Jan 072012

A paper by Kashani et al. in the latest issue of American Journal of Physical Anthropology brings 2012 to a bold start with its claim (excerpt from abstract below) of having decisively rejected Paleolithic European involvement in Amerindian origins, as argued by Stanford and Bradley (2004) and other advocates of the Solutrean hypothesis:

The similarities in ages and geographical distributions for C4c and the previously analyzed X2a lineage provide support to the scenario of a dual origin for Paleo-Indians. Taking into account that C4c is deeply rooted in the Asian portion of the mtDNA phylogeny and is indubitably of Asian origin, the finding that C4c and X2a are characterized by parallel genetic histories definitively dismisses the controversial hypothesis of an Atlantic glacial entry route into North America.

The apparent confidence of this assertion belies some fairly fundamental objections. First, it hinges on a problematic synecdoche, treating a single pillar of argumentation (prehistoric European provenance of mtDNA haplogroup X2 in the Americas) as a stand-in for the entire Solutrean edifice. Granted, this is a line of support whose rejection wouldn’t do any good for Stanford and Bradley’s case, but it’s worth realizing that these latter authors formulated their original argument in non-genetic terms; nowhere do Kashani et al. directly engage with their comparative-technological claims about the origins and diffusion of the Clovis toolkit, for instance. Second, while concordance of entry time and geographic distribution between X2a and a second rare and restricted haplogroup is certainly consistent with simultaneous or near-simultaneous entry along the same migratory pathway (namely, direct entry from Beringia to North America east of the Rockies via “the ice-free corridor between the Laurentide and Cordilleran ice sheets”), compelling reasons to reject alternative scenarios able to generate the same patterns ought to be tightly in hand before one speaks of definitive dismissal.

I’m a skeptic of trans-Atlantic Ice Age dispersal (or at least how it’s commonly imagined) myself, but it’s apparent to me that more cautious wording would have been prudent — particularly since the geographic correspondences are far from perfect, and the temporal alignment is more plausible overlap than bullseye hit.

C4c and X2a: Geographic patterns

C4c was isolated and confirmed as a founding Amerindian haplogroup quite recently (Tamm et al., 2007; Malhi et al., 2010), and would seem to be even rarer than X2a. Kashani et al. (2012) provide an overview of the localities in Table 1:

These are mirrored in the inset map of Fig. 1 (accompanying a maximum parsimony tree based on complete mtDNA sequences):

X was recognized by Forster et al. as an Amerindian mtDNA haplogroup in 1996, and has since then been identified in the Ojibwa [aka Chippewa] (25.7% in Manitoulin Island and 25% in northern Ontario: Scozzari et al., 1997), Sioux (14.6%), Nuu-Chah-Nulth [aka Nootka] (11.1 and 13.3%), Navajo (6.5%), and Yakima (4.8%) (Brown et al., 1998). Smith et al. (1999) detected X in the Chippewa (amongst whom it once again saw its highest frequency), Cheyenne/Arapaho, Micmac, Blackfoot, Kiowa, Jemez Pueblo, Pomo, and Sioux. With one notable exception — discussed later in this post — all samples of Amerindian X so far documented in the literature appear to be branches of X2a, which has two primary sub-divisions, X2a1 and X2a2 in the Great Lakes and Great Plains, and a possible third further west, among the Nuu-Chah-Nulth and Yakima (Perego et al., 2009).

Contrast these X2a maps against C4c’s earlier on. The first, at left (Smith et al., 1999), summarizes presence data without regard for frequency (the Navajo records were taken as indicative of “extensive and recent admixture with Pueblo tribes living in the Southwest” and, presumably for the same reason, are omitted):

The second, right beside it, is a spatial frequency distribution adapted from Fig. 3 of Perego et al. (2009): The left half represents data from “general mixed populations of national states” (taken from this database), and the right half exclusively Native American groups; sample locations show up as dots. I think the difference between the two has something to do with the lack of Métis inclusion in the Native American dataset, in contrast to their probable presence in the “general mixed” one.

The most notable discrepancy between the C4c and X2 distributions is, of course, the existence of a highly divergent sequence of the former in Columbia; X2a has to date never been recovered outside of North America. However, if we accept indications that it was found amongst aDNA from 7-8 ka at the Windover site in Florida (Hauswirth et al., 1994), it seems clear that both haplogroups had formerly much more expansive distributions. If the Great Lakes/Great Plains zone of overlap between the two was a refugium rather than the locus of initial expansion (“the terminal part of the glacial corridor”, Kashani et al. suggest), the strength of suppositions about entry route diminishes in consequence. And really, it’s anyone’s guess as to how much agreement X2a and C4c distributions exhibited with each other and with their present-day counterparts in prehistory.

Setting all that aside, I won’t dispute that C4c’s geographic pattern, from what we see here, differs significantly from those of other minor lineages like D4h3, which is limited to the Pacific coast of the Americas (Perego et al., 2009).

C4c and X2a: Divergence time estimates

Using the molecular clock model of Soares et al. (2009), Kashani et al. determined the maximum likelihood divergence for C4c as a whole to correspond to a divergence time of 13.8 ± 3.8 ky; the ML divergence for C4c1 (the primary subclade, containing all but the lone Columbian and two Canadian branches) translated into a divergence time of 9.7 ± 2.6 ky. The corresponding ρ divergence time estimates (based on the average distances of the haplotypes from the root) were 12.3 ± 2.9 ky for C4c and 8.5 ± 1.9 ky for C4c1.

Applied to all available full mitochondrial genomes of X2a, the same analyses yielded ML and ρ divergence time estimates of 18.6 ± 5.5 ky and 18.4 ± 5.1 ky, respectively — dates the authors judged to “overlap with those observed for C4c when taking into account standard errors”.

Tidy though this picture might be, it doesn’t account for one major potential complication: In 2009, as the present paper indeed notes, Perego et al. discovered X2g, a novel member of X2 that was neither X2a nor any of the Old World branches (X2b-X2f), in a single Ojibwa individual, and concluded that this probably indicated “an additional and very rare Native American founder”. I haven’t yet estimated TMRCA for the combined clade of X2a/X2g as per Soares et al. (2009), but you’re free to wager how much overlap their coalescence times would have with those C4c intervals from the following phylogeny (modified from Fig. S1 of Perego et al., 2009):

The node labels are divergence time estimates based on the ρ-statistic. (Again, do realize that the simple clock model employed here assumed a linear relationship between coding-region substitutions or synonymous transitions on the one hand and years on the other; Soares et al. took a more sophisticated tack that aimed to correct for purifying selection.)

A lot of stories could be spun about X2g’s implications for the age and history of Amerindian X2 (multiple chronologically distinct arrivals, unprecedentedly ancient differentiation within America, etc.), but it’ll be hard to be place much confidence in any of them so long as this sequence remains a singleton.

Closing remarks

You’ll notice that I so far haven’t devoted any space to the paleoceanographic and technological rebuttals to Stanford and Bradley and co. advanced by Straus et al. (2005) and Westley and Dix (2008), among others. Nor have I brought up the SNP-based STRUCTURE and ADMIXTURE analyses of Amerindians and north Eurasians by Dioegenes and the Eurogenes Genetic Ancestry Project whose intriguing implications are summed up here.

No need to read too much into this; I didn’t intend for this post to be a comprehensive position paper on the Solutrean hypothesis. I merely aimed to examine Kashani et al.’s work on its own merits, and to thereby decide whether their “scenario in which C4c and X2a are characterized by parallel genetic histories” was compelling enough to “definitively [dismiss] the controversial Solutrean hypothesis of an Atlantic glacial entry route into North America for X2a”. In lieu of any new X2a sequence data, and in lieu of any truly eye-popping geographic agreements (which even then could very well have more to do with post-arrival processes than unity of migration route into America), I hesitate to say that it lived up to the abstract’s promise.

In all fairness, Kashani et al. provide a nuanced and appropriately cautious framework for the interpretation of their findings — which is especially well-taken in light of the sample sizes — in their Discussion section:

Native American groups have undergone expansions and extinctions both prior and after the arrival of Europeans. Moreover, mtDNA is particularly prone to genetic drift, especially in tribal populations, thus one or even more mtDNA haplogroups can be easily lost in Native American groups (Torroni et al., 1994). […] Obviously a fundamental contribution to the resolution of this issue can be provided only by extensive ancient DNA studies performed all over the Americas.

[…] Therefore, although a temporally distinct arrival for C4c cannot be completely dismissed, the similarities in both ages and geographical distributions between C4c and X2a suggest that these two lineages possibly arrived together from Beringia, with the same Paleo-Indian group(s)…

It’s a little puzzling, then, to consider the forcefulness of their wording elsewhere.


I expect we’ll be hearing a good deal more about the “Atlantic route” in coming months. The University of California Press will be coming out with Across Atlantic Ice: The Origin of America’s Clovis Culture, coauthored by Stanford and Bradley, in late February 2012. (Sample chapter and synopsis to be found here.) I’m looking forward to seeing the “original archaeological analysis, paleoclimatic research, and genetic studies” they’ll bring to the table.

As for Kashani et al.’s aDNA comment, they ought to have said “all over the Americas — and Eurasia”. While X2a and X2g might still be exclusively American, neither extant nor ancient coverage of Siberia is superb, and ancient European X2 discoveries (e.g., the 2011 papers of Lacan et al. and Deguilloux et al) admonish us not to overlook the western half of the continent either.

Works cited

Brown, M., Hosseini, S., Torroni, A., Bandelt, H., Allen, J., Schurr, T., Scozzari, R., Cruciani, F., & Wallace, D. (1998). mtDNA haplogroup X: An ancient link between Europe/Western Asia and North America? American Journal of Human Genetics, 63(6), 1852-1913.

Deguilloux, M.-F., Soler, L., Pemonge, M.-H., Scarre, C., Joussaume, R., & Laporte, L. (2011). News from the west: Ancient DNA from a French megalithic burial chamber. American Journal of Physical Anthropology, 144(1), 108-118.

Forster, P., Harding, R., Torroni, A., & Bandelt, H. (1996). Origin and evolution of Native American mtDNA variation: a reappraisal. American Journal of Human Genetics, 59(4), 935-980.

Hauswirth, W., Dickel, C., Rowold, D., & Hauswirth, M. (1994). Inter- and intrapopulation studies of ancient humans. Experientia, 50(6), 585-676.

Kashani, B., Perego, U., Olivieri, A., Angerhofer, N., Gandini, F., Carossa, V., Lancioni, H., Semino, O., Woodward, S., Achilli, A., & Torroni, A. (2012). Mitochondrial haplogroup C4c: A rare lineage entering America through the ice-free corridor? American Journal of Physical Anthropology, 147(1), 35-44.

Lacan, M., Keyser, C., Ricaut, F.-X., Brucato, N., Duranthon, F., Guilaine, J., Crubézy, E., & Ludes, B. (2011). Ancient DNA reveals male diffusion through the Neolithic Mediterranean route. Proceedings of the National Academy of Sciences of the United States of America, 108(24), 9788-9879.

Malhi, R. S., Cybulski, J. S., Tito, R. Y., Johnson, J., Harry, H., & Dan, C. (2010). Brief communication: Mitochondrial haplotype C4c confirmed as a founding genome in the Americas. American Journal of Physical Anthropology, 141(3), 494-497.

Perego, U., Achilli, A., Angerhofer, N., Accetturo, M., Pala, M., Olivieri, A., Kashani, B., Ritchie, K., Scozzari, R., Kong, Q.-P., Myres, N., Salas, A., Semino, O., Bandelt, H.-J., Woodward, S., & Torroni, A. (2009). Distinctive Paleo-Indian migration routes from Beringia marked by two rare mtDNA haplogroups. Current Biology, 19(1), 1-9.

Scozzari, R., Cruciani, F., Santolamazza, P., Sellitto, D., Cole, D., Rubin, L., Labuda, D., Marini, E., Succa, V., Vona, G., & Torroni, A. (1997). mtDNA and Y chromosome-specific polymorphisms in modern Ojibwa: implications about the origin of their gene pool. American Journal of Human Genetics, 60(1), 241-245.

Smith, D., Malhi, R., Eshleman, J., Lorenz, J., & Kaestle, F. (1999). Distribution of mtDNA haplogroup X among Native North Americans. American Journal of Physical Anthropology, 110(3), 271-355.

Soares, P., Ermini, L., Thomson, N., Mormina, M., Rito, T., Röhl, A., Salas, A., Oppenheimer, S., Macaulay, V., & Richards, M. (2009). Correcting for purifying selection: an improved human mitochondrial molecular clock. American Journal of Human Genetics, 84(6), 740-799.

Stanford, D. J., & Bradley, B. A. (2004). The North Atlantic ice-edge corridor: a possible Palaeolithic route to the New World. World Archaeology, 36(4), 459-937.

Stanford, D. J., & Bradley, B. A. (2012). Across Atlantic Ice: The Origin of America’s Clovis Culture. Berkeley, California: University of California Press.

Straus, L. G., Meltzer, D. J., & Goebel, T. (2005). Ice Age Atlantis? Exploring the Solutrean-Clovis ‘connection’. World Archaeology, 37(4), 507-1039.

Tamm, E., Kivisild, T., Reidla, M., Metspalu, M., Smith, D., Mulligan, C., Bravi, C., Rickards, O., Martinez-Labarga, C., Khusnutdinova, E., Fedorova, S., Golubenko, M., Stepanov, V., Gubina, M., Zhadanov, S., Ossipova, L., Damba, L., Voevoda, M., Dipierri, J., Villems, R., & Malhi, R. (2007). Beringian standstill and spread of Native American founders. PLoS ONE, 2(9), e829.

Westley, K., & Dix, J. (2008). The Solutrean Atlantic Hypothesis: A view from the ocean. Journal of the North Atlantic, 1(1), 85-183.

  5 Responses to “Overselling the story? “Parallel genetic histories” of mtDNA C4c and X2 as ostensible refutation of the Solutrean hypothesis”

  1. Dienekes has repeatedly claimed that Lake Baikal formed the prehistoric boundary between Caucasoids and Mongoloids. If early West Eurasians really did venture that far east, do we really have to invoke Solutreans to explain the presence of Caucasoid genes in pre-contact Amerindians?

  2. The Minnesotan:

    Soares et al. (2009) estimate a coalescence time of 20.9 ka for X2, based on the value inferred for X2a, X2b, and X2c, and 31.8 ka for X as a whole (corresponding to the divergence of X1 and X2). Compare to the estimates of 27.4 ± 6.2 ka for X and 21.6 ± 4.0 ka for X2 calculated by Reidla et al. in 2003 using a more complete range of samples, but a less sophisticated mutation rate model (1 mutation per 5,139 years in the coding region).

    Most of the clade diversity of X2 is extra-American (with X2b-X2f distributed from Europe to West and Inner Asia and Siberia) and X1, which is basally oriented relative to the X2 branches, is exclusive to North Africa, the Near East, and East Africa. So even if X2 arrived in the New World via Beringia, the natural conclusion is that its journey ultimately began in the Near East.

    Even if its original bearers were diagnostically caucasoid, 20-30 ky is a very long time — more than long enough for a wandering mtDNA haplogroup to become dissociated from the morphology/genomic affinities of its “starting point”. Yes, the anthropometry and genetic analysis of South Siberian remains does point to an ancient West Eurasian presence in the area, but as far as I know no earlier than the Bronze and Iron Ages (see, e.g., Keyser et al., 2009).

    At any rate, Alexander Kozintsev (1999) points out that, “even today there is little hope of finding direct ancestors of Amerindians in Siberia, as human skeletal remains older than 10,000 years are quite scanty”. Some researchers highlighted cranial similarities between Neolithic/Chalcolithic Baikalians and Amerindians, but the most parsimonious explanation is that this is a mostly superficial similarity — genuine, if relatively late, caucasoid admixture in the former happened to approximate a deviation from “classical” mongoloid norms in the latter that had an entirely different basis.

    As Kozintsev explains: “Contrary to expectation, the earliest known inhabitants of the Baikal area, who lived there in the sixth and fifth millennia BC, were extremely flat-faced Mongoloids. Their descendants began to look less Mongoloid (and accordingly more Amerindian) only in the fourth millennium BC, and this tendency continued in the [3rd] millennium BC.”

    And, of course, more extensive Amerindian genetic studies ought to be conducted, but — as far as West Eurasian ancestry goes — it’d take ancient DNA results to convince me that what one sees in contemporary populations is anything other than post-Columbian admixture … just like how African ancestry in the contemporary inhabitants of Veracruz, Mexico, isn’t anything close to a reasonable basis for speculating about trans-Atlantic voyages by Malian ships.

  3. Hi, (I’m an anthropologist, but please don’t look up my site(s), they’ve been hacked – am working on fixing the problem)

    Some of the most remarkable genetic discoveries that have emerged in the last couple of years (and that have largely gone unnoticed) are the striking matches between Amerindian genes and Neanderthal ones. I’m talking about blood group O, the basal B006 lineage of human X chromosome identical to that of Neanderthals, a couple of HLA markers, etc. mtDNA haplogroup X shares a lot of mutations with the corresponding Neanderthal sequences. For example, human mtDNA 16278 and 16223 sites carry the same T allele in the so-called “haplogroup X” as the 3 Neanderthal sequences from Germany, Croatia and the Caucasus. Site 16189 carries a C in human “haplogroup X” exactly like in the 3 Neanderthal sequences. These sites are hypervariable in humans but fixed in Neanderthals. See Hagelberg 2003 “Recombination or mutation rate heterogeneity” for an interesting discussion.

    Notably, there seems to be some parallelism in the distribution of mtDNA X and X chromosome’s B006 (frequent in North America, absent in Siberia, present in Europe but less frequent than in the Americas). I wouldn’t trust the dates attached to mtDNA X2 or C4c. They very well may be relic lineages underdiversified because of some selective/ecological constraints. Neanderthals are also very uniform genetically – it doesn’t mean they are a young population. Add to this the high frequency of shovel-shaped incisors in both Neanderthals and American Indians and you start wondering…

    The Solutrean-Clovis connection may be a typological parallelism, not a shared innovation, but as such rooted in the same Middle Paleolithic pattern that’s much older than its most advanced archaeological manifestations (Solutre and Clovis). In Eurasia it was later replaced by the blade complex associated with Upper Paleolithic (Kara-Bom at 43,000 in South Siberia, Aurignacian in Europe, Dabban in North Africa, etc.) that didn’t leave a trace in the Americas. This Middle Paleolithic/Neanderthal pattern (centered on flakes) is so well documented and well-dated in Europe but it appears to be archaeologically recent in the Americas but without any roots in Siberia. At the same time, while it went completely extinct in Europe, American Indians have survived into the present.

    There’s something that we fundamentally misunderstand about this deep prehistory and the origin of American Indians. We’ve long assumed that we have all the facts but the most compelling and striking ones are only just coming up.

  4. Hi, this is to let you know that my website http://www.kinshipstudies.org is up and running.

    German Dziebel

  5. But who says mtDNA lineage X is “Caucasoid”? It’s root distribution is very different from West Eurasian mtDNA lineages. It’s part of the pre-Neolithic substratum in the Middle East. This concept may be too murky for those not familiar with the issue.

    BTW, the X in the Americas not being Solutrean is a long dead issue, has been since clearer phylogeny of X sub-lineages was mapped out years ago. The closest twig to the American X is a sample from Iran.

 Leave a Reply



You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>