Abstract and Figures

The Solutrean hypothesis for the origin of the Clovis archaeological culture contends that people came from south-western Europe to North America during the Last Glacial Maximum. This hypothesis has received numerous critiques, but little objective testing, either of cultural or genetic evidence. We contest the assertion that there is NO genetic evidence to support this hypothesis, and detail the published evidence, consistent with a pre-Columbian western Eurasian origin for some founding genetic markers, specifically mtDNA X2a, and some autosomal influence, found in ancient and modern Native American populations. The possibility that the inferred pre-Columbian western autosomal influence came more directly than through Siberia is not even considered in such studies. The mtDNA X2a evidence is more consistent with the Atlantic route and dates suggested by the Solutrean hypothesis and is more parsimonious than the assumption of a single Beringian entry, that assumes retrograde extinction of X in East Eurasia.
Content may be subject to copyright.
This article was downloaded by: [Smithsonian Institution Libraries]
On: 02 November 2014, At: 10:21
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
World Archaeology
Publication details, including instructions for authors and subscription
Solutrean hypothesis: genetics, the
mammoth in the room
Stephen Oppenheimera, Bruce Bradleyb & Dennis Stanfordc
a University of Oxford, Institute of Social and Cultural Anthropology
b Department of Archaeology, University of Exeter
c Department of Anthropology, Smithsonian Institution
Published online: 31 Oct 2014.
To cite this article: Stephen Oppenheimer, Bruce Bradley & Dennis Stanford (2014) Solutrean hypothesis:
genetics, the mammoth in the room, World Archaeology, 46:5, 752-774, DOI: 10.1080/00438243.2014.966273
To link to this article: http://dx.doi.org/10.1080/00438243.2014.966273
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our
licensors make no representations or warranties whatsoever as to the accuracy, completeness, or
suitability for any purpose of the Content. Any opinions and views expressed in this publication
are the opinions and views of the authors, and are not the views of or endorsed by Taylor &
Francis. The accuracy of the Content should not be relied upon and should be independently
verified with primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities
whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use
can be found at http://www.tandfonline.com/page/terms-and-conditions
Solutrean hypothesis: genetics, the
mammoth in the room
Stephen Oppenheimer, Bruce Bradley
and Dennis Stanford
The Solutrean hypothesis for the origin of the Clovis archaeological culture contends that people came
from south-western Europe to North America during the Last Glacial Maximum. This hypothesis has
received numerous critiques, but little objective testing, either of cultural or genetic evidence. We contest
the assertion that there is NO genetic evidence to support this hypothesis, and detail the published
evidence, consistent with a pre-Columbian western Eurasian origin for some founding genetic markers,
specically mtDNA X2a, and some autosomal inuence, found in ancient and modern Native American
populations. The possibility that the inferred pre-Columbian western autosomal inuence came more
directly than through Siberia is not even considered in such studies. The mtDNA X2a evidence is more
consistent with the Atlantic route and dates suggested by the Solutrean hypothesis and is more parsimo-
nious than the assumption of a single Beringian entry, that assumes retrograde extinction of X in East
Archaeogenetics; migrations; Solutrean hypothesis; Palaeo-Americans; mtDNA; aDNA.
In the past decade there has been a signicant shift in our general understanding of how, when
and by whom the Americas may have rst been settled. We have gone from a relatively simple
model of a group of Siberian hunters following large animals into the North American continent
and marking their presence with a distinctive set of artefacts and behaviours (Clovis First) to the
proposition that there were earlier peoples that may have come from different places at different
times (Collins et al. 2013). That there were people in the Americas before Clovis (13,500
12,800 BP), is now generally accepted as proven (e.g., Gilbert et al. 2008). What is still very
contentious is who these people were and where they came from. There is still a tendency to
World Archaeology Vol. 46(5): 752774 Debates
© 2014 Taylor & Francis ISSN 0043-8243 print/1470-1375 online
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
emphasize a one route entry model, specically from Northeast Asia (single or multiple, coastal
and/or inland). Yet these hypotheses have yet to be well articulated. An additional route has long
been proposed, which suggests that Pleistocene people came to the Americas not only from
north-eastern Asia but also across the Atlantic from south-western Europe during the last glacial
maximum circa 22,00017,000 BP (Bradley and Stanford 2004; Stanford and Bradley 2012).
While this proposition has been clearly presented as a hypothesis (Stanford and Bradley 2012),
it has elicited strong opposition, based on some specic critiques, both of the ideas and evidence
(Straus, Meltzer, and Goebel 2005; Eren et al 2013). There have been four main objections: (1)
the distance between the proposed origin point (the region where the Pyrenees meet the Bay of
Bisque) and the entry point in what is the now submerged Grand Banks off of Newfoundland;
(2) the perceived difculty of making a North Atlantic crossing during the LGM; (3) the time
gap between the originating archaeological culture (Solutrean) and Clovis; and (4) the claimed
lack of genetic evidence of this proposed founding population. There have of course been other
challenges to the model, such as the suggested dissimilarities between the artefact assemblages
(Straus, Meltzer, and Goebel 2005), etc. Some of these are published debates and will rightly
continue to be points of discussion. This article is specically focused on the issue of the genetic
evidence, based mainly on the long-standing, mitochondrial evidence for the anomalous phy-
logeography of Native American haplogroup X2a and its most recent common ancestors in West
Eurasia and North Africa (data in Reidla et al. 2003; Fernandes et al. 2012). This evidence is
phylogeographically most-parsimonious with a specic additional Pleistocene West Eurasian
entry to Northeast America.
But we also discuss recent analyses, here, that we feel have either: (1) ignored that possibility
(e.g., Reich et al. 2012); or, (2) led to over-interpretation of evidence to reject it (Fagundes et al.
2008; Hooshiar Kashani et al. 2012; Rasmussen et al. 2014), or (3) have interpreted their own/
new autosomal evidence for pre-Columbian, West Eurasian inuence in the Americas, assuming
it came in via Beringia, without formally testing that route against the alternative Atlantic route
(Rasmussen et al. 2014; Raghaven et al. 2014; Patterson et al. 2012).
Archaeological discussion
Recent archaeo-genetic developments discussed here include genetic-inference claims of three
Late Pleistocene human remains (two skeletal and one faecal) in Siberia (Malta), Oregon
(Paisley Cave) and Montana (Anzick Clovis Cache). While our focus is on the DNA results,
a few comments about their archaeological contexts are in order. The Malta site (Raghavan
et al. 2014) has been studied off and on since the 1920s and, based on the dates and artefact
assemblages, especially the mobiliary art, it may be interpreted as related to the Gravettian,
which spread from West into Central Eurasia, west of Lake Baikal. That the aDNA results from
the single Malta individual in Central Eurasia show some connections to West Eurasia, in
having a basal U mtDNA haplotype and a Y haplotype basal to R is therefore of no particular
surprise. That there are also autosomal traits shared with Native Americans is also not unex-
pected since Malta is indeed in East Eurasia (Fig. 1). However, the Malta remains, at ~24 ka,
are geographically closer to the origins of the Gravettian in Eastern Europe than they are to the
Bering Strait (Fig. 1). That the Malta individual has something important to contribute to the
debate is without question, but the interpretation that the pre-Columbian West Eurasian
Solutrean hypothesis: genetics 753
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
admixture (14 to 38 per cent) into Native Americans (Raghavan et al. 2014) that they estimated,
could only have entered the Americas via Beringia rather than by some other route, e.g., trans-
Atlantic, seems unsubstantiated. The latter route was not mentioned let alone tested/excluded as
a model in that analysis. In any case, there is no established Western Eurasian archaeological
trail east of Malta, and into Alaska (as implicit in the archaeo-genetic Upper Palaeolithic route
claim in the Malta paper), early enough or archaeologically congruent enough to be ancestral to
the older-than-Clovis cultural remains in eastern North America (Collins et al. 2013). The other
site mentioned (Afontova Gora) is even further west than Malta and Soviet researchers
interpreted that if anything the directionality indicated by Malta to Afontova Gora is to the
northwest not toward Beringia.
The mtDNA results (A2 & B2) of the analysis of older-than-Clovis (circa 14,500 BP)
reported human coprolites from Paisley Cave in Oregon clearly show East Asian afnities
(Gilbert et al. 2008) consistent with re-calibrated estimates for the ages of those branches
(Soares et al. 2009). This is also supported by the non-Clovis archaeological materials that
have been found associated with them. How this may or may not relate to the origin(s) of Clovis
remains to be debated and cannot be asserted.
Finally, we have some real concerns about how the Anzick Cache results (Rasmussen et al. 2014)
have been used to support a single western entry, and more specically how the genome analysis
has been used to decide that the study refutes the possibility that Clovis originated via a European
(Solutrean) migration to the Americas(Rasmussen et al. 2014, 228). We are not challenging the
aDNA results but we do have concerns about the way they have been over-interpreted towards one
single migration model (see below) and the condent representation of the site archaeology and
stratigraphy. We question the interpretation that the analysed human remains were found in direct
associationwith Clovis artefacts (Rasmussen et al. 2014, 225). We have two concerns. First, the
association was not conrmed by controlled archaeological excavation and recording. A subsequent
archaeological investigation of the site did not conrm the association (Taylor 1969,149).Whilea
later site examination also found no intact deposits, and other arguments were used to assert an
association (Owsley et al. 2001,11718). Second, the directly dated human remains (14c cal
average12,70712,556 BP) and Clovis antler artefacts (14c cal average13,03912,763 BP)
(Rasmussen et al. 2014) yielded non-overlapping dates (at 2 sigma [95.4% probability]) indicating
that the artefacts are older than the person. At a minimum there is a 56-year gap but it could be as
much as 483 years. To simply say that the age difference was because the artefacts were heirlooms is
unsupported speculation and assumes that the 56-year gap is the best estimate. However, even if a
direct association is accepted, this still doesnt demonstrate that the childs genetic makeup was
specically indicative of the origins of Clovis culture or inform on the geographic origins of Clovis
and cannot be used to refute the Solutrean hypothesis (see below).
Genetics and culture
Archaeological literature critiquing the Solutrean hypothesis (SH) has come most vociferously from
Clovis-Firsters (e.g., see Fiedel and Morrow 2012,382;Morrow2014). Surprisingly, in view of this
and the mitochondrial X2a geographic anomaly (data in Reidla et al. 2003; Fernandes et al. 2012),
mentioned above, some recent autosomal studies condently ignore or dismiss the SH route on
summary evidence. For instance, one recent autosomal-based publication (Reich et al. 2012),
754 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
reconstructing the genetic prehistory of the rst Americans from modern populations, relies on a
solely-Beringian ancestry route as a given and ignores a possible supplementary ancestral popula-
tion-source route for their four selected First Nations groups in Canada, who collectively form a
separate phylogenetic branch from other mainland Native Americans they analysed (Reich et al.
2012,Fig 1c). Another autosomal study on a ~13 ka aDNA sample (Rasmussen et al. 2014)in
western Montana claims our genome analysis refutes the possibility that Clovis originated via a
European (Solutrean) migration to the Americas(Rasmussen et al. 2014, 228), but omits to test for
trans-Atlantic sources of West Eurasian Pleistocene gene ow to the Americas, which is inferred in
their own analysis (see below).
Migrating groups of people carry their culture, language and genes, but rarely do these markers
survive in the same proportions unless, perchance, they were the only settlers. All three proxy
records of the past leave some record in the soil and/or among descendant societies, in different
strengths, accuracy and precision of dating. So, it is essential that each proxy should be explored and
dated independently before inferences on interdisciplinary congruence are made (Oppenheimer
2004). Further, while predictions from hypotheses derived from one proxy can, without prejudice,
be tested for independent support from another, items of evidence selected from one, cannot simply
be used to disproveor, alternatively, patch-up narrative deciencies in the other. Most importantly
genetics should not be used, as in the past, to nd support for the current most inuential
archaeological paradigm; rather it should objectively test all testable migration-models.
Material culture and languages can change dramatically with small intrusions of new people.
Linguists even have a name for it: language switch (Vasek 1983). Genes also mix, but leave
long-lasting records, both in individuals and in populations. The importance of the individual
lineage record applies particularly to non-recombining uniparental lineages, which give an
uncorrupted record of their own history, have phylo-geographic continuity and persistence
and, unlike culture, offer direct proxy records of discrete prehistoric population migrations.
Uniparental lineages (mtDNA & NRY) each have one quarter the effective population size of
any autosomal locus. Their consequent greater tendency to local extinction than autosomes
means that modern persistence and lack of recombination, increases rather than decreases the
phylogeographic specicity of surviving founding lineages and their descendants.
So, multidisciplinary reconstruction, requires a separation and enforced independence of each
proxy dating for the past, in order to avoid self-referential circularity of arguments
(Oppenheimer 2004). Advocates of the SH accept the mitochondrial evidence that the majority
(4+/5) of founding lineages (i.e., AD, with the exception of X2a) arrived in Alaska from
Beringia at least before the Holocene, i.e., allowing plural (e.g., both east and west) Palaeolithic
routes and entries with subsequent genetic and cultural admixture. The possibility of multiple
colonizations unfortunately increases that risk of confusion. The implications of this, in deter-
mining which ancestors brought which early tool traditions and who ended up using them, are
more complicated than testing a single-entry scenario. Perhaps unsurprisingly, critics of the SH
often implicitly couch their dispute in terms of either Beringia/or SH.
Anzick study: a cultural or genetic refutation of the Solutrean hypothesis?
The risks of relying on one-to-one direct links between material culture and genes can, perhaps
best, be illustrated with reference to a recent, important archaeo-genetic publication in Nature
Solutrean hypothesis: genetics 755
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
(Rasmussen et al. 2014), which offers an extraordinary genetic window into the Pleistocene,
only ~1,500 years after the current earliest evidence of human occupation in the American
Northwest. Implicit conations of the culture-genes bond linked-with the either/or argument are
introduced in the introductory abstract (Rasmussen et al. 2014, 225):
the people who manufactured Clovis tools remain under debate. It is generally believed
that these people ultimately derived from Asia and were directly related to contemporary
Native Americans. An alternative, Solutrean, hypothesis posits that the Clovis predecessors
emigrated from southwestern Europe during the Last Glacial Maximum.
The authors did extensive genome sequencing on a male infant, deposited approximately 12,680
years ago in proximity of Clovis artefacts at the Anzick site in western Montana. They inferred
(Rasmussen et al. 2014, 227) that the Anzick-1 sample is genetically more closely related to
Central and South Americans than to any other populations.
They later argue further (Rasmussen et al. 2014, 228) that the infant
is closely related to all indigenous American populations. As such, contemporary Native
Americans are effectively direct descendants of the people who made and used Clovis tools
and buried this child. In agreement with previous archaeological and genetic studies, our
genome analysis refutes the possibility that Clovis originated via a European (Solutrean)
migration to the Americas.
The rst of these claims conrms a reliance on the validity of claiming a direct bond of descent
between material culture and genes as proof, which is weak in the context of their refutation,
both generally and specically, in this case. For the claim to hold more water, would indeed
require only a single-source American entry, which is duly asserted a few sentences later, or if
plural-entries, that Clovis was the earliest culture and this site was near the founding date,
neither of which apply. The former, singular, assertion is unproven and is not conrmed by the
combination of diverse East Eurasian mtDNA founders and West Eurasian X2a (see below). For
the latter, i.e., with plural entries, the site would have to be coincident with the earliest directly
dated settlement of northwest America or at least with the earliest nding of Clovis tools,
neither of which it was.
By their own estimate (Rasmussen et al. 2014), the Anzick child was actually around 1,500
years younger than the rst conrmed American human remains (Gilbert et al. 2008;Fig. 2)
showing directly dated human coprolites from Paisley Cave in Oregon, over 14,000 years old
and pre-dating Clovis technology by over 1,000 years, thus directly refuting the old Clovis First
claim. The coprolite paper shows various Clovis dates tightly clustering around 13,000 kya
(Gilbert et al. 2008;Fig. 2), the oldest antedating the Anzick human remains (average 12,680
kya) by 320 years, the latter post-dating the coprolites by around 1,500 years.
The dates of appearance of Clovis points across Northern USA (Waters and Stafford 2007)
are also surprisingly synchronized, so there is no clear West-East or East-West direction. If, as
the SH argues, East Coast Atlantic settlement occurred before the evolution of hallmark Clovis
technology and Clovis spread material culture from East to West (but not necessarily with many
people see Bradley and Collins 2013), it would have reached pre-existing settlements in
Montana well before the Anzick child was born. At that time people in the West were indeed
756 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
likely to be largely of Beringian origin, consistent with Anzick also inferred from lower
modern-western X rates in mtDNA studies (Fig. 1; see also, e.g., Malhi et al. 2008,Fig. 2;
Perego et al. 2009, Fig. 3). However, by analogy, guns and horses spread from the East early in
the colonial period, but fairly soon they were not just used by Europeans; so the genes of the
nal owner could not, therefore, identify the ultimate origins of the animal or artefact type.
Concerning the second claim above –‘our genome analysis refutes the possibility that Clovis
originated via a European (Solutrean) migration to the Americas’–it is difcult to see which
part of their genome analysis specically tests, let alone refutes the SH, given the complete
absence of any further mention of the SH, or any supplementary Atlantic route, elsewhere in the
paper and online material. Furthermore, as discussed in more detail below, their own study
actually nds evidence of pre-Columbian West Eurasian admixture.
Separating genetics and culture and testing model predictions
Given the risks of reading too much into associations of genes and material culture, when
cultural and genetic admixture resulting from putative multiple entries both tend to confound
inferences and conclusions, it is important to ask what unique, uncontaminated biological
insight genetics has to offer, for what is largely an archaeological-cultural argument about
geographic sources and dates of migration. To focus on this, some details of the archaeological
discourse need to be put to one side, as opaque to genetic enquiry. These include technological
sources, lithic evolution and chronological gaps and the question of whether independent
evolution, from any convincing ancestral Pre-Clovis source point style in northeast Asia, trumps
a phylogeny connecting it to Solutrean technology at the LGM on the other side of the Atlantic.
Having done that trimming, the only clear relevant directional message that currently remains
is genetic and even stronger with complete sequencing and recent data (Reidla et al. 2003;
Fernandes et al. 2012), namely that one of the ve oldest Pleistocene founding American
mitochondrial lineages, X2a, is unlikely to have come from East Eurasia on phylogeographic
parsimony grounds, with the only possible alternatives being West Eurasia and North Africa.
Currently, the SH is the only archaeological model which predicts or offers any credible
geographic parallel for this directional anomaly which has remained as the elephant (or
mammoth) in the room for over sixteen years.
Testing predictions of hypotheses is a useful tool of the scientic method. Ignoring anom-
alous evidence, not testing it objectively, rather retesting current dogma, lacks balance.
American X2a does not of course provethe specics of the SH, but does oblige geneticists
looking at other parts of the genome, to test available hypotheses to explain the substantial
presence of pre-Columbian West Eurasian genetic signatures found in Native American popula-
tions, rather than assuming they all arrived from Beringia (e.g., Reich et al. 2012).
The key genetic questions and predictions arising from the SH are: (1) is there non-cultural
evidence for a Palaeolithic migration from the western Old World across the Atlantic to the
Americas; (2) if so, from where and (3) when? As mentioned above, non-recombining unipar-
ental genetic lineages (mitochondrial DNA and the Y chromosome), do each have an uncor-
rupted, cumulative record of their own prehistoric origins. Unlike material culture, and the rest
of our vast genome, they preserve direct, unambiguous proxy-source sequence-records of
discrete prehistoric population migrations. For, unlike the rest of the genome, they do not
Solutrean hypothesis: genetics 757
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
recombine (the genome-wide splicing and dicing which happens after every ovum fertilization,
when different populations meet), and each chromosome retains its own ancestral record intact.
New, random, non-lethal mutations on these chromosomes occurring over thousands of years
allow a detailed lineage tree to be reconstructed, on which fresh mutations are geographically
highly-specic, allowing the migration of new branches (mutations) to be inferred phylogeo-
graphically, and dated, using formal methods of founder analysis (Avise 2000; Richards et al.
2000). It is important to realize that while uniparental markers can only partly inform on
population admixture history and relative size, the distribution of their branches in time and
space are specic markers of migration in their own right.
Mitochondrial DNA evidence
That modern uniparental lineage branch-distributions have far greater regional and interconti-
nental geographic specicity than individual autosomal loci has been amply demonstrated over
the past twenty-seven years. Evidence for the modern human out-of-Africa migration was rst
demonstrated unambiguously in 1987 using mitochondrial DNA phylogeography (Cann,
Stoneking, and Wilson 1987). It was later conrmed to be a single migration, with all non-
Africans belonging to one founding L3African lineage (haplogroup) entirely conned within
its two non-African branches M & N, while all the rest of the L tree haplogroups remain in
Africa (reviewed in Oppenheimer 2003,2012).
For the handful of American founding mitochondrial lineages it could be nearly that
simple. It has been agreed for twenty-one years that the four main Palaeolithic founding
lineages in the Americas are derived from four ancient haplogroups, which are well
represented in East Eurasia generically as A, B, C and D (Torroni, Schurr, et al. 1993;
Torroni, Sukernik, et al. 1993). These four haplogoups clearly evolved in East Eurasia long
before the LGM and are absent in West Eurasia (Soares et al. 2009, Fig. 6). Furthermore,
with a couple of minor back-ow-exceptions across the Bering Strait, the American deri-
vative-branches of these four haplogroups are only found in the Americas, though with
relatively different geographic distributions. The ve deepest branches (specically A2, B2,
C1a&b & D1) date genetically in America around 14-kya (Soares et al. 2009,Fig.6),
consistent with the dates of the 14-kya coprolites from Oregon which, neatly, feature A2 and
B2 aDNA haplotypes (Gilbert et al. 2008).
So much is common ground, but there is a fth Palaeolithic American haplogroup: X2a,
derived from West Eurasian X(Brown et al. 1998; Reidla et al. 2003). This uniquely-
American founder lineage X2a, is a sub-clade of X2-225 (Fernandes et al. 2012), and is of
Pleistocene age: 14,080 yr; (10,32117,914) by Maximum Likelihood (ML) and 21,289 yr
(11,04032,035) by Rho, on complete mtDNA sequences (Fernandes et al. 2012). Although
these estimates are clearly Pleistocene, with one of possibly greater local antiquity than those of
haplogroups A-D in America, X2a has a completely different ancestral source and distribution,
originating near the Mediterranean region, and is near-absent from East Eurasia (Figs 1 and 2:
see also Reidla et al. 2003; Fernandes et al. 2012).
The X haplogroup is clearly West Eurasian in origin, as is X2-225 (Figs 1 and 2). Along
with two other early relict primary branches of ex-African macro-group N (N1 & N2), X
most likely originated within the Arabian Peninsula(Fernandes et al. 2012, 347) close to
758 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
the out-of-Africa movement to Southwest Asia, and evidently never made it past the Indus
Valley towards East Eurasia (Fernandes et al. 2012). The largest, but not oldest, X branch,
X2, most likely originated in the eastern Mediterranean region around 22 kya, rapidly
expanding in diversity and geographic distribution at that time (Fernandes et al. 2012).
X2 has many primary branches distributed throughout West Eurasia, most sharing a further
mutation at np 225 (Fig. 2; Fernandes et al. 2012). The unique phylogeography and
Palaeolithic founding age of X2a in America, means it physically separated near to the
founding age of its parent X2aj around the LGM in West Eurasia. Three large X2-225
branches (X2e, X2ajandX2bd) also expanded at that time (2218 kya) and spread mainly
west. Confusingly, np 225A subsequently reverted independently in both sub-clades X2e2
andX2a1(Fernandesetal.2012:Fig. 1), but this does not of course alter the West Eurasian
ancestry of X2a nor affect the Altai X2e2 red herring (next paragraph).
The most geographically conservative of these three branches X2e, expanded locally in
the Levant and into the Caucasus and eventually, during the historic period, to the Russian
Altai(Derenkoetal.2001,2007), where it is only represented as a single highly-derived
closely-related founding-cluster (a unique derivative of X2e2: Fig. 2)ofthreesimilar
haplotypes dating to ~2.6 kya (Fernandes et al. 2012). This recent, derived and geographi-
cally isolated clade on the south-western verge of East Eurasia, is the most easterly
representative of X and is clearly not ancestral to X2a, nonetheless is regularly promoted
as such, by SH critics (e.g., Straus et al. 2005;Bamforth2013,Morrow2014) as evidence
that the X2a ancestor travelled from the Mediterranean, through Siberia, leaving no trail,
half-way round the globe, to Beringia. Even if the phylogeny could miraculously be re-
arranged, it is wishful and misleading to argue that the Russian Altai is remotely close to
Beringia or even integral to East Eurasia, let alone that X2a and X2g went east to America
The most western European distribution among these three large X2-225 branches is seen in
X2bd (Age: 18,983 yr (14,85623,191) by ML), which spread west along both the north and
south Mediterranean coasts to Morocco, Spain and up the Atlantic coast and widely throughout
Europe, including the British Isles (e.g., Orkney). X is present in Basques, both in Mid-
Holocene aDNA (Izagirre and Rua 1999, Table 2) and X2b & X2c in modern populations. It
would of course be helpful to know if the X2-225 sub-clade was present in relevant parts of
Pleistocene Europe as well as in the Mid Holocene. However, given the small handful of Iberian
samples analysed for that period (Brandt et al. 2013; Skoglund et al. 2014), and the fact that,
although X2 is widespread in West Eurasia, even modern rates of X2 in Western Europe are low
(<3%) and heterogeneous (Fig. 1), with higher rates in the Mediterranean region, so absence of
Hg X in relevant Late Pleistocene aDNA analyses in West Eurasia cannot be taken to indicate
absence at that time. Unfortunately, little high-resolution X data is available for Iberia for any
period. There is, however, a recent report from the Franco-Cantabrian region noting a great
diversity and high frequency for subhaplogroup X2(García et al. 2011,39). In particular 6/91
samples (6.6 per cent) from the Vizcaya province in the northern Basque Country were
characterized as X2 (based on HVS-1 markers) mostly X2b & X2c, one X2b sequence having
a near HVS-1 match in Morocco (García et al. 2011, Table S1). South Portugal has an X
frequency of 3.6% (González et al. 2003), and Spain an X2 rate of 4.2% (Crespillo et al. 2000).
Orkney also represents an Atlantic X2 hotspot with 7.2% X2 (Helgason et al. 2001).
Solutrean hypothesis: genetics 759
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
Figure 1 Worldwide distribution of Mitochondrial Haplogroup X2. © Stephen Oppenheimer, Bruce Bradley and Dennis Stanford, 2014. All Rights Reserved.
Insets: (A) Across Asia hypothesis; (B) Across the Atlantic hypothesis (same scale). Collated and corrected for repeat citations from all sources cited in: Table 1,
Reidla et al. 2003; Table S4, Perego et al. 2009; Table 3, Helgason et al. 2001; Table 3, Bolnick and Smith 2007; Scozzari et al. 1997; Malhi and Smith 2002.
Anzick dates: (Rasmussen et al. 2014) and Malta dates: (Raghavan et al. 2014).
760 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
This apparently rapid coastal spread of X2bd from the Levant to the Atlantic is consistent with
the Pleistocene use of boats by modern humans there and elsewhere. The Pleistocene antiquity of
such maritime skills is well-attested archaeologically elsewhere in Oceania, using radiocarbon and
luminescence-dating (reviewed in Oppenheimer 2012), thus clearly does not preclude such spread
further along the edge of Atlantic pack ice at the LGM. For Melanesia, the rst archaeological
evidence of occupation of the island of New Guinea has been radiocarbon dated to 49 ka BP
(Summerhayes et al. 2010) and 40 ka BP to the east, in New Britain island (Leavesley et al. 2002).
For Australia, the carbon ceiling has been increased to around 48 ka using the ABOX-SC method
(Turney et al. 2001). Non-carbon dating techniques, however, suggest an earlier human colonization
of Australia by 5060 ka (Roberts et al. 1994;RobertsandJones2000;Birdetal.2002,2004). One
need not look only to Southeast Asia for evidence of early watercraft. There has been much
discussion of their use in the colonization of western Scandinavia during and after the nal glacial
retreat (Bjerck and Breivik 2012; Bonsall, Pickard, and Groom 2013).
The third, and most diverse, primary X-225 branch is X2aj(Fig. 2; 19,409 yr; 15,547
23,342, by ML), which includes X2a in the Americas and has several relict X2j branches found
in 3 Egyptians (Fernandes et al. 2012) and a single haplotype from Iran (Reidla et al. 2003). The
latter shares two further mutations with the north-African X2j cluster as well (at np sites 16,179
and 16,357), thus conrming X2aj as a solid Old World common ancestral node for the
American X2a branch with its nearest relative X2j being located in Egypt on the opposite
side of the globe from Beringia across the North Pole. To critics, a possible Most Recent
Common Ancestor for X2j and X2a in North Africa or Southwest Asia, might fail to conrm an
Iberian cast-off point for X2a en route to the Americas but, as mentioned, proving the specics
Figure 2 Schematic tree of Mitochondrial Haplogroup X.
Ages indicated are maximum likelihood estimates obtained with the complete mtDNA genome (in ka).
(Redrawn and modied with permission from Fernandes et al. 2012,Fig. 1.)
Solutrean hypothesis: genetics 761
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
of the SHis not the purpose of this discussion. Given the absolute paucity, and reality of X2j, it
simply conrms Western Eurasia/Mediterranean as the home of the X2-225 clade and its
relevant branch X2aj and certainly gives no support to a Beringian route for X2a.
In addition to the X2a branch, there is another Native American X2-225 branch represented
by a single haplotype X2gunique to the Americas and the Ojibwa Nation who also possess the
greatest X2a diversity in the Americas. In their denitive review of X origins, Fernandes et al.
(2012, 352) remark:
The rare X2g, also found only in Native Americans, indicates that the spread from the Near
East toward the Americas could have begun as early as the emergence of the X2-225 clade
(~21 ka), given that this root could have been the only founder sequence.
This glacial age of X2-225 (20,786 yr; 17,35824,269 by ML) is signicantly older than the
founding age estimates for American A-D at ~14 ka, using the same calibration method, and
thus appears to inate the potential founding date of X2 in the Americas. In our view this
statement can also be taken not to exclude the more direct and shorter route from the Near East
across the Atlantic (Fig. 1B). The hypothetical transit of the Atlantic would necessarily have
taken a shorter time than Egypt to Alaska overland.
In spite of its unique antiquity, X2a has a focused distribution in the Americas. It is limited to
North America and to groups speaking Northern Amerindlanguages in Greenberg et al.s
(1986) tri-partite American language classication, as referred to in the autosomal studies
discussed here (exceptions being Navajo and Apache, both southern Athabaskan (Na-Dene)
and Pueblo Jemez who speak a Central Amerind language). X2a has not otherwise been found
in Greenbergs other two northern language macro-groups Na-Dene or Eskimo-Aleut. More
specically, among dened First Nations groups X2a is found at highest frequency (Fig. 1),
branch antiquity and diversity (Fernandes et al. 2012, source data in Fig. S1 and SOM) in
Northeast America. The highest frequencies and lineage diversity are among the Algonquian-
speaking (Northern Amerind) First Nations of the Great Lakes region (Fernandes et al. 2012:
tribal afliations sourced from Fig. S1, tree). Three such groups (Wisconsin Chippewa, and N.
& S. Ojibwa) have X2a frequencies between 25 to 30 per cent, while another three Algonquian-
speaking groups range from 4 to 7 per cent (Perego et al. 2009, Table S4). Indeed the two
primary branch nodes of American X2a (X2a1 ~8.5 ka & X2a2 ~4.4 ka) and also X2g (Fig. 2),
are predominantly represented among Algonquian-speakers, as are 7/10 tertiary branches of X2a
(data in Fernandez et al. 2012, Tree S1). Pre-Columbian aDNA evidence for X2a1a has also
been found in the same modern linguistic region (in Illinois: Norris Farm in Fig. 2) at 4 per cent
(Stone and Stoneking 1998). X was also tested for in six other archaeological site aDNA
collections (data in Mahli et al. 2004; Bolnick and Smith 2007), dating between 200 BP and
2,150 BP in sufcient numbers to make frequency estimates, including a total of three pre-
Columbian sites with X present at frequencies from 3.9 to 14.3 per cent (Fig. 1).
Eastern Canada and the Great Lakes region thus currently appear to have the highest regional
frequencies, modern branch-diversity and aDNA support, consistent with an eastern primary
region of X2a dispersal. The same cannot be said for the Northwest of the continent (Fig. 1).
The only other modern Native American groups that have signicant frequencies of X2a are
three other widely-separated, culturally-unrelated, phylogenetically-derivative clusters (see
below) to the west and south (Fig. 1), in order of proximity to the Great Lakes: (1) Nearer
762 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
to the southwest of the Great Lakes: Sioux (Siouan: 7.3 per cent), (2) Far west Vancouver
Island and Washington State: Yakima (Sahaptin: 4.8 per cent) and Nuu-Chah-Nulth (Wakashan:
8.8 per cent); Pre-Columbian aDNA evidence for X, from 5001,500 BP, has also been found at
Vantage in Washington State (1/7; Malhi and Smith 2002; Malhi 2004), (3) Southwestern USA
Navajo (3 per cent) and Apache (0.5 per cent) (both southern Athabaskan languages) and
Jemez (Pueblo Kiowa-Tanoan: 10 per cent).
Extensive surveys of X2a traces found in the general mixed populations of North American
States (Perego et al. 2008, Fig. 3, lower left; Table S3), excluding autochthonous Native
American groups, show a distinctive distribution, mostly very low rates (3 per cent), with
the highest rates (1.742.52 per cent) focused in a moderate but distinct shift-to-the-west from
the Great Lakes region to formerly Siouan-speaking regions. This shift centers on North Dakota-
South Dakota, Montana-Wyoming with lower rates in the mainly Algonquian-speaking
Canadian Prairie Provinces of Manitoba, Saskatchewan and Alberta (1.80 per cent). X2a is
also found in former Apache territory (Oklahoma: 0.68 per cent). Exceptions to this non-tribal
western-shift pattern are, perhaps unsurprisingly, found in other north-eastern, formerly
Algonquian-speaking areas (Minnesota 0.93 per cent; Wisconsin 0.52 per cent; Pennsylvania
1 report: Perego et al. 2008, Fig. 3, lower left; Table S3).
Complete sequence phylogeny (Fernandes et al. 2012, Fig. S1) of X2a shows that identied
haplotypes from Native Americans of the three non-Algonquian-speaking regions (above),
belong mainly to just one of the seven subgroups of X2a1 (namely X2a1a, age: 7,689 yr;
4,83510,592 by ML). The other six X2a1 subgroups all include Algonquian-speakers as the
majority, while X2a1b has 2/8 haplotypes located away from the Great Lakes (Pennsylvania and
Montana) and X2a2 has 1/4 types from Navajo (Fernandes et al. 2012, Fig. S1). Both of the two
X2a1a embedded subgroups include Sioux, while one of these (X2a1a1 ~4.3 ka) has Jemez and
Nuu-Chah-Nulth on one branch and Sioux on the other (the same Washington cluster from Nuu-
Chah-Nulth in Vancouver Island, inferred as X2a1a1 from HVS sequence/RFLP in Brown et al.
1998). A south-eastern United States aDNA dispersal location is Florida, where pre-Columbian
(~ 8 ka) aDNA evidence for X (inferred from sequence as X2a1a1*) has also been found (Smith
et al. 1999), consistent with the implied Early Holocene date of expansion of X2a1a.
As mentioned, the Great Lakes Algonquian-speaking region encompasses the highest frequen-
cies and the oldest and deepest X2a (and X2g) branch-ancestry. The default phylogeographic
inference would therefore be that the derived cluster X2a1a, and its haplotypes currently found in
the three geographically-dispersed Native American Groups, ultimately derives from an ancient
dispersal event towards the south and west of the Great Lakes, in the early Holocene, agged by
X2a1a, best represented among Sioux. However, more sequences would be required to explore
this inference. There is, however, no phylogeographic support for a west-to-east spread of X2a.
This new evidence-based perspective that the region of greatest antiquity of American X2a
appears to be in northeast Canada and that the three other younger more-derived X2a regions in
North America may have been colonised by more recent Early Holocene dispersals to the south,
east and west, suggests an ultimately eastern introduction of X2a and X2g to the Americas
consistent with an additional trans-Atlantic migration suggested by archaeological evidence
(Stanford and Bradley 2012).
Since the Lakes region would have been largely covered by periglacial lakes adjoining the
Laurentide Ice sheet before the Holocene, this raises the logical questions of where the putative
rst east coast settlements most likely occurred and which rivers they might have used to
Solutrean hypothesis: genetics 763
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
migrate to the Great Lakes area. The St Laurence River is certainly an access possibility and
X2a (X2a1 as inferred from HVS sequence in Smith et al. 1999), has been found in 50 per cent
(3/6) of a small sample of Algonquian-speaking Mikmaqs who live around its estuary (Malhi,
Schultz, and Smith 2001). Moreover X2a1 has been identied in aDNA from the extinct
Beothuk of nearby Newfoundland (Kuch et al. 2007).
Finally, there are other European mtDNA lineages among rst Nations, but they have been
ignored, routinely left out of reports and certainly not studied (Malhi, Schultz, and Smith 2001,
23). One of the earlier sources (Scozzari et al. 1997), however, surveyed Ojibwa living on
Manitoulin Island and reported an X frequency of 25.7 per cent and an H frequency of 5.7 per
cent, which was assumed to be of recent European maternal origin.
Arguments made for X2a arriving via Beringia
It is difcult to nd formal arguments in favour of X2a arriving via Beringia, rather than
summary-dismissive contra arguments to the Atlantic source of X2a. Two recent genetics papers
have argued in favour of X2a arriving via Beringia with the other four haplogroups (A-D).
One of these (Fagundes et al. 2008), claims to refute the SH on the basis of theoretical
modelling of a single entry because, on their estimates, American X has a similar age and
diversity to the other American mtDNA founders. They argue it could have gone through a
similar LGM bottleneck to the other lineages in Beringia and thus claim their
results strongly support the hypothesis that haplogroup X, together with the other four main
mtDNA haplogroups, was part of the gene pool of a single Native American founding popula-
tion; therefore they do not support models that propose haplogroup-independent migrations, such
as the migration from Europe posed by the Solutrean hypothesis. (Fagundes et al. 2008,589)
No other support for a shared (X2a + A-D) migration history is given, and no comparative likelihood
tests of alternative haplogroup-independent migrationsare offered. The fact that the LGM would
have had similar effects on groups arriving on the east coast of America, with similar date constraints
is not mentioned. Twelve of the eighty-six mtDNA complete genomes used for age estimates were
X2a, compared with the twenty-four used in Fernandes et al. (2012). In fact their overall estimate for
the transition period within Beringia/America was fairly broad at 5,000 years:
As for the absence of X in East Eurasia, the reverse of the Old World A-D distribution which
is exclusively found there, this is surprisingly not seen as a parsimony constraint, simply
asserted, with no other discussion, as the expected result of drift:
It is likely that this haplogroup is absent in eastern Siberian populations because of drift
effects, which impact rare variants more strongly. Thus, its probability of being lost through
random effects would be high. (Fagundes et al. 2008, 589)
Drift and extinction of X2a & X2g ancestry in East Eurasia (Fig. 1) cannot simply be invoked as
highly likely (unlike A-D), in this way, without considering phylogeographic parsimony cost to
their underlying hypothesis. If X2 was so liable to extinction to start with, how did it manage to
cross Eurasia, with multiple bottlenecks and still achieve substantial frequency in Northeast
764 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
America? No evidence or similar example is given from the mtDNA phylogeographic literature
for the assumption of high probability of retrograde land-based lineage extinction nearly half
way round the world. It would be difcult to nd one. The cost in phylogeographic parsimony
to their core hypothesis is not considered, in comparison to a transatlantic founder migration of
X2a and the single X2g haplotype, both of which have the same MRCA in the West Eurasian/
Mediterranean region and obviously have no need to invoke retrograde drift-extinction in the
Atlantic (inset A vs B: Fig. 1).
In spite of such theoretical claims, long-distance uniparental lineage migration nearly always
leaves, not only a trail, but progressive, geographically-dened mutational branch markers. One of
the best examples of such trail-persistence, can be found in the progression of B4a1a from SE
Asia through the Melanesian islands to Eastern Polynesia in the Pacic. In spite of serial founding
effects and drift a progressive genetic trail can be traced along the island chain evolving through
B4a1a1, to B4a1a1a to B4a1a1a1. In spite of evident drift, nowhere is the trail actually lost en
route (Fig. S1 in Soares et al. 2011). If a phylogeographically specic genetic trail can be left
across the Pacic Islands in spite of serial founder effects, it is more likely, than not, to be left
somewhere in the landmass of East Eurasia or at least have left specictraces.
Another recent article (Hooshiar Kashani et al. 2012) also argues against the SH on the basis
of modelled shared population genetic histories within America as summarized in the abstract
(below) but, surprisingly in this case, their argument is based on claimed simple associations
between two uncommon founders, being very similar to each other in America and different
from the other founders:
Taking into account that C4c is deeply rooted in the Asian portion of the mtDNA phylogeny
and is indubitably of Asian origin, the nding that C4c and X2a are characterized by parallel
genetic histories denitively dismisses the controversial hypothesis of an Atlantic glacial
entry route into North America. (Hooshiar Kashani et al. 2012, 35)
Their parallel evidence that X2a & C4c together formed a dual founding migration within
America (thus by inference discrete from the other A-D lineages) is based on three types of
association between mtDNA X2a and C4c in North America (clearly not on their lack of
association in NE Asia): (1) rarity, (2) co-distribution restricted to the North American con-
tinent, and (3) similar founding age. While simple associations may sometimes be considered as
suggestive, they are not normally accepted as denitive evidence. However, the genetically
unlinked regional/temporal associations offered here seem too non-specic even for suspicion,
let alone for the above refutation. To enlarge:
1. Firstly, shared rarity: C4c is indeed very rare (total sixteen complete sequences, three in
Canada in their Table 1), being rst discovered in Columbia in 2007, and much rarer than
X2a, which achieves frequencies up to 30 per cent in several Great Lakes First Nations
(Fig. 1). Moreover shared-rarity is present in many multiple haplogroup migrations and is
not regarded as a feature indicating unique common ancestry as a separate migration
within the same timescale as other co-founders, in formal uniparental Founder Analysis
(e.g., in Richards et al. 2000);
2. Co-distribution restricted to North America: While, with the exception of two basal
C4c instances in Columbia, South America (Tamm et al. 2007), both haplogroups are
Solutrean hypothesis: genetics 765
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
restricted to North America, they do not closely co-distribute regionally within that
continent. While X2a is clearly most frequent and diverse in the Canadian Great Lakes
region with phylogeographic evidence of some minor spread southeast from there (Fig. 1),
C4c has only three Canadian instances (one in British Columbia and two in Manitoban
mixed/Metis groups), with the rest widely scattered further south in the USA, excepting
the US west coast (their Fig. 1).
3. Parallel/dual Founding Ages of X2a & C1c: Using the current complete mtDNA
sequence calibration method (Soares et al. 2009), they estimated the root age of C4c to
be 13.8 +/3.8 ky (by Maximum Likelihood) and 12.3 +/2.9 ky (by Rho), which are
statistically indistinguishable from the Soares et al. (2009) ML estimates for ve other
American founding lineages with ultimate East Eurasian ancestors (A2:14.6 ky; B2: 14.6
ky; C1a:13.0 ky; C1b: 14.5 ky; and D1: 13.5 ky). By contrast, the estimates Hooshiar
Kashani et al. (2012) make for the age of American X2a (table 2) were older: 18.6 +/5.5
ky (ML) and 18.4 +/5.2 ky (Rho), and could not thus be described as more parallel to
C4c than to any other founding lineage.
To conclude on mtDNA, our discussion of evidence from the X2a/X2g lineages suggests that
there is a case to explore for those predictions of Pleistocene gene ow from West Eurasia to the
Americas, most parsimoniously by the direct route across the Atlantic. If so, it would seem
appropriate to look at other parts of the human genome for such inuence. Such questions have
not been formally addressed, at least not in relation to the Atlantic route suggested indepen-
dently by the SH and the phylogeography of X2a/X2g lineages. There is indeed some recent
evidence for pre-Columbian West Eurasian autosomal admixture but it is not clear why this is
presumed to have travelled across Siberia, as discussed in the next section.
Autosomal evidence for pre-Columbian West Eurasian admixture
Moving rst to the Rasmussen et al. (2014) claim introduced above: our genome analysis
refutes the possibility that Clovis originated via a European (Solutrean) migration to the
Americas.While not explaining the basis of their SH genomic refutation, the authors do
actually nd evidence for Pleistocene West Eurasian autosomal admixture in the Americas,
which would be predicted by the SH.
They make this inference by several analytic approaches. First they acknowledge a relative
genetic afnity of Anzick to Western Eurasians (Rasmussen et al. 2014: Supplementary
Information Sect. 15.6), suggesting admixture and gene ow prior to the rst migration into
Beringia and the Americas. However, while this relative West Eurasian afnity may have
arrived via Beringia, as they have presumed, they did not formally explore the alternative
direction of gene ow, namely that the minor, apparent evidence for West Eurasian afnity that
they detected, might have come more directly via the Atlantic (e.g., before Clovis tools
evolved), and subsequently admixed with the Siberian co-ancestors of all Americans. This
omission of any comparative test of direction weakens rather than strengthens their claim to
refute the SH.
766 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
Furthermore they also found that the ancient population that Anzick represented, was not
basal to that for modern Northern Amerinds (NA), analysed from further east in Canada,
although otherwise closely related with them (Rasmussen et al. 2014, 2267 and SI Sect.
15.4). A basal relationship of Anzick to modern Northern Amerinds should have been expected
from the sole Beringian entry hypothesis, given the relative proximity of NA to the east of the
Pleistocene Anzick child, and also since Anzick was actually basal to all other Native American
groups further south. This picture, of the ancestry of NA (and Chipewyan) groups distinct from
all other mainland groups, nds echoes in the modern genetic population phylogeny (Reich
et al. 2012,Fig 1c) mentioned above.
One of the possible explanatory models Rasmussen et al. (2014) tested for this anomaly of
NA ancestry was articulated as: The NA groups carry ancestry from a previously undocumen-
ted stream of gene ow from the Old World.Testing the model of undocumented stream of
gene ow, they looked for and found no evidence for Siberian gene ow into the Northern
Amerinds(Rasmussen et al. 2014, SI Sect. 15.4, 27). But, again they did not explore the
alternative explanatory possibility of an earlier trans-Atlantic stream of gene ow which, by
virtue of geography, would inuence the NA in East Canada to a greater extent than the SA
populations, in this context. These omissions of testing predictions of, or of evidence for, or
against, the SH hardly help to conrm their refutation of it.
Similarly to the Anzick report the Malta paper (Raghavan et al. 2014) infers pre-Columbian
West Eurasian admixture in mainland Native Americans. The Malta paper abstract states: Our
ndings reveal that western Eurasian genetic signatures in modern-day Native Americans derive
not only from post-Columbian admixture, as commonly thought, but also from a mixed ancestry
of the First Americans(Raghavan et al. 2014, 87).
The genetic afnity they noted between Native Americans and the Pleistocene Malta sample
was also detected to a signicant extent using modern-day Western Eurasian populations
instead, with West Europeans (Orcadian (highest), French Basque, French, North Italians and
Russians) having greater afnity than the Middle East or Central and South Asia (Raghavan
et al. 2014; Fig 3b; SI Sect 14.5). In spite of this the Northeast Asian route was still presumed
for these West Eurasian afnities, with no mention of an Atlantic route.
From such D-statistic analyses, they inferred that Native Americans have mixed origins,
resulting from admixture between peoples related to modern-day east Asians and western
Eurasians(Raghavan et al. 2014, 89). They also modelled past population admixture history
in relation to the Malta sample using MixMapper v1.0 and modeled Karitiana as having 14
38% western Eurasian ancestry and 6286% east Asian ancestry(the Karitiana are an isolated
South American group with no other apparent genetic evidence of recent European admixture).
Again, the trans-Atlantic possibility of gene ow was ignored as indicated by the statement: it
is likely that populations related to MA-1 [Malta] mediated the mixture event between western
Eurasians and Native Americans(Raghavan et al. 2014: SI Sect. 12).
Ancient admixture between Europeans and Karitiana and other groups was also explored by
Patterson et al. (2012), using f3-analysis on population triplesfrom the HGDP dataset. Without
going into detail, an extract from a key section entitled Evidence for Northeast Asian-related
genetic material in Europeagain reveals the assumption of Beringian route for such admixture:
We single out from Table 5 the score for French arising as an admixture of Karitiana, an
indigenous population from Brazil, and Sardinians. The Z-score of 18.4 is unambiguously
Solutrean hypothesis: genetics 767
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
statistically signicant. We do not of course think that there has been substantial gene ow
back into Europe from Amazonia. The only plausible explanation we can see for our signal
of admixture into the French is that an ancient northern Eurasian population contributed
genetic material to both the ancestral population of the Americas and the ancestral population
of northern Europe. This was quite surprising to us, and in the remainder of the article this is
the effect we discuss. (Patterson et al. 2012, 1083)
We are concerned that it is presumed in all these analyses that the inferred pre-Columbian west-
Eurasian admixture into Native Americans arrived via Beringia and there was no mention of the
possibility that this pre-Columbian west-Eurasian admixture could, alternatively, have come
across the Atlantic. There was nothing in the analysis, which formally or explicitly determined
the geographic direction of inferred, shared West Eurasian population ancestry, from Siberia or
from the Atlantic.
Y chromosome
The SH and a high frequency of X2a in northeastern Native American groups, might both
predict unexpectedly high Europeanor at least high West Eurasian Y haplogroup frequencies
there (e.g., Straus et al. 2005, 522). That is indeed the case overall (e.g., Hammer et al. 2005,
Fig 3D). In geographic plots, R1 frequencies in native populations, of the Great Lakes/
Algonquian-speakers stand out as the great majority, having among the highest worldwide R1
rates (e.g., Malhi et al. 2008; and World frequency map as of 1 June 2014), even higher than
non-western Europe and far higher than other Native Americans (c. 010 per cent). When
further characterized in the USA (Hammer et al. 2005), 97 per cent of R1 had the M269 SNP
(unambiguous Single Nucleotide Polymorphism), which denes R1b1b, the main West
European Y-haplogroup, which possibly originated there before the LGM (Morelli et al.
2010). The less-reliable P25 was used in an earlier US study (Zegura et al. 2004).
The main problem with interpreting unusually high Y-R1 frequencies in Northeast Native
Americans, is that the published SNP characterization is still too poorly resolved in all studies to
differentiate ancient migration from post-Columbian introduced R1. STR (Short Tandem
Repeats more rapidly mutating, but less reliable than SNPs) characterization of R1, however,
indicates a substantial proportion of derived, STR-haplotypes not shared with Europeans
(Bolnick, Bolnick, and Smith 2006, Fig 6b; Zegura et al. 2004, Fig. 5). This would not be
expected if those R1 STR types were all recent European introductions, and could be consistent
with being derived from more ancient founders. While better phylogenetic resolution is needed,
these results, far from refuting the SH, are more consistent with its predictions than solely with
massive recent male replacement.
The Solutrean hypothesis (SH) currently offers an archaeological explanation for the origins of the
majority of pre-Clovis cultural assemblages and their in situ evolution into Clovis; no such credible
cultural-evolutionary sequence has been offered for Palaeolithic East Eurasia as their cultural
768 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
source. In parallel, the SH offers the only credible route-explanation for the unique, substantial
presence of West Eurasian-derived X2g and X2a in the Great Lakes region of north-east America,
and their antiquity. The East Eurasian ancestry of 4/5 major American founding mitochondrial
lineages (A-D) is undisputed, but does not affect the above inferences.
As predicted by X2a, West Eurasian Y haplogroups have much higher absolute rates in north-
east Native Americans than those in the rest of the Americas. Three recent autosomal studies
have all found substantial evidence of pre-Columbian, West Eurasian autosomal admixture in
Native American populations. While this has all been presumed to have arrived via Beringia, the
alternative trans-Atlantic route for such gene ow was not considered or tested.
Genetic critiques of the SH and the X2a/X2g evidence for an additional Atlantic migration,
including the Rasmussen, Fagundes and Hooshiar Kashani refutations, are argued in the form
of counter-assertions, including hypothetical summary-descriptive similarities of X2a with A, B,
C and D and a presumed high likelihood of extinction of X in East Eurasia, rather than
hypothesis-testing and rules of evidence. However, while the SH prediction of the X2a/X2g
phylogeographic-evidence continues to attract controversy, much more work is required for the
rest of the genome, including balanced model-testing.
We thank Martin Richards and some anonymous reviewers for useful comments on our review.
Stephen Oppenheimer
University of Oxford
Institute of Social and Cultural Anthropology
Bruce Bradley
Department of Archaeology
University of Exeter
Dennis Stanford
Department of Anthropology
Smithsonian Institution
Avise, J. C. 2000. Phylogeography. Cambridge, MA: Harvard University Press.
Bamforth, D. 2013. Across Atlantic Ice: The Origin of Americas Clovis Culture by Dennis J. Stanford,
Bruce A. Bradley.Book Review Current Anthropology, 54 (1): 1057.
Bird, M. I., G. Hope, and D. Taylor. 2004. Populating PEP II: The Dispersal of Humans and Agriculture
Through Austral-Asia and Oceania.Quaternary International 11819:14563. doi:10.1016/S1040-6182
Solutrean hypothesis: genetics 769
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
Bird, M. I., C. S. M. Turney, L. K. Field, R. Jones, A. Palmer, R. Cresswell, and S. Robertson. 2002.
Radio-Carbon Analysis of the Early Archaeological Site of Nauwalabila I, Arnhemland, Australia:
Implications for Sample Suitability and Stratigraphic Integrity.Quaternary Science Reviews 21: 1061
75. doi:10.1016/S0277-3791(01)00058-0.
Bjerck, H. B., and H. M. Breivik, 2012. Off-Shore Pioneers: Scandinavian and Patagonian Lifestyles
Compared in the Marine Ventures Project.Antiquity 86 (333). Accessed August 25, 2014. http://antiquity.
Bolnick, D. A., D. I. Bolnick, and D. G. Smith. 2006. Asymmetric Male and Female Genetic Histories
among Native Americans from Eastern North America.Molecular Biology and Evolution 23 (11): 2161
74. doi:10.1093/molbev/msl088.
Bolnick, D. A., and D. G. Smith. 2007. Migration and Social Structure among the Hopewell: Evidence
from Ancient DNA.American Antiquity 72 (4): 62744.
Bonsall, C., C. Pickard, and P. Groom, 2013. Boats and Pioneer Settlement: The Scottish Dimension.
Norwegian Archaeological Review 46 (1): 8790.
Bradley, B., and M. Collins. 2013. Imagining Clovis as a Cultural Revitalization Movement.In
Paleoamerican Odyssey, edited by K. Graff, C. Ketron, and M. Waters, 24756. College Station: Center
for the Study of the First Americans, Department of Anthropology, Texas A&M University.
Bradley, B., and D. Stanford. 2004. The North Atlantic Ice-Edge Corridor: A Possible Paleolithic Route to
the New World.World Archaeology 36: 45978.
Brandt, G., W. Haak, C. J. Adler, C. Roth, A. Szécsényi-Nagy, S. Karimnia, S. Möller-Rieker, et al. 2013.
Ancient DNA Reveals Key Stages in the Formation of Central European Mitochondrial Genetic
Diversity.Science 342: 25761.
Brown, M. D., S. H. Hosseini, A. Torroni, H. J. Bandelt, J. C. Allen, T. G. Schurr, R. Scozzari, F. Cruciani,
and D. C. Wallace. 1998. mtDNA haplogroup X: An Ancient Link between Europe/Western Asia and
North America?The American Journal of Human Genetics 63 (6): 185261.
Cann, R. L., M. Stoneking, and A. C. Wilson. 1987. Mitochondrial DNA and Human Evolution.Nature
325: 316. doi:10.1038/325031a0.
Collins, M., D. Stanford, D. Lowery, and B. Bradley. 2013. North America Before Clovis: Variance in
Temporal/Spatial Cultural Patterns, 24,000 to 13,000 BP.In Paleoamerican Odyssey, edited by K. Graff,
C. Ketron, and M. Waters, 52140. College Station: Center for the Study of the First Americans,
Department of Anthropology, Texas A&M University.
Crespillo, M., J. A. Luque, M. Paredes, R. Fernandez, E. Ramirez, and J. L. Valverde. 2000.
Mitochondrial DNA Sequences for 118 Individuals from Northeastern Spain.International Journal of
Legal Medicine 114: 1302.
Derenko, M., T. Grzybowski, B. Malyarchuk, J. Czarny, D. Miscicka-Sliwka, and I. A. Zakharov. 2001.
The Presence of Mitochondrial Haplogroup X in Altaians from South Siberia.American Journal of
Human Genetics 69 (1): 23741.
Derenko, M., B. Malyarchuk, T. Grzybowski, G. Denisova, I. Dambueva, M. Perkova, C. Dorzhu, et al.
2007. Phylogeographic Analysis of Mitochondrial DNA in Northern Asian Populations.American
Journal of Human Genetics 81: 102541. doi:10.1086/522933.
Eren, M., R. Patten, J. OBrien, and D. Meltzer. 2013. Refuting the Technological Cornerstone of the Ice-
Age Atlantic Crossing Hypothesis.Journal of Archaeological Science 40: 293441.
Fagundes N. J. R., R. Kanitz, R. Eckert, A. C. S. Valls, M. R. Bogo, M. Francisco, F. M. Salzano, et al.
2008. Mitochondrial Population Genomics Supports a Single Pre-Clovis Origin with a Coastal Route for
the Peopling of the Americas.The American Journal of Human Genetics 82: 58392.
770 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
Fernandes, V., F. Alshamali, M. Alves, M. D. Costa, J. B. Pereira, N. M. Silva, L. Cherni, et al. 2012. The
Arabian Cradle: Mitochondrial Relicts of the First Steps along the Southern Route out of Africa.
American Journal of Human Genetics 90: 34755. doi:10.1016/j.ajhg.2011.12.010.
Fiedel, S., and J. Morrow. 2012. Comment on Clovis and Western Stemmed: Population Migration and
the Meeting of Two Technologies in the Intermountain Westby Charlotte Beck and George T. Jones.
American Antiquity 77 (22): 37685.
García, O., R. Fregel, J. M. Larruga, V. Ávarez, I. Yurrebaso, V. M. Cabrera, and A. M. González. 2011.
Using Mitochondrial DNA to Test the Hypothesis of a European Post-Glacial Human Recolonization
from the Franco-Cantabrian refuge.Heredity 106: 3745.
Gilbert, M. P., D. L. Jenkins, A. Götherstrom, N. Naveran, J. J. Sanchez, M. Hofreiter, P. F. Thomsen, et al.
2008. DNA from Pre-Clovis Human Coprolites in Oregon, North America.Science 320: 7869.
González, A. M., A. Brehm, J. A. Pérez, N. Maca-Meyer, C. Flores, and V. M. Cabrera. 2003.
Mitochondrial DNA Afnities at the Atlantic Fringe of Europe.American Journal of Physical
Anthropology 120: 391404.
Greenberg, J. H., C. G. Turner II, S. L. Zegura, L. Campbell, J. A. Fox, W. S. Laughlin, E. J. E. Szathmary,
et al. 1986. The Settlement of the Americas: A Comparison of the Linguistic, Dental, and Genetic
Evidence.Current Anthropology 27 (5): 47797.
Hammer, M. F., V. F. Chamberlain, V. F. Kearney, D. Stover, G. Zhang, T. Karafet, B. Walsh, and A. J.
Redd. 2005. Population Structure of Y Chromosome SNP Haplogroups in the United States and Forensic
Implications for Constructing Y Chromosome STR Databases.Forensic Science International 164: 45
55. doi:10.1016/j.forsciint.2005.11.013.
Helgason, A., E. Hickey, S. Goodacre, V. Bosnes, K. Stefánsson, R. Ward, and B. Sykes. 2001. mtDNA
and the Islands of the North Atlantic: Estimating the Proportions of Norse and Gaelic Ancestry.American
Journal of Human Genetics 68: 72337.
Hooshiar Kashani, B., U. A. Perego, A. Olivieri, N. Angerhofer, F. Gandini, V. Carossa, H. Lancioni, et al.
2012. Mitochondrial Haplogroup C4c: A Rare Lineage Entering America through the Ice-Free Corridor?
American Journal of Physical Anthropology 147: 359.
Izagirre, N., and C. de la Rúa. 1999. An mtDNA Analysis in Ancient Basque Populations: Implications
for Haplogroup V as a Marker for a Major Paleolithic Expansion from Southwestern Europe.American
Journal of Human Genetics 65: 199207.
Kuch, M., D. R. Gröcke, M. C. Knyf, M. P. T. Gilbert, B. Younghusband, T. Young, I. Marshall, E.
Willerslev, M. Stoneking, and H. Poinar. 2007. A Preliminary Analysis of the DNA and Diet of the
Extinct Beothuk: A Systematic Approach to Ancient Human DNA.American Journal of Physical
Anthropology 132: 594604.
Leavesley, M. G., M. I. Bird, L. K. Field, G. M. Dos Santos, P. A. Hausladen, and M. L. Di Tada. 2002.
Buang Merabak: 40,000 Radiocarbon Years of Human Occupation in the Bismarck Archipelago, Papua
New Guinea.Australian Archaeology 55: 557.
Malhi, R. S., K. E. Breece, B. A. Schultz Shook, and F. A. Kaestle. 2004. Patterns of mtDNA Diversity in
Northwestern North America.Human Biology 76 (1): 3354. doi:10.1353/hub.2004.0023.
Malhi, R. S., A. Gonzalez-Oliver, K. B. Schroeder, B. M. Kemp, J. A. Greenberg, S. Z. Dobrowski, and D.
G. Smith. 2008. Distribution of Y Chromosomes among Native North Americans: A Study of Athapaskan
Population History.American Journal of Human Genetics 137: 41224. doi:10.1002/ajpa.20883.
Malhi, R. S., B. A. Schultz, and D. G. Smith. 2001. Distribution of Maternal DNA Lineages among
Native American. Tribes of Northeastern North America.Human Biology 73: 1755.
Malhi, R. S., and D. G. Smith. 2002. Brief Communication: Haplogroup X Conrmed in Prehistoric
North America. American Journal of Physical Anthropology 119: 846. doi:10.1002/ajpa.10106.
Solutrean hypothesis: genetics 771
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
Morelli L., D. Contu, F. Santoni, M. B. Whalen, P. Francalacci, and F. Cucca. 2010. A Comparison of Y-
Chromosome Variation in Sardinia and Anatolia is More Consistent with Cultural Rather than Demic
Diffusion of Agriculture.PLoS ONE 5 (4): e10419. doi:10.1371/journal.pone.0010419.
Morrow, J. 2014. Book Review of Across Atlantic Ice.Lithic Technology 39 (1): 768.
Oppenheimer, S. 2003. Out of Eden: The Peopling of the World. London: Constable.
Oppenheimer, S. 2004. The Express Train from Taiwan to Polynesia: On the Congruence of Proxy Lines
of Evidence.World Archaeology: Debates in World Archaeology 36 (4): 591600.
Oppenheimer, S. 2012. Out-of-Africa, the Peopling of Continents and Islands: Tracing Uniparental Gene
Trees across the Map.Philosophical Transactions of the Royal Society B: Biological Sciences 367: 770
84. doi:10.1098/rstb.2011.0306.
Owsley, D., D. Hunt, I. Macintyre, and M. Logan. 2001. Clovis and Early Archaic Crania from the
Anzick Site (24PA506), Park County, Montana.Plains Anthropologist 46 (176): 11524.
Patterson, N., P. Moorjani, Y. Luo, S. Mallick, N. Rohland, Y. Zhan, and T. Genschoreck, et al. 2012.
Ancient Admixture in Human History.Genetics 192: 106593.
Perego, U. A., A. Achilli, N. Angerhofer, M. Accetturo, M. Pala, A. Olivieri, B. H. Kashani, et al. 2009.
Distinctive Paleo-Indian Migration Routes from Beringia Marked by Two Rare mtDNA Haplogroups.
Current Biology 19: 18. doi:10.1016/j.cub.2008.11.058.
Raghavan, M., P. Skoglund, K. E. Graf, M. Metspalu, A. Albrechtsen, I. Moltke, S. Rasmussen, et al.
2014. Upper Palaeolithic Siberian Genome Reveals Dual Ancestry of Native Americans.Nature 505:
Rasmussen, M., S. Anzick, M. Waters, P. Skoglund, M. DeGiorgio, T. Stafford, Jr, S. Rasmussen, et al.
2014. The Genome of a Late Pleistocene Human from a Clovis Burial Site In Western Montana.Nature
506: 2259.
Reich, D., N. Patterson, D. Campbell, A. Tandon, S. Mazieres, N. Ray, M. V. Parra, et al. 2012.
Reconstructing Native American Population History.Nature 488: 3714. doi:10.1038/nature11258.
Reidla, M., T. Kivisild, E. Metspalu, K. Kaldma, K. Tambets, H.-V. Tolk, J. Parik, et al. 2003. Origin and
Diffusion of mtDNA Haplogroup X.American Journal of Human Genetics 73: 117890.
Richards, M., V. Macaulay, E. Hickey, E. Vega, B. Sykes, V. Guida, C. Rengo, et al. 2000. Tracing European
Founder Lineages in the Near Eastern mtDNA Pool.American Journal of Human Genetics 67: 125176.
Roberts, R. G., and R. Jones. 2000. Chronologies of Carbon and of Silica: Evidence Concerning the
Dating of the Earliest Human Presence in Northern Australia.In Humanity from African Naissance to
Coming Millennia, edited by P. V. Tobias, 24352. Florence: Florence University Press.
Roberts, R. G., R. Jones, N. A. Spooner, M. J. Head, A. S. Murray, and M. A. Smith. 1994. The Human
Colonization of Australia: Optical Dates of 53,000 and 60,000 Years Bracket Human Arrival at the Deaf
Adder Gorge, Northern Territory.Quaternary Science Reviews 13: 57583. doi:10.1016/0277-3791(94)
Scozzari, R., F. Cruciani, P. Santolamazza, D. Sellito, D. E. C. Cole, L. A. Rubin, D. Labuda, et al. 1997.
MtDNA and Y Chromosome-specic Polymorphisms in Modern Ojibwa: Implications About the Origin
of Their Gene Pool.American Journal of Human Genetics 60: 2414.
Skoglund, P., H. Malmström, A. Omrak, M. Raghavan, C. Valdiosera, T. Günther, P. Hall, et al. 2014.
Genomic Diversity and Admixture Differs for Stone-Age Scandinavian Foragers and Farmers.Science
344: 74750. doi:10.1126/science.1253448.
Smith, D. G., R. S. Malhi, J. G. Lorenz, J. A. Eshleman, and F. A. Kaestle. 1999. Distribution of mtDNA
Haplogroup X among Native North Americans.American Journal of Physical Anthropology 110: 27384.
772 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
Soares, P., L. Ermini, M. Mormina, T. Rito, A. Röhl, S. Oppenheimer, V. Macaulay, and M. Richards.
2009. Correcting for Purifying Selection: An Improved Human Mitochondrial Molecular Clock.
American Journal of Human Genetics 84: 120. doi:10.1016/j.ajhg.2009.05.001.
Soares, P., T. Rito, J. Trejaut, M. Mormina, C. Hill, E. Tinkler-Hundal, M. Braid, et al. 2011. Ancient
Voyaging and Polynesian Origins.The American Journal of Human Genetics 88: 19. doi:10.1016/j.
Stanford, D., and B. Bradley. 2012 Across Atlantic Ice: The Origin of Americas Clovis Culture. Berkeley:
University of California Press.
Stone, A.C., and M. Stoneking. 1998. MtDNA Analysis of a Prehistoric Oneota Population: Implications
for the Peopling of the New World.American Journal of Human Genetics 62: 115370.
Straus, L., D. Meltzer, and T. Goebel. 2005. Ice Age Atlantis? Exploring the SolutreanClovis
Connection.World Archaeology 37: 50732.
Summerhayes, G. R., M. Leavesley, A. Fairbairn, H. Mandui, J. Field, A. Ford, and R. Fullagar. 2010.
Human Adaptation and Plant Use in Highland New Guinea 49,000 to 44,000 Years Ago.Science 330:
7880. doi:10.1126/science.1193130.
Tamm, E., T. Kivisild, M. Reidla M. Metspalu, D. G. Smith, C. J. Mulligan, C. M. Bravi, et al. 2007.
Beringian Standstill and Spread of Native American Founders.PLoS One 2: e829.
Taylor, D. 1969 The Wilsall Excavations: An Exercise in Frustration.Proceedings of the Montana
Academy of Sciences 29: 147150.
Torroni, A., T. G. Schurr, M. F. Cabell, M. D. Brown, J. V. Neel, M. Larsen, D. G. Smith, C. M. Vullo, and
D. C. Wallace. 1993. Asian Afnities and Continental Radiation of the Four Founding Native American
mtDNAs.American Journal of Human Genetics 53: 56390.
Torroni, A., R. Sukernik, T. G. Schurr, Y. B. Starikovskaya, M. F. Cabell, M. H. Crawford, A. G.
Comuzzie, and D. C. Wallace. 1993. mtDNA Variation of Aboriginal Siberians Reveals Distinct
Genetic Afnities with Native Americans.American Journal of Human Genetics 53: 591608.
Turney, C., M. I. Bird, L. K. Field, R. G. Roberts, M. Smith, C. E. Dortch, R. Grün, et al. 2001. Early
Human Occupation at Devils Lair, Southwestern Australia 50,000 Years Ago.Quaternary Research 55:
313. doi:10.1006/qres.2000.2195.
Vasek, A. 1983. On Language Shift and Language Switch.Dissertation, Brenenske University: Sbornik
praci lozocke Fakulty Brenenske Univerzity - Studia Minora Facultatis Philosophicae Universitas
Brunensis. A 31.
Waters, M. R., and T. W. Stafford. 2007. Redening the Age of Clovis: Implications for the Peopling of
the Americas.Science 315: 11226.
World frequency map: http://en.wikipedia.org/wiki/File:Haplogroup_R_(Y-DNA).PNG
Zegura, S. L., T. M. Karafet, L. A. Zhivotovsky, and M. F. Hammer. 2004. High-Resolution SNPs and
Microsatellite Haplotypes Point to a Single, Recent Entry of Native American Y Chromosomes into the
Americas.Molecular Biology and Evolution 21 (1): 16475. doi:10.1093/molbev/msh009.
Stephen Oppenheimer, BA (Physiol), BM, BCh, MA, DM (Oxon), DTM&H (Liver), FRCP,
FRCPC, (UK): Institute of Social and Cultural Anthropology, School of Anthropology, University
of Oxford. 197297: Paediatrician, 1987: Professor of Paediatrics in Malaysia, then Hong Kong.
Published 104 peer-reviewed papers, focused on the interactions between iron, nutrition, genetics
(particularly α
-thalassaemia) and malaria; 1982, developed a long-term interest in the use of
Solutrean hypothesis: genetics 773
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
unique genetic mutations to trace migrations. 1997present: Wrote three multidisciplinary books
reconstructing prehistoric migrations using genetics, archaeology, cultural anthropology and
Palaeoclimate: Eden in the East: The Drowned Continent of Southeast Asia (Weidenfeld &
Nicolson, 1998); Out of Eden: The Peopling of the World (Constable, 2003); The Origins of the
British: A Genetic Detective Story (Constable, 2006); published a further 66 peer-reviewed
papers, testing the original reconstructions in his three published books on ancient migrations.
Bruce Bradley is Professor of Prehistory and Director of the Experimental Archaeology
Masters Programme and has extensive experience with Stone Age technologies and experi-
mental archaeology. He was trained in 4-eld anthropology at the University of Arizona. His
early research was focused on the North American Southwest and Great Plains. Since then his
research has included the Upper Palaeolithic of Russia and France, and horse domestication in
Central Asia. His current areas of research deal with the early peopling of the New World,
prehistoric Pueblo archaeology of the American Southwest and skill learning and cognition in
hominins. He is also a noted expert intknapper. Bruce is active in bringing his archaeological
and anthropological interests to the public through presentations, teaching, interaction with
Native American communities and participation in documentaries.
Dennis Stanford is Curator of North and South American Paleolithic, Asian Paleolithic and
Western United States archaeological collections and Director of the Smithsonians Paleoindian/
Paleoecology Program. His research interests include origins and development of New World
Paleo-Indian cultures in relation to changing climate and ecosystems during the terminal
Pleistocene, interdisciplinary Quaternary studies, stone tool technology, experimental and public
archaeology. He has conducted eldwork in Siberia, China, Alaska, the Rocky Mountains,
Plains and Southeastern States; as well as Central and South America and south-western Europe.
Dennis has produced exhibitions and public programs; prepared scholarly and popular publica-
tions, and lms; served as advisor for graduate and PhD students and instructed undergraduates,
interns, and volunteers in eld and laboratory work.
774 Stephen Oppenheimer et al.
Downloaded by [Smithsonian Institution Libraries] at 10:21 02 November 2014
... The recent claim that the Anzick cache was a Clovis burial and that the genetic make-up of the supposedly 'associated' child disproves the Solutrean Hypothesis (Rasmussen et al.2013) is ridiculous. As an archaeologist, I am unable to question the genetic analysis and leave this to others, however, the archaeological evidence and the contextual-associative interpretations of the Anzick child are highly suspect (Oppenheimer et al. 2014). The circumstances of the find and recovery of materials from the Anzick site make any claims of specific contexts impossible to determine. ...
... If the 'Everybody out of Beringia as the only migration Hypothesis' has as wide a range of supporting evidence we strongly encourage somebody to publish it rather than just continue to assert it. Curiously, it is geneticists that propose a 'ghost' population in Eastern Eurasia to account for the total lack of X haplogroup let alone the ancestral pre-19,000 year old X2 root group (Oppenheimer et al. 2014). If this isn't 'evidence you wish you had' what is? ...
... I discuss these nuclear autosomal studies and others (i.e. those on mitochondrial DNA and the Y chromosome) in a joint review (Oppenheimer, Bradley & Stanford, 2014) on the Solutrean Hypothesis published in World Archaeology. ...
Full-text available
Comments, with references, by film contributors on the Guardian review by Jennifer Raff (also a film contributor) of the recent Canadian documentary "Ice Bridge" and published in the Guardian newspaper (UK) online: Jennifer Raff's Review title: "Rejecting the Solutrean hypothesis: the first peoples in the Americas were not from Europe: A recent Canadian documentary promoted a fringe idea in American archaeology that's both scientifically wrong and racist"
... A principal evidência dessa rota de migração são as semelhanças das pontas Clóvis, encontradas na América do Norte entre 13,4 e 12,7 mil anos (Miller et al., 2013), com os bifaciais da tradição Solutrense, existentes na Europa entre 25 e 16 mil anos Stanford, 2004;Bradley, 2013). Outra evidência apontada pelos pesquisadores é a presença, mesmo que em baixa frequência, do haplogrupo mitocondrial X em nativos americanos do nordeste da América do Norte, sendo este haplogrupo muito comum na população europeia atual (Perego et al., 2009;Oppenheimer et al., 2014). Porém, a ocorrência do haplogrupo X em nativos americanos atuais não necessariamente correspondeu à sua distribuição no passado. ...
... haplotipo exclusivo X2a é semelhante à dos outros quatro haplogrupos principais em nativos americanos (A, B, C e D), os quais são tipicamente encontrados na Ásia, e não na Europa, sugerindo que os cinco haplogrupos possivelmente vieram em uma mesma migração (Raff;Bolnick, 2015). Os proponentes da migração via Europa propõem que a ocorrência em alta frequência do haplogrupo R1 do cromossomo Y no leste da América do Norte seria uma evidência adicional de conexão entre Europa e Américas (Oppenheimer et al., 2014). Todavia, outros autores defendem que o haplogrupo R1 encontrado em nativos americanos é resultado de miscigenação recente com europeus (Zegura et al., 2004). ...
Full-text available
Resumo A ocupação das Américas tem sido discutida cientificamente desde o século XIX, gerando uma infinidade de modelos explicativos. Por outro lado, há limitações das evidências empíricas das primeiras ocupações, causadas por problemas depreservação e pela baixa visibilidade arqueológica. Nesse aspecto, o uso de dados biológicos de populações humanas atuais e antigas tem fornecido informações cruciais para a interpretação dessas primeiras ocupações. Sob uma perspectivabioarqueológica, ou seja, através do estudo dos remanescentes biológicos humanos em contexto arqueológico, este texto sintetiza o atual entendimento sobre a rota de entrada, a data de entrada inicial, o número de migrações, a subsistência eos rituais mortuários dos primeiros americanos. Os resultados desse panorama sintético indicam que há temáticas de alto consenso (rota de entrada), de consenso intermediário (data de entrada) e de baixo consenso (número de migrações).Por outro lado, temáticas como a saúde e o modo de vida dos habitantes antigos das Américas ainda carecem de estudos mais aprofundados. Este texto ressalta a importância do conhecimento bioarqueológico para a formulação de modelosde ocupação, buscando incorporar de forma equilibrada evidências da América do Sul e do Norte.
... There seems little doubt that the initial pulse of people into North America originated in northeast Asia, a conclusion supported by a series of data sets, including linguistic data (e.g., Nichols 2015), blood groupings (e.g., Szathmary 1979), dental traits (e.g., Turner 2002), paleogenetics (e.g., Achilli et al. 2013;Fiedel 2017;Llamas et al. 2016;Raff and Bolnick 2014;Raghavan et al. 2015; but see Oppenheimer, Bradley, and Stanford 2014), and archaeological data (e.g., Graf 2014). A less likely alternative is the Solutrean model that proposed people first entered the New World from Europe (Oppenheimer, Bradley, and Stanford 2014;Stanford and Bradley 2012; but see Walker and Clinnick 2014), although most researchers do not accept this model. ...
... There seems little doubt that the initial pulse of people into North America originated in northeast Asia, a conclusion supported by a series of data sets, including linguistic data (e.g., Nichols 2015), blood groupings (e.g., Szathmary 1979), dental traits (e.g., Turner 2002), paleogenetics (e.g., Achilli et al. 2013;Fiedel 2017;Llamas et al. 2016;Raff and Bolnick 2014;Raghavan et al. 2015; but see Oppenheimer, Bradley, and Stanford 2014), and archaeological data (e.g., Graf 2014). A less likely alternative is the Solutrean model that proposed people first entered the New World from Europe (Oppenheimer, Bradley, and Stanford 2014;Stanford and Bradley 2012; but see Walker and Clinnick 2014), although most researchers do not accept this model. ...
Paleoindian osseous technology includes a number of artifact categories, including at least four types of “rods” (straight uni-beveled, straight bi-beveled, curved uni-beveled, and bi-pointed). An analysis of the distribution and dating of known examples of such artifacts from contexts across North America leads to the proposal that osseous rod technology was first developed by Clovis-era people; that the bi-beveled type was confined to Clovis times, that the two uni-beveled types were retained in later Paleoindian tool kits, and that the bi-pointed type was used primarily in late Paleoindian times. A number of different functions of these artifacts have been proposed, the bi-beveled type most often as foreshafts and the uni-beveled types most often as projectile points. Other proposed functions include prybars, hafting wedges, flaking tools, or components of composite fishing spears, an idea expanded on here.
... Clovis research trends over the past two decades remain focused on searching for evidence, and the cultural origins, of Clovis technology. The dissemination of proposed hypotheses, investigations, as well as replies and rebuttals provide intellectual forage that draw lively and bitter debate on the subject (Bradley and Stanford 2004;Curry 2012;Eren, et al. 2013;Haynes, et al. 2007Lohse, et al. 2014a;Morrow, et al. 2012;O'Brien, et al. 2014a;Oppenheimer, et al. 2014;Rasmussen, et al. 2014;Stanford and Bradley 2002;Straus 2000;Straus, et al. 2005;Waters, et al. 2011b;Waters and Stafford 2007). ...
... The state of Clovis research also fundamentally influences molecular genetic research of the earliest Americans (Oppenheimer et al., 2014), and research on human genome sequences (Rasmussen, et al., 2014) which is intimately linked to alternative migration hypotheses (Goebel et al. 2008;Stanford and Bradley 2012). ...
Full-text available
This research examines the technology behind Clovis biface production from Clovis manufacturing areas at the Gault Site, Texas,(41BL323), with specific focus on flake striking platform preparation traits. Lithic analysts agree that platform bearing flakes retain clues into knapping technologies (Andrefsky 2005:86). Clovis experts agree that Clovis knappers invested effort before removing flakes by preparing platforms (Bradley, et al.2010:66; Morrow 1995) for exerting control during biface manufacture, including mastering control of overshot flaking (Bradley 2010:466). Evidence shows that Clovis knappers were highly skilled in their craft and preferred high quality raw materials to manufacture their tools and frequently produced overshot flakes. While basic manufacturing traits are present, Clovis represents a complex bifacial reduction technology (Bradley, et al. 2010:64). The data here elucidates differences in the application of reduction techniques used by Clovis. These data reveal no set pattern in the application of platform preparation traits used by Clovis knappers, but identified trends in the use of preparing platforms in flake types and phases that highlight Clovis biface reduction sequences, which may have followed a systematic ‘template.’ Therefore, a consistent approach may have been used to produce Clovis bifaces, but individual platform preparation traits were not. In addition to this study, a supplemental study was conducted concerning the intentionality of Clovis overshot flaking. This separate study revealed these flakes regularly exhibit the removal of stacks, hinges, deep flake scars and other error traits. As such, overshot flakes were a technique that served a dual purpose of removing errors while simultaneously thinning the biface. This research has contributed to a greater understanding of Clovis biface technology reduction processes and flake removal techniques used at the Gault Site.
... Furthermore, the genomes of ancient and recent Native Americans show no early input from Upper Paleolithic Western Europeans. The X2a clade of mitochondrial DNA, proffered as supposedly incontrovertible evidence of a European connection (Oppenheimer et al. 2014), is only present in North America (mainly found today among Algonquianspeaking groups, particularly the Ojibwa/Chippewa ). ...
The study of the peopling of the Americas has been transformed in the past decade by astonishing progress in paleogenomic research. Ancient genomes now show that Native American ancestors were formed in Siberia or the Amur region by admixture of ca. 15–30% Ancient North Eurasian genes with those of East Asians. The Anzick infant, buried with Clovis bifaces at 12,900 cal BP, belonged to a group that was ancestral to later Native Central and South Americans. Fishtail points, derived from Clovis, mark the arrival and rapid expansion of Clovis-descended Paleoindians across South America, also evident in the sharp increase of radiocarbon dates, continent-wide, at 13,000–12,500 cal BP. In both North and South America, extinction of most genera of megafauna was virtually simultaneous with Paleoindian expansion. Human hunting must have been involved, perhaps in concert with other indirect impacts. Contrary to the alternative bolide impact theory, there is no evidence of a dramatic human population decline after 12,800 cal BP. Ancient genomes show that divergent lithic traditions after 13,000 cal BP need not be attributed to a separate Pacific Rim migration stream. Several recent finds raise the possibility that pre-Clovis people might have reached the Americas before 20,000 cal BP, but these precursors must have either failed to thrive, or were ultimately replaced by proto-Clovis or Clovis people. Consilient paleogenomic and archaeological data indicate that initial colonization by Paleoindian ancestors of living Native Americans occurred after 14,500 cal BP.
... The controversies of the first settlement revolve around three distinct paradigms relating to the early humans settled in the New World [25]. These questions are best answered by the assessment of distinct hypotheses that build up the framework of archaeological and cultural evidence among which the following are the most relevant: (a) the Clovis hypothesis [26], (b) the coastal entry hypothesis [27] and (c) the Solutrean Hypothesis [28]. The Clovis hypothesis sets out as the most commonly accepted hypothesis that the first settlement of the Americas started during the late Ice Age when hunter-gatherers were entering North America through the Bering Land Bridge. ...
Full-text available
Background: Land-use change, resulting from urban expansion, has led to the unprecedented importance of rethinking our interactions with the environment in general. The consequences of anthropogenic activity on natural land, as well as sustainable land use types, is a significant concern to the detriment of small towns. In this article, I assess the changes in the urban footprint morphology of Northumberland County while evaluating the consequences of urban fragmentation on archaeological sites. Northumberland has a total population of 85,598 divided into seven municipalities and the Alderville First Nation reserve. Northumberland County is a commuting distance of one hour from Toronto, one of North America’s most important economic hubs. It is of the highest importance to measure and understand the current characteristics of land-use change and its impacts on the sustainability of small towns and local communities. Methods: A total of archaeological sites are used to test the impacts of urbanization and land-use change throughout the perimeter of Northumberland county. Spatial analysis is conducted employing spatial autocorrelation techniques (global and local) to assess the existence of hotspots of archaeological sites and the potential endangerment of these sites due to urban pressure. A hexagonal topology is used to derive a regional understanding of impacted regions within the county, allowing the richness of archaeological heritage sites as well as the possible concerns of these spatial clusters to be examined. Results: It is shown that a significant number of archaeological sites are present in Northumberland county, with a clear indication of a hotspot around Rice Lake. Future urbanization as well as land-use fragmentation may have an impact on the sustainability of these archaeological sites. It is shown that hexagonal geometries may establish relevant spatial surfaces to compare urban and archaeological sites as criteria to understand the endangerment of archaeological sites. The availability of the Global Urban Footprint (GUF) creates a particularly rigorous framework for the detection of impact on archaeological sites. Conclusion: The findings suggest that archaeological sites in small towns can benefit from high-resolution urban land-use footprints to detect impacts on archaeological sites and the relations of land-use change and intrasite management. This is of utmost importance in times of significant anthropogenic change brought by increasing urban pressure on small and medium-sized towns.
... Genetic studies conclusively show that the first Americans did not originate from Europe (9,21) as posited by the Solutrean hypothesis (22,23). Genetic evidence also does not support a successful occupation of the Americas before~17.5 ...
Full-text available
North and South America were the last continents to be explored and settled by modern humans at the end of the Pleistocene. Genetic data, derived from contemporary populations and ancient individuals, show that the first Americans originated from Asia and after several population splits moved south of the continental ice sheets that covered Canada sometime between ~17.5 and ~14.6 thousand years (ka) ago. Archaeological evidence shows that geographically dispersed populations lived successfully, using biface, blade, and osseous technologies, in multiple places in North and South America between ~15.5 and ~14 ka ago. Regional archaeological complexes emerged by at least ~13 ka ago in North America and ~12.9 ka ago in South America. Current genetic and archaeological data do not support an earlier (pre-17.5 ka ago) occupation of the Americas.
We present a theoretical framework for applying Geographic Information System least-cost path modeling to initial migration into the Americas. When integrated with theory, geospatial models can serve as a powerful tool for exploring how ancient humans may have moved through marine environments. We show how a strong theoretical foundation can be used to conduct Maritime Least Cost Path analysis to better understand where, when, and how migration and movement events may have taken place in the deep past. This paper examines possible peopling events along the Pacific Coast of North America during the Late Pleistocene. We ask what the role of modeling is in the search for early sites in the Americas and how models can most effectively be employed by looking at two case studies of migration into Prince Rupert Harbor and the Dundas Islands in British Columbia, Canada around 16,000 cal. yr BP. Our analysis created paths of least resistance for moving through landscapes based on different environmental, physiological, and cultural factors of boat-based travel. We examine the predicted paths created by our model to determine what they say about possible routes of maritime migration and consider the implications for understanding constraints and human agency in migration events.
Full-text available
Founder analysis is a method for analysis of nonrecombining DNA sequence data, with the aim of identification and dating of migrations into new territory. The method picks out founder sequence types in potential source populations and dates lineage clusters deriving from them in the settlement zone of interest. Here, using mtDNA, we apply the approach to the colonization of Europe, to estimate the proportion of modern lineages whose ancestors arrived during each major phase of settlement. To estimate the Palaeolithic and Neolithic contributions to European mtDNA diversity more accurately than was previously achievable, we have now extended the Near Eastern, European, and northern-Caucasus databases to 1,234, 2, 804, and 208 samples, respectively. Both back-migration into the source population and recurrent mutation in the source and derived populations represent major obstacles to this approach. We have developed phylogenetic criteria to take account of both these factors, and we suggest a way to account for multiple dispersals of common sequence types. We conclude that (i) there has been substantial back-migration into the Near East, (ii) the majority of extant mtDNA lineages entered Europe in several waves during the Upper Palaeolithic, (iii) there was a founder effect or bottleneck associated with the Last Glacial Maximum, 20,000 years ago, from which derives the largest fraction of surviving lineages, and (iv) the immigrant Neolithic component is likely to comprise less than one-quarter of the mtDNA pool of modern Europeans.
Full-text available
Americas Redefining the Age of Clovis: Implications for the Peopling of the This copy is for your personal, non-commercial use only. clicking here. colleagues, clients, or customers by , you can order high-quality copies for your If you wish to distribute this article to others here. following the guidelines can be obtained by Permission to republish or repurpose articles or portions of articles): July 7, 2014 www.sciencemag.org (this information is current as of The following resources related to this article are available online at
Full-text available
Clovis technology, as known in the durable record, consists of several distinctive flaked-stone reduction strategies as well as the manufacture of a range of bone, ivory, and antler tools. Stone was flaked to produce large flakes from bifacial cores: blades from two core-reduction technologies, and bifaces for varying purposes including the distinctive points known as Clovis. All these were complex technologies, which demanded expert knowledge and significant skill to achieve, even at a basic level. Special characteristics such as the extraordinary selection of exotic raw materials, production of oversized bifaces for caching, controlled full-face and overshot biface flaking, and flat-backed blade core maintenance are some of the features that indicate “deep” technologies that must have had significant and distinguishable antecedents in the archaeological record. These specific technologies span multiple ecological zones from the sub-arctic to the tropics, indicating an astonishing consistency and a system imposed on environmental factors rather than controlled by them. These features and behaviors are used to propose that Clovis was the product of culture change known as a revitalization movement. This anthropological concept is introduced in detail and then used to suggest that Clovis may not have been a single culture but a disparate set of cultures unified by a technologically coded belief system. © 2014 by Kelly E. Graf, Caroline V. Ketron, and Michael R. Waters. All rights reserved.
Full-text available
Significant new archaeological discoveries made over recent decades present a wide range of contrasting cultural patterns across North America during various portions of time between 27,000 and 13,000 cal yr BP. Each of these patterns is now represented by multiple sites within distinctive environmental settings. The extent and variance of this rich archaeological fabric indicates a much more complex process of peopling North America (and the Western Hemisphere), with multiple cultural origins and a long period of human presence prior to the advent of the distinctive Clovis manifestation, or coexisting with but differing from Clovis. The focus is on North America with limited mention of selected datasets in South America. © 2014 by Kelly E. Graf, Caroline V. Ketron, and Michael R. Waters. All rights reserved.
Full-text available
Beck and Jones (2010) assert that Clovis "was not first in the Intermountain West"; Western Stemmed points are older than fluted points; and the stemmed point makers derived from a hypothetical pre-13,000 cal B.P. Pacific Coast migration. A less tendentious review of the data suggests instead that Western Stemmed follows Clovis in this region, as previously inferred by Willig and Aikens.
The Anzick archaeological site in Montana, known since 1968, has been recognized since 1974 as Clovis age because of the lithic material recovered. Now it is possible to assert that some, but not all, of the recovered skeletal material is also Clovis age. The site contained more than 100 stone and nonhuman bone artifacts in association with the partial skeletal remains of a young child. All were covered with red ocher. The partial cranium of a second youth, not stained red, was also recovered. Recently, radiocarbon dating of these two crania produced two separate ages, 2000 years apart. Forensic examination of the remains, completed in 1999, has determined their ages at death. This site, an important addition to New World prehistory, contains not only the largest single assemblage of Clovis artifacts but also the only known Clovis skeletal remains.
Who were the first humans to inhabit North America? According to the now familiar story, mammal hunters entered the continent some 12,000 years ago via a land bridge that spanned the Bering Sea. Distinctive stone tools belonging to the Clovis culture established the presence of these early New World people. But are the Clovis tools Asian in origin? Drawing from original archaeological analysis, paleoclimatic research, and genetic studies, noted archaeologists Dennis J. Stanford and Bruce A. Bradley challenge the old narrative and, in the process, counter traditional-and often subjective-approaches to archaeological testing for historical relatedness. The authors apply rigorous scholarship to a hypothesis that places the technological antecedents of Clovis in Europe and posits that the first Americans crossed the Atlantic by boat and arrived earlier than previously thought. Supplying archaeological and oceanographic evidence to support this assertion, the book dismantles the old paradigm while persuasively linking Clovis technology with the culture of the Solutrean people who occupied France and Spain more than 20,000 years ago. Who were the first humans to inhabit North America? According to the now familiar story, mammal hunters entered the continent some 12,000 years ago via a land bridge that spanned the Bering Sea. Distinctive stone tools belonging to the Clovis culture established the presence of these early New World people. But are the Clovis tools Asian in origin? Drawing from original archaeological analysis, paleoclimatic research, and genetic studies, noted archaeologists Dennis J. Stanford and Bruce A. Bradley challenge the old narrative and, in the process, counter traditional-and often subjective-approaches to archaeological testing for historical relatedness. The authors apply rigorous scholarship to a hypothesis that places the technological antecedents of Clovis in Europe and posits that the first Americans crossed the Atlantic by boat and arrived earlier than previously thought. Supplying archaeological and oceanographic evidence to support this assertion, the book dismantles the old paradigm while persuasively linking Clovis technology with the culture of the Solutrean people who occupied France and Spain more than 20,000 years ago.
Who were the first humans to inhabit North America? According to the now familiar story, mammal hunters entered the continent some 12,000 years ago via a land bridge that spanned the Bering Sea. Distinctive stone tools belonging to the Clovis culture established the presence of these early New World people. But are the Clovis tools Asian in origin? Drawing from original archaeological analysis, paleoclimatic research, and genetic studies, noted archaeologists Dennis J. Stanford and Bruce A. Bradley challenge the old narrative and, in the process, counter traditional-and often subjective-approaches to archaeological testing for historical relatedness. The authors apply rigorous scholarship to a hypothesis that places the technological antecedents of Clovis in Europe and posits that the first Americans crossed the Atlantic by boat and arrived earlier than previously thought. Supplying archaeological and oceanographic evidence to support this assertion, the book dismantles the old paradigm while persuasively linking Clovis technology with the culture of the Solutrean people who occupied France and Spain more than 20,000 years ago.