. However, they observed significant European
admixture in their sample, whereas for our sample Cook
Island paternal ancestry was documented for at least two
generations and, thus, was more likely to reflect native
Polynesian ancestry. Moreover, we found the DYS390.3del/
RPS4Y711T haplotype at a frequency of 70% in Western
Samoa, and Deka et al. [32] reported a frequency of 60% of
alleles 17–22 at DYS390 in American Samoa (although
they did not characterise these alleles by DNA sequencing or SNP analysis, they are likely to belong to the
DYS390.3del/RPS4Y711T haplotype). Thus, we conclude
that this is the major Y-chromosome haplotype in Polynesia.
Outside Polynesia, the DYS390.3del/RPS4Y711T haplotype was found to be present only in Melanesia and eastern
Indonesia. As the DYS390.3 deletion in Polynesia was
always observed on RPS4Y711T chromosomes, whereas
RPS4Y711T chromosomes without the DYS390.3 deletion
were observed in nearly all of the populations analysed here,
with the exception of Melanesia and Polynesia (data not
shown), we conclude that the DYS390.3 deletion occurred
on a RPS4Y711T chromosome. RPS4Y711T chromosomes
are widespread among East and Southeast Asian, Indian,
Australian and native American populations, but are not
found in Europe or Africa ([28,30,33]; M.K. and M.S.,
unpublished data). Also, the DYS390.3 deletion has not
been identified anywhere else other than in the populations
mentioned above when more than 600 individuals from 35
populations from Asia, Africa, Europe and America have
been sequenced ([27,34]; this study; M.K., unpublished
data), with the exception of a single Turk and a single
Ovambo from Namibia [34]. This gives rise to the conclusion that the DYS390.3del/RPS4Y711T haplotype is indeed
restricted to Melanesia, Polynesia and eastern Indonesia,
and thus can serve as a suitable marker for investigating
population history in this region of the world.
We have used Y-STR haplotypes to infer the time and
place of origin of the DYS390.3 deletion on the
RPS4Y711T background and, thus, the direction of gene
flow. Haplotype diversity and the mean number of pairwise differences were higher in Melanesia than in Polynesia or Indonesia and a coalescence-based approach
indicated that the deletion arose about 11,500 years ago.
These results therefore indicate that the major Y-chromosome haplotype in Polynesians originated in Melanesia.
A second hypothesis is that the DYS390.3del/RPS4Y711T
haplotype arose in Proto-Polynesians (presumably Austronesian speakers) in Melanesia as they were expanding on
their way to Polynesia. Support for this hypothesis comes
from the frequency distribution of the DYS390.3del/
RPS4Y711T haplotype in Melanesia; it was present in
coastal Papua New Guinea and island Melanesia but nearly
absent from highland Papua New Guinea. However, this is
contradicted by the amount of Y-STR diversity in Melanesia and the related age of the haplotype (about
11,500 years), as Proto-Polynesians did not reach Melanesia
until about 3,500 years ago [9].
A third hypothesis is that the DYS390.3del/RPS4Y711T
haplotype arose in Polynesia, and therefore the presence
of this haplotype in Melanesia represents a substantial
back-migration from Polynesia. The high frequency of
this haplotype in Polynesia, and the central position of
Polynesian types in the Y-STR haplotype network,
provide some support for this explanation. However, the
diversity analyses suggest a Melanesian origin, and the
date for the origin of this haplotype (11,500 years) substantially predates any other evidence for human occupation of Polynesia. Bottleneck effects provide a likely
alternative explanation for the elevated frequency in Polynesia and for the central position of Polynesian types in
the Y-STR network. Moreover, there is no other genetic
evidence and little archaeological evidence to suggest a
substantial back-migration from Polynesia to Melanesia;
although one study [31] does suggest back-migration as an
explanation for the high frequency of the mtDNA 9 bp
deletion marker in island and coastal Papua New Guinea,
others do not favour this interpretation [10–12]. A Melanesian origin thus remains the most likely explanation for
the DYS390.3del/RPS4Y711T haplotype in Polynesians.
What about the other Y-SNP haplotypes found in Polynesia? The M122C/M9G haplotype was observed in 7.1% of
Cook Islanders, but to date the M122 marker has not been
analysed in other Polynesian populations. This haplotype
has a high frequency in East and Southeast Asia, as
observed here and elsewhere [35,36], but has not been
found in Africa, America or Europe [35]. Thus, the
M122C mutation probably arose in Asia, on an M9G
Y chromosome, about 11,100 years ago as indicated by the
coalescence approach. The M122C/M9G haplotype is
found in Melanesia, but only in coastal Papua New
Guinea and island Melanesia; it is absent from highland
Papua New Guinea and nearly absent from Australia (1 in
95). This distribution corresponds closely to that of the
mtDNA 9 bp deletion marker [10–12], which is thought to
reflect the presumed Austronesian expansion from Asia/
Taiwan through coastal and island Melanesia into the
Research Paper Melanesian origin of Polynesian Y chromosomes Kayser et al.
Pacific [4,6,9]. The M122T/M9G haplotype also shows a
reduction in Y-STR haplotype diversity and in the mean
number of pairwise differences from mainland East Asia to
Southeast Asia to Oceania (coastal and island Papua New
Guinea, Polynesia). Furthermore, the detected moderate
population growth and the estimated start of population
expansion at about 6,000 years ago is in perfect agreement
with archaeological data, which suggest that the Austronesian expansion started about 6,000 years ago from Asia/
Taiwan [4,6,9]. Unfortunately, the frequency of the
M122C/M9G haplotype in our Melanesian and Polynesian
samples is too low to permit accurate analysis for signs of
regional population expansion. Although the M122C/M9G
haplotype in Polynesia could reflect an Austronesian expansion, it could also reflect simply a direct Melanesian contribution, as with the DYS390.3del/RPS4Y711T haplotype.
The third Y-chromosome haplotype observed in the Cook
Islanders was M9G, at a frequency of 10.7%. This haplotype is not useful to investigate population relationships in
Polynesia as it is the common ancestor of the majority of
non-African haplotypes [35,37]. Thus, M9G alone is not
suitable to characterise a distinct Y-chromosomal haplotype,
as many Y-SNPs have occurred on an M9G background
[37], including, for example, the M122C mutation. In particular, M9G could reflect recent European admixture; in
fact, three Cook Islanders that were typed but not included
in our study because family records indicated paternal
European ancestry all had the M9G haplotype. Additional
markers on the M9G background are needed to address the
possible Melanesian origin of M9G in Polynesia.
Our study showed reduced Y-STR haplotype diversity
within the DYS390.3del/RPS4Y711T haplotype in Polynesians when compared with Melanesians and Indonesians. Furthermore, the overall Y-SNP and Y-STR
haplotype diversity was found to be lowest in Polynesia
when compared with the other 17 populations analysed
here (M.K. and M.S., unpublished data). Reduced genetic
diversity in Polynesians has also been reported for many
other genetic markers, indicating a Polynesian bottleneck
[11,12,38,39]. Moreover, when dividing the total sample
set of individuals carrying the DYS390.3del/RPS4Y711T
haplotype into Polynesians and non-Polynesians, the population growth rate of Polynesians was estimated to be
four times larger than for non-Polynesians, with a population expansion starting 2,200 years ago. This is in agreement with the hypothesis of a bottleneck in the
colonisation of Polynesia, which would result in a stronger
signal of population growth coming out of the bottleneck.
This date is also in remarkably good agreement with
archaeological data suggesting that the Cooks and surrounding islands were settled about 2,200 years ago [9].
We conclude that the major Y-chromosome haplotype in
Polynesians has an origin from Melanesia, and in fact this
1243
may be the case for all Y-chromosome haplotypes in Polynesia, as all of the haplotypes in Polynesia are also found
in Melanesia. Recently, Su et al. [40] found no evidence
for a Melanesian origin of Polynesian Y chromosomes,
because their major Melanesian Y-chromosomal haplotype
(H17, characterised by mutations at M4, M5 and M9 [37])
was not found in Polynesia. We also found this haplotype
in high frequency in Melanesia (M.K. and M.S., unpublished data) and concur that it is absent from Polynesia;
however, the absence of this Melanesian haplotype can be
explained by founder effects and genetic drift. Su et al.
[40] did not analyse the markers RPS4Y711 and DYS390;
the DYS390.3del/RPS4Y711T haplotype would be
included in their complete ancestral (with respect to the
analysed markers) haplotype H1, which they reported
were found in frequencies of 15–33% in Melanesians and
30–48% in Polynesians. Therefore, their results may actually be compatible with our conclusions.
Other genetic studies have also provided evidence for
Melanesian gene flow into Polynesia. For example, the
α-haemoglobin –α3.7 III deletion is restricted to (mainly
coastal) Papua New Guinea, island Melanesia, Micronesia
and Polynesia, and probably originated in Melanesia
[19–22]. Although some HLA genes in Polynesia are
claimed to show Asian rather than Melanesian origin
[13–15], the HLA data are equivocal. A recent phylogenetic analysis of DRB1–DQB1 haplotypes grouped
Polynesians with Melanesians [41]. Also, a particular
allele, HLA DRB1-0901, that was observed at high frequency (26–45%) in Polynesians [41,42] and at moderate
frequency (10–15%) in mainland Asia [43–45] was originally reported to be absent or rare in Melanesian populations [41,44,46] but, recently, this allele has been observed
at a frequency of 18% in the Trobriands from island
Melanesia and 14% in the Roro from the coast of Papua
New Guinea [42]. Thus, for the HLA DRB1-0901 allele, a
Melanesian contribution to Polynesia is not ruled out,
especially as the DYS390.3del/RPS4Y711T haplotype is
also found in these same populations.
Studies of the mtDNA 9 bp deletion marker, and the associated ‘Polynesian’ sequence motif in hypervariable
region I of the mtDNA control region, have suggested a
Taiwanese origin for Polynesian mtDNAs [10–12]. This is
in agreement with the express-train to Polynesia hypothesis [4,6,9], but seems to disagree with the Y-chromosome
data presented here and elsewhere [40]. One explanation
might be sex-specific migration patterns [47]. However,
although most Polynesian mtDNAs have the 9 bp deletion, about 3.5–10% do not [12,48]. Significantly, these
Polynesian mtDNA types that lack the 9 bp deletion are
found in Papua New Guinea but not elsewhere [12,48];
moreover, all Polynesian mtDNA types are also found in
Melanesia, as is the case with Polynesian Y chromosomes.
This would suggest that Polynesian mtDNA ancestors did
1244
Current Biology Vol 10 No 20
not simply migrate through New Guinea, but rather that
they interbred with the local Melanesian populations,
leaving behind mtDNA types with the 9 bp deletion and
incorporating Melanesian mtDNA types. This scenario is
compatible with the Y chromosome results.
Conclusions
Most, if not all, Polynesian genes examined to date are
also found in Melanesia, and thus a Melanesian origin is
not only demonstrable for some genes (such as the Y-chromosomal DYS390.3del/RPS4Y711T haplotype), it cannot
be ruled out for any gene. Although this would appear to
argue against the express-train hypothesis and in favour of
the alternative entangled-bank hypothesis of Polynesian
origins, it is also true that some of the genes in Melanesia
do reflect an expansion out of Southeast Asia that appears
to be associated with the expansion of the Austronesian
language family. This has been demonstrated for mtDNA
[10–12] and may also be true for the Y-chromosomal
M122C/M9G haplotype. The overall picture appears to be
more complicated than a simple express-train hypothesis
of Polynesian ancestors moving quickly through Melanesia with little genetic impact; a more apt metaphor might
be a ‘slow-boat’ model, in which Polynesian ancestors
moved gradually across Melanesia, mixing extensively
with local Melanesian populations, and thereby not only
left behind their genes, but also incorporated many
Melanesian genes. This scenario is compatible with cultural indications of interactions between Polynesian ancestors and Melanesians, and with linguistic evidence for
‘pauses’ during the spread of Austronesian languages
through the Pacific [2,49,50].
Materials and methods
DNA samples
DNA samples from 611 male individuals from 18 populations of the following geographic locations were analysed. Polynesia: 28 Cook
Islanders (13 from Rarotonga, 4 from Mauke, 3 from Manihiki, 1 from
Penrhyn, 2 from Mitiaro, 3 from Aitutaki, 1 from Atiu and 1 from
Mangaia); Papua New Guinea: 31 highland New Guineans and 31
coastal New Guineans, both described elsewhere [51], 16 Tolai from
Vunapope New Britain, and 54 Trobriand Islanders from Tauwema
Kaileuna Island; Australia: 60 aborigines from Arnhem Land (Aus1) and
35 from Great Sandy Desert (Aus2), both described elsewhere [52];
Southeast Asia: 34 eastern Indonesians from the Moluccan Islands (20
from Hiri and 14 from Ternate), 31 from the Nusa Tenggara Islands (8
from Alor, 2 from Flores, 11 from Roti and 10 from Timor, all described
elsewhere [11]), 53 Javanese from a rural area near Jakarta, 40 individuals from southern Borneo and 18 Malay, both described elsewhere
[10]; Eastern Asia: 11 southern Vietnamese, 26 Han Chinese from
Taiwan, 25 southern Koreans (the latter three groups are first generation immigrants to the USA), 36 Han Chinese from Hanzhou (province
Zhejiang, south coast of China), 43 Taiwan aborigines (10 Ami, 10
Atayal, 10 Bunum, 13 Paiwan) and 39 Philippinos; the latter were
described elsewhere [10]. Care was taken to include unrelated males
only for all populations and to assure native ancestry for at least one or
two generations for the following areas: Cook Islands, Trobriand
Islands, New Britain, coastal and highland Papua New Guinea, eastern
Indonesia. For statistical analyses, published data for 33 Cook
Islanders from Rarotonga and 58 Papua New Guineans from Port
Moresby [26] and 10 Western Samoans [27,31] were included.
Genotyping and DNA sequencing
The Y-STRs DYS19 (or DYS394), DYS389I, DYS389II, DYS390,
DYS391, DYS392 and DYS393 were analysed by PCR using published primers, allele nomenclature and protocols [53,54]. Fragment
length analysis was performed using either an A.L.F. express (Pharmacia) or an ABI PRISM 377 DNA Sequencer (PE Biosystems). DNA
sequencing of DYS390 was performed on both strands, either directly
from the PCR product or after reamplification, using published primers
[53]. Prior to sequence analysis, PCR products were purified using the
QUIquick PCR purification kit (Qiagen) or, before and after reamplification, using a 3% NuSieve (FMC Bioproducts) agarose gel and the
Wizard Plus Minipreps DNA purification system (Promega). Sequencing reactions were prepared with the Big Dye Reaction Terminator
Cycle Sequencing Kit (PE Biosystems) and purified by isopropanol
precipitation prior to analysis on an ABI PRISM 377 DNA Sequencer.
The Y-SNPs analysed include M4, M5, M9, M16, M21, M119, M122
and RPS4Y711 [28,37]. M9, M122 and RPS4Y711 were analysed by
PCR using the following standard conditions: 0.4 µM of each primer, 1×
GeneAmp PCR buffer (PE Biosystems), 1 U AmpliTaq Gold DNA polymerase (PE Biosystems), 0.2 µM dNTPs (Pharmacia Biotech), 147 µM
bovine serum albumin (Sigma), 10–100 ng DNA and a hot-start PCR of
11 min at 95°C (initial denaturation), followed by 30–50 cycles of
30 sec at 94°C, 30 sec at the locus-specific annealing temperature, and
45 sec at 72°C, followed by a final step of 10 min at 72°C. For M9, the
PCR primers 5′-GCAGCATATAAAACTTTCAGG-3′ and 5′-GAAATGCATAATGAAGTAAGCG-3′ were used with an annealing temperature
of 54°C. The M9 C→G mutation was detected by single-strand oligonucleotide (SSO) hybridisation as described elsewhere [10] using the
5′-biotin-labelled probes: 5′-GATGGTTGAATCCTCTTTAT-3′ for the
ancestral C allele, and 5′-ATAAAGAGCATTCAACCATC-3′ for the
mutated G allele, with two stringent washes at 55°C for 10 min for each
of the probes. Alternatively, the M9 C→G mutation was screened by
PCR–RFLP using 10–20 µl PCR product, 1× buffer 2 (New England
Biolabs), 10 U Hinf I (New England Biolabs) at 37°C overnight, resulting
in one undigested fragment (164 bp) for the mutant G allele or two
digested fragments (100 bp, 64 bp) for the ancestral C allele. For
M122, the PCR primers 5′-GTTGCCTTTTGGAAATGAATAAATC-3′
and 5′-CACTTGCTCTGTGTTAGAAAAGATAGC-3′ were used with an
annealing temperature of 58°C. The M122 T→C mutation was detected
by PCR–RFLP using 10–20 µl PCR product, 1× buffer 2 (Promega),
1.1 µM bovine serum albumin and 10 U Hsp92II (Promega) at 37°C
overnight. Fragments were resolved on a 3% NuSieve agarose gel,
resulting in one undigested fragment (109 bp) for the mutant C allele, or
two digested fragments (58 bp, 51 bp) for the ancestral T allele. For
RPS4Y711, the PCR-primers 5′-CTGTACTTACTTTTATCTCCTC-3′
and 5′-CAGCAACAGTAAGTCGAATG-3′ were used with an annealing
temperature of 55°C. The RPS4Y711 C→T mutation was detected by
PCR–RFLP using 10–20 µl PCR product, 1× buffer 2 Bsl I (New
England Biolabs), 5 U Bsl I (New England Biolabs) at 55°C overnight.
Fragments were resolved on a 3% NuSieve agarose gel, resulting in one
undigested fragment (91 bp) for the mutant T allele or two digested
fragments (34 bp, 57 bp) for the ancestral C allele. For the Y-SNPs M4,
M5, M16, M21 and M119, primers, PCR conditions, and genotyping
methods will be described in detail elsewhere (M.K. and M.S., unpublished work) and are available from the authors.
Statistical analyses
Y-STRs were analysed with respect to haplotype diversity and the
associated standard deviation, mean number of pairwise differences
between haplotypes, pairwise RST values and associated p values
based on 10,000 permutations, using the software package Arlequin
version 2.000 [55]. A median-joining network [56] based on Y-STR
haplotypes was calculated using the Network 2.0b software
(http://www.fluxus-engineering.com/sharenet.htm). For network calculation, locus-specific weights were given according to the recently
observed mutation rates for the Y-STRs used here [57] so that loci with
the highest mutation rates were given the lowest weights (ratio of
DYS393:DYS392:DYS19:DYS389I:DYS389II:DYS391:DYS390 =
10:10:5:5:2:2:1). As the DYS389II PCR product also contains
Research Paper Melanesian origin of Polynesian Y chromosomes Kayser et al.
DYS389I, for all statistical analyses, a simple subtraction of the
DYS389I repeat length from that of DYS389II was done.
Bayesian-based coalescence analyses of Y-STR haplotypes were performed using the software BATWING (http://www.maths.abdn.
ac.uk/~ijw/batwing). The principles of this Markov chain Monte Carlo
based inference method were described elsewhere [58]. We chose a
two-phase population model, where in the past the population size was
of constant size, N, followed by a period of exponential growth until
present. We assigned gamma-distributed prior distributions to the
mutation rates of the seven loci adjusted to the corresponding estimates in Kayser et al. [57]. For the initial effective population size, we
used a lognormal prior distribution with mode 148, median 403 and
mean 665, corresponding to a small initial founder population. The
population growth rate prior (per N generations) was an exponential
distribution with mean 1, which covers the simple constant population
size model as well as reasonable growth rates for human population.
The length of the growth period (in units of N times generation time)
also had an exponential prior with mean 1.
Acknowledgements
We thank the original donors of samples, and E. Hagelberg, S. Ulijaszek,
K. Katayama, N. Saha, A.S.M. Sofro, K. Bhatia, J. Kuhl, N. Kretchmer,
D. Bugawan, J. Martinson, B. Budowle and C. Tyler-Smith for providing
DNA or blood samples. C. Tyler-Smith is additionally acknowledged for
sharing data before publication. We thank S. Pääbo and M. Nagy for
helpful comments. This work was funded by a grant from the National
Science Foundation to M.S. and by the Max Planck Society.
References
1. Blust R: The Austronesian homeland: a linguistic perspective.
Asian Perspect 1984/85, 26:45-67.
2. Blust R: Subgrouping, circularity and extinction: some issues in
Austronesian comparative linguistics. Symp Ser Inst Linguist Acad
Sinica 1999, 1:31-94.
3. Diamond JM: Taiwan’s gift to the world. Nature 2000, 403:709-710.
4. Bellwood PS: Man’s Conquest of the Pacific: The Prehistory of
Southeast Asia and Oceania. Oxford: Oxford University Press; 1978.
5. Spriggs M: The Lapita cultural complex: origins, distribution,
contemporaries and successors. In Out of Asia. Edited by Kirk R,
Szathmary E. Canberra: Journal of Pacific History; 1985:185-206.
6. Bellwood PS: The Polynesians. London: Thames and Hudson; 1987.
7. Kirch PV: The Lapita Peoples. Cambridge, Massachusetts:
Blackwell; 1995.
8. Diamond JM: Express train to Polynesia. Nature 1988, 336:307-308.
9. Bellwood PS: The colonization of the Pacific: some current
hypotheses. In The Colonization of the Pacific: A Genetic Trail.
Edited by Hill AVS, Serjeantson SW. Oxford: Oxford University Press;
1989:1-59.
10. Melton T, Peterson R, Redd AJ, Saha N, Sofro ASM, Martinson J,
et al.: Polynesian genetic affinities with Southeast Asian
populations as identified by mtDNA analysis. Am J Hum Genet
1995, 57:403-414.
11. Redd AJ, Takezaki N, Sherry ST, McGarvey ST, Sofro ASM, Stoneking
M: Evolutionary history of the COII/tRNA-lys intergenic 9 base
pair deletion in human mitochondrial DNAs from the Pacific. Mol
Biol Evol 1995, 12:604-615.
12. Sykes B, Leiboff A, Low-Beer J, Tetzner S, Richards M: The origins of
the Polynesians: an interpretation from mitochondrial lineage
analysis. Am J Hum Genet 1995, 57:1463-1475.
13. Serjeantson SW, Ryan DP, Thompson AR: The colonization of the
Pacific: the story according to human leukocyte antigens. Am J
Hum Genet 1982, 34:904-918.
14. Serjeantson SW: Migration and admixture in the Pacific. J Pac Hist
1984, 19:123-146.
15. Serjeantson SW: HLA genes and antigens. In The Colonization of
the Pacific: A Genetic Trail. Edited by Hill AVS, Serjeantson SW.
Oxford: Oxford University Press; 1989:120-173.
16. Terrell J: Prehistory in the Pacific Islands. Cambridge: Cambridge
University Press; 1986.
17. Anson D: Lapita pottery of the Bismarck Archipelago and its
affinities. Archeol Oceania 1986, 21:157-165.
18. Terrell J: What Lapita is and what Lapita isn’t. Antiquity 1989,
63:623-626.
1245
19. Hill AVS, Bowden DK, Trent RJ, Higgs DR, Oppenheimer SJ, Thein
SL, et al.: Melanesians and Polynesians share a unique
α-thalassemia mutation. Am J Hum Genet 1985, 37:571-580.
20. Flint J, Hill AVS, Bowden DK, Oppenheimer SJ, Sill PR, Serjeantson
SW, et al.: High frequencies of α-thalassemia are the result of
natural selection by malaria. Nature 1986, 321:744-749.
21. Hill AVS, Gentile B, Bonnardot JM, Roux J, Weatherall DJ, Clegg JB:
Polynesian origins and affinities: globin gene variants in eastern
Polynesia. Am J Hum Genet 1987, 40:453-463.
22. O’Shaughnessy DR, Hill AVS, Bowden DK, Weatherall DJ, Clegg JB:
Globin gene in Micronesia: origin and affinities of Pacific Island
peoples. Am J Hum Genet 1990, 46:144-155.
23. Heyerdahl T: Kontiki: Across the Pacific by Raft. Chicago: Rand
McNally; 1950.
24. Bonatto SL, Redd AJ, Salzano FM, Stoneking M: Lack of ancient
Polynesian-Amerindian contact. Am J Hum Genet 1996, 59:253-256.
25. Spurdle AB, Woodfield DG, Hammer MF, Jenkins T: The genetic
affinity of Polynesians: evidence from Y chromosome
polymorphisms. Ann Hum Genet 1994, 58:251-263.
26. Hurles ME, Irven C, Nicholson J, Taylor PG, Santos FR, Loughlin J,
et al.: European Y-chromosomal lineages in Polynesia: a contrast
to the population structure revealed by mtDNA. Am J Hum Genet
1998, 63:1793-1806.
27. Forster P, Kayser M, Meyer E, Roewer L, Pfeiffer H, Benkmann H,
et al.: Phylogenetic resolution of complex mutational features at
Y-STR DYS390 in aboriginal Australians and Papuans. Mol Biol
Evol 1998, 15:1108-1114.
28. Bergen AW, Wang C-Y, Tsai J, Jefferson K, Dey C, Smith KD, et al.:
An Asian-Native American paternal lineage identified by RPS4Y
resequencing and by microsatellite haplotyping. Ann Hum Genet
1999, 63:63-80.
29. Pandya A: Human Y-chromosomal DNA variation. DPhil thesis. Oxford,
Oxford University; 1998.
30. Karafet TM, Zegura SL, Posukh O, Osipova L, Bergen A, Long J, et al.:
Ancestral Asian source(s) of New World Y-chromosome founder
haplotypes. Am J Hum Genet 1999, 64:817-831.
31. Hagelberg E, Kayser M, Nagy M, Roewer L, Zimdahl H, Krawczak M,
et al.: Molecular genetic evidence for the human settlement of the
Pacific: analysis of mitochondrial DNA, Y chromosome and HLA
markers. Phil Trans R Soc Lond B 1999, 354:141-152.
32. Deka R, Jin L, Shriver MD, Yu LM, Saha N, Barrantes R, et al.:
Dispersion of human Y chromosome haplotypes based on five
microsatellites in global populations. Genome Res 1996,
6:1177-1184.
33. Santos FR, Pandya A, Tyler-Smith C, Pena SD, Schanfield M, Leonard
WR, et al.: The central Siberian origin for native American Y
chromosomes. Am J Hum Genet 1999, 64:619-628.
34. Forster P, Rohl A, Lunnemann P, Brinkmann C, Zerjal T, Tyler-Smith C,
et al.: A short tandem repeat-based phylogeny for the human Y
chromosome. Am J Hum Genet 2000, 67:182-196.
35. Su B, Xiao J, Underhill P, Deka R, Zhang W, Akey J, et al.:
Y-chromosome evidence for a northward migration of modern
humans into Eastern Asia during the last Ice Age. Am J Hum
Genet 1999, 65:1718-1724.
36. Qian Y, Qian B, Su B, Yu J, Ke Y, Chu Z, et al.: Multiple origins of
Tibetan Y chromosomes. Hum Genet 2000, 106:453-454.
37. Underhill PA, Jin L, Lin AA, Mehdi SQ, Jenkins T, Vollrath D, et al.:
Detection of numerous Y chromosome biallelic polymorphisms
by denaturing high-performance liquid chromatography. Genome
Res 1997, 7:996-1005.
38. Flint J, Boyce AJ, Martinson JJ, Clegg JB: Population bottlenecks in
Polynesia revealed by minisatellites. Hum Genet 1989, 83:257-263.
39. Lum JK, Cann RL, Martinson JJ, Jorde LB: Mitochondrial and nuclear
genetic relationships among Pacific Island and Asian populations.
Am J Hum Genet 1998, 63:613-624.
40. Su B, Jin L, Underhill P, Martinson J, Saha N, McGarvey ST, et al.:
Polynesian origins: insights from the Y chromosome. Proc Natl
Acad Sci USA 2000, 97:8225-8228.
41. Mack SJ, Bugawan TL, Moonsamy PV, Erlich JA, Trachtenberg EA,
Paik YK, et al.: Evolution of Pacific/Asian populations inferred
from HLA class II allele frequency distributions. Tissue Antigens
2000, 55:383-400.
42. Zimdahl H, Schiefenhovel W, Kayser M, Roewer L, Nagy M: Towards
understanding the origin and dispersal of Austronesians in the
Solomon Sea: HLA class II polymorphism in eight distinct
populations of Asia-Oceania. Eur J Immunogenet 1999, 26:405-416.
1246
Current Biology Vol 10 No 20
43. Tsuji K, Aizawa M, Sazazuki T: HLA 1991: Proceedings of the 11th
International Histocompatibility Workshop and Conference. Oxford:
Oxford University Press; 1992.
44. Gao X, Serjeantson SW: Analysis of 2600 HLA-DR, DQ haplotypes
in Asia-Oceania. In HLA 1991: Proceedings of the 11th International
Histocompatibility Workshop and Conference. Edited by Tsuji K,
Aizawan M, Sasazuki T. Oxford: Oxford University Press; 1992:229-232.
45. Hashimoto M, Kinoshita T, Yamasaki M, Tanaka H, Imanishi T, Ihara H,
et al.: Gene frequencies and haplotypic associations within the
HLA region in 916 unrelated Japanese individuals. Tissue Antigens
1994, 44:166-173.
46. Gao X, Bhatia K, Trent RJ, Serjeantson SW: HLA-DR, DQ nucleotide
sequence polymorphisms in five Melanesian populations. Tissue
Antigens 1992, 40:31-37.
47. Perez-Lezaun A, Calafell F, Comas D, Mateu E, Bosch E, MartinezArias R, et al.: Sex-specific migration patterns in central Asian
populations, revealed by analysis of Y-chromosome short tandem
repeats and mtDNA. Am J Hum Genet 1999, 65:208-219.
48. Lum JK, Rickards O, Ching C, Cann RL: Polynesian mitochondrial
DNAs reveal three deep maternal lineage clusters. Hum Biol
1994, 66:567-590.
49. Cann RL: Talking trees tell tales. Nature 2000, 405:1008-1009.
50. Gray RD, Jordan FM: Language trees support the express-train
sequence of Austronesian expansion. Nature 2000, 405:1052-1055.
51. Stoneking M, Jorde LB, Bhatia K, Wilson AC: Geographic variation
in human mitochondrial DNA from Papua New Guinea. Genetics
1990, 124:717-733.
52. Redd AJ, Stoneking M: Peopling of Sahul: mtDNA variation in
aboriginal Australian and Papua New Guinea populations. Am J
Hum Genet 1999, 65:808-828.
53. Kayser M, Caglià A, Corach D, Fretwell N, Gehrig C, Graziosi G,
et al.: Evaluation of Y chromosomal STRs: a multicenter study. Int
J Legal Med 1997, 110:125-133,141-149.
54. Redd AJ, Clifford SL, Stoneking M: Multiplex DNA typing of shorttandem-repeat loci on the Y chromosome. Biol Chem 1997,
378:923-927.
55. Schneider S, Roessli D, Excoffier L: Arlequin ver 2.000: a Software
for Population Genetics Data Analysis. University of Geneva,
Switzerland: Genetics and Biometry Laboratory; 2000.
56. Bandelt H-J, Forster P, Rohl A: Median-joining networks for
inferring intraspecific phylogenies. Mol Biol Evol 1999, 16:37-48.
57. Kayser M, Roewer L, Hedman M, Henke L, Henke J, Brauer S, et al.:
Characteristics and frequency of germline mutations at
microsatellite loci from the human Y chromosome, as revealed by
direct observation in father/son pairs. Am J Hum Genet 2000,
66:1580-1588.
58. Wilson IJ, Balding DJ: Genealogical inference from microsatellite
data. Genetics 1998, 150:499-510.
Because Current Biology operates a ‘Continuous Publication
System’ for Research Papers, this paper has been published
on the internet before being printed. The paper can be
accessed from http://biomednet.com/cbiology/cub — for
further information, see the explanation on the contents page.