The Na-Dene Have Paleo-Eskimo Ancestors

The Na-Dene Language Area (via Wikipedia)

Background

The Na-Dene are North American Native America/First Peoples tribes who share a language family that is sufficiently distinct that Greenberg when classifying the languages of the indigenous peoples of the Americas treated the Na-Dene languages (along with the Inuit languages) as the only languages which he did not include in an Amerind macro language family.

Most Na-Dene tribe are found in Alaska, and the Pacific Northwest of Canada and the United States, but the most famous, including the Apache and the Navajo, are found in the American Southwest. Archaeological and linguistic estimates suggest that the Na-Dene of the American Southwest arrived there around 1000 CE.  The earliest archaeological remains definitively considered Na-Dene date to about 1500 BCE, but with considerable uncertainty regarding the classification of earlier relics which might or might not have been related.

After Greenberg made his linguistic classification, genetic evidence corroborated his conclusion to some extent and demonstrated that Inuits and at least some Na-Dene were somewhat distinct from other indigenous populations of the Americas.

Canadian archaeological research also revealed that today's Arctic Inuit people (colloquially known as Eskimos and sometimes also called "Neo-Eskimos" in the anthropology literature) are in cultural continuity with an archaeological culture called the Thule which appeared in Alaska around 1000 CE and migrated east from there across Northern Canada (around the same time that the Na-Dene migrated to the American Southwest and around the time that the Anazasi a.ka. Ancient Puebloans saw their civilization collapse in a major drought).

Before then, two successive archaeological cultures of "Paleo-Eskimos", the first called the Saqqaq, who were followed by the Dorset culture were predominant in the region that the Inuits occupy today. Ancient DNA recovered from Paleo-Eskimo remains have made it possible, first with mtDNA and later with whole genomes, to clarify the genetic relationship between the members of these cultures and later ones.

Also after Greenberg's work, subsequent linguists have found that the Na-Dene languages derive from an Alaskan origin and also appear to be part of the same language family as the Central Siberian Yenesians, of whom a few thousand Ket people are now the sole surviving representatives.

One burning questions, arguably the biggest in the anthropology and prehistory of the New World, has been, who are the ancestors of the Na-Dene and when did they arrive?

In many respects, the Na-Dene are similar genetically to the rest of the Founding population of the Americas which entered North and South America from a refugium on the Beringian land bridge around 14,500 years ago.  And, since the Na-Dene mostly live close to where the Founding population entered the Americas, any genetic distinctiveness could have been due to an early wave of serial founder effects and population structure in the founding population.

Another possibility was that the Na-Dene people arrived in a subsequent wave of migration of their own, followed by some amount of admixture with existing North American populations that was distinct from the Paleo-Eskimos or the Inuits.

A third possibility is that the Na-Dene people are derived from one or more of the Paleo-Asian or Inuit populations.

These questions have now been answered.

The New Results

We now know that the vast majority of Native American ancestry comes from the Founding population, supplemented in the pre-Columbian era mostly by a Paleo-Eskimo wave ca. 2,400 BCE, and an Inuit wave in the first millenium CE.

Using a relatively novel method of focusing in rare genetic variants, a new study has fairly conclusively established that many Na-Dene individuals have significant (up to 30%) ancestry from Paleo-Eskimos admixed with a majority of founding population genes and a smaller contribution of Inuit admixture in some individuals, that the Saqqaq and Dorest derive from a single wave of migration that arrived in North America ca. 2,400 BCE, and that the Paleo-Eskimo wave of migration from a Trans-Baikal population that had a genetic affinity to populations that ended up in Western and Central Siberia, where the Yenesians live.

Previous studies using methods based on average genetic similarity, uniparental markers and common genetic variants, had been unable to find the subtle genetic connections between the Na-Dene and Central Siberian populations, due to roughly 75 generations of dilution with neighboring populations on each side of the Pacific. 

The Founding population of the Americas has a roughly 45% ancestry from a Siberian population in the same general region, with the other 55% or so derives from a population closer to Southeast Asians than to Siberians, from about 14,000 years earlier, also complicated the task because the Paleo-Eskimo wave also had ancestry from both of the regions that the Founding population of the Americas did (the fact that Native Americans have a mix of East Asian ancestry and Siberian ancestry was apparent almost immediately after the first studies of contemporary Native American mtDNA and was confirmed when whole genome data became available). (The Soultrean hypothesis continues to be completely discredited.)

The conclusion that the Na-Dene have Paleo-Eskimo ancestry confirms one previous study based on ancient DNA which had estimated a lower percentage (about 16%) of non-Founding population ancestry in the Na-Dene.  But, the older study isn't strictly inconsistent with the current result, as some individuals in Na-Dene tribes have essentially no Paleo-Eskimo ancestry, presumably due to marriages and inductions of non-Na-Dene Native Americans into their tribes, while most have significant Paleo-Eskimo ancestry.

Conventional wisdom regarding the Inuits was confirmed but with some precision regarding how much admixture there is between Inuits, Na-Dene populations, and other Native Americans.

There were technical issues involved in parsing out post-Columbian European admixture, but there was no evidence for any points of entry for any Native American population (including the Paleo-Eskimo and Neo-Eskimo waves) from any place other than Alaska.

The study also places the ancestors of the Paleo-Eskimos in the larger narrative context of the prehistoric archaeological cultures of Siberia in a manner that explains a pre-New World migration major admixture event of these ancestors around 5,000 BCE that blended an East Asian population with a Siberian one.

The study and its abstract are as follows (hat tip to Eurogenes): 

Prehistory of Native Americans of the Na-Dene language family remains controversial. Genetic continuity of Paleo-Eskimos (Saqqaq and Dorset cultures) and Na-Dene was proposed under the three-wave model of America's settlement; however, recent studies have produced conflicting results. Here, we performed reconstruction and dating of Na-Dene population history, using genome sequencing data and a coalescent method relying on rare alleles (Rarecoal). We also applied model-free approaches for analysis of rare allele and autosomal haplotype sharing. All methods detected Central and West Siberian ancestry exclusively in a fraction of modern day Na-Dene individuals, but not in other Native Americans. Our results are consistent with gene flow from Paleo-Eskimos into the First American ancestors of Na-Dene, and a later less extensive bidirectional admixture between Na-Dene and Neo-Eskimos. The dated gene flow from Siberia to Na-Dene is in agreement with the Dene-Yeniseian language macrofamily proposal and with the succession of archaeological cultures in Siberia.

Flegontov et al., Na-Dene populations descend from the Paleo-Eskimo migration into America,bioRxiv (September 13, 2016).

Major Open Questions

What remain the big unanswered questions about New World genetics and prehistory after this study?

1.  Was there population structure in the Beringian refugium population that is discernible in modern First Peoples genetics?

There are genetic variants (e.g. several sub-halogroups of mtDNA haplogroup X2b - others kinds of mtDNA X2b are European) which are present in many non-Na-Dene/non-Inuit tribal and linguistic groups of North America (and sometimes in the Na-Dene as well) that are almost entirely absent from indigenous South American populations. 

This is easily explained with founder effects in a population that took a fast track along the Pacific Coast to South America, and by the subsequent lack of interaction between the populations.  But, some of the genetic variants also show regional/linguistic group variation within North America and seem to be correlated with each other to some extent.

These populations, unlike the Na-Dene and Inuits, overwhelmingly have pure "Founding population" ancestry.  

But, do these division also merely reflect founder effects, or do they reflect a deeper population structure in the Beringian refugium population? 

2. How are the Amerind Languages related to each other?

It is reasonable to assume that the Berginian refugium population spoke languages that all belonged to a single language family (or perhaps a couple of language families) around 16,000 years ago when they entered the Americas (after having been isolated from mainland Asia for thousands of years).

Most North American indigenous languages (about 270 out of 300) have been classified by linguists into 29 language families and meaningful progress has been made in recent decades at classifying these language families into larger language super-families even as the vast majority of these languages have gone extinct or become moribund.  Progress has also been good in understanding the connections between Meso-American languages.

But, far less progress has been made in South America, despite the fact that a far larger share of these languages are living languages (about 300 living languages and four times as many extinct languages) that are merely threatened with extinction.  A large share of all South American languages are isolates, and the languages families that have been identified tend to be small (i.e. five or fewer languages).

Is it possible to build larger language families?  

Is it possible to reconstruct the prehistory that caused the larger language families to be able to expand and to draw conclusions about the proto-languages of these language families? Larger language families generally reflect the success of related groups of archaeological cultures relative to their peers, and the diversity of languages within a language family can also often be used to crudely estimate when the proto-language of the language family was spoken.

For example, are there linguistic traces of the urban civilization that existed for a time in the Amazon or the very geographically expansive Mississippian culture

Realistically, it is impossible to reconstruct the original proto-AmerInd languages at this time depth, particularly in the absence of any written language that could convey information about the oral language in the vast majority of cases. Mayan and perhaps a few others languages of pre-Columbian farming cultures had a written form, but even in these cases, many of the writing systems convey no information about the spoken languages.  But, there is some slight evidence of some linguistic structure in the Founding population of the Americans, based upon the distribution of glottal phonemes in the Americas.

Indeed, the big lesson to be drawn from the AmerInd languages is that over the course of 14,500 years of isolation, almost every language feature known to humanity will arise independently in all manner of combinations despite a common or only modestly diversified starting point.  Equally important, what ecological circumstances drive the size and diversity of languages and why do the languages of the New World seem more diverse than the seemingly much older click languages of Africa, or the indigenous languages of Australia?  The experience of the Americas seems more akin to the diversity seen in Papua New Guinea (despite the fact that it independently developed agriculture thousands of years ago) than the relative lack of linguistic diversity found in Africa and Australia.

3.  What more can we discern about the routes by which the "virgin territory" of North America was settled?  

For example, it appears that the Clovis culture (which was not the first archaeological culture Native Americans in North America and occurred after all of North America had been populated by a population arriving from Alaska) progressed from east to west, which is not what one would naively expect. (This question is not unrelated to the previous two).

Similarly, did all South Americans make their way there via Central America in the Founding era, or were there major migrations into or out of South America later on and/or by different routes? By what routes was the part of South America to the east of the Andes settled?

In general, how static or mobile were the populations of the Americas in the pre-Columbian era?

4.  Was there a modern human population in the Americas that predated the founding population that is ancestral to the vast majority of modern indigenous people of the Americas?

There are a few sites near the Atlantic coast of South America and the American South that are claimed to predate the earliest confirmed Native American presence in the Americas by several thousand years.  But, there is no evidence of mass extinctions at that time, and there is no archaeological trace that old of hominins of any kind in intermediate locations necessary to reach those sites.  And, it isn't unprecedented for original dating estimates to be substantially revised later due to methodological issues (like the use of very old wood by people who come along thousands of years later, or disturbances that reshuffle the strata in which relics are located).  This has happened in key Neanderthal sites in Europe, in Indonesia, and elsewhere on many occasions in the last few decades.  

In an arguably related issue, some of the earliest pre-Clovis Native American skeletal remains (Paleo-Indians) are phenotypically distinct (and arguably archaic) relative to Clovis and later Native American skeletal remains. Ancient DNA has been recovered and tested from many of these remains, however, and revealed absolutely typical founding population DNA.

Finally, there is a single tribal group of indigenous people deep in the Amazon near the eastern foothills of the Andes Mountains who have a trace amount (2%) of ancestry in their whole genomes (but no anomalous uniparental markers) that looks "Paleo-Asian" (similar to the Onge people of the Andaman Islands, for example) rather than like the whole genomes of all other Native Americans descended from the founding population of the Americans.

These disparate and inclusively pieces of evidence have been used to argue that there was a small wave of modern humans who arrived in the Americas thousands of years before the Founding population of existing Native Americans, which didn't thrive, lived mostly along the warmer areas of the Atlantic Coast, and was ultimately almost totally replaced apart from a tiny trace bit of autosomal DNA in a relict tribe that escaped not only contact with Europeans but with everyone else for millennia.

5. Is Our Catalog Of Minor Pre-Columbian Contacts With The Americas Complete?

There are a variety of proposed pre-Columbian contacts with the Americas after the Founding population arrived, some more credible than others.  All of the credible claims pre-date Columbus by only about 1,200 years or less (many dating to just a few centuries before Columbus), with one exception.

The exception is one study from 2009 which claims to have radiocarbon dated a New World custard apple found in India to ca. 2000 BCE. (Pokharia, Anil Kumar; Sekar, B.; Pal, Jagannath; Srivastava, Alka (2009). "Possible evidence of pre-Columbian transoceaic voyages based on conventional LSC and AMS 14C dating of associated charcoal and a carbonized seed of custard apple (Annona squamosa L.)". Radiocarbon. 51 (3): 923–930.).  But, there are a variety of ways this outlier result could be explained. For example, the claim that the seed came from India rather than arriving from the New World could be false, or the seed could have been very similar to a custard apple seed, but was actually from an Old World species that might be extinct now.

There are now two verified sites in Canada where Vikings settled before dying off in pre-Columbian times around 1000 CE. They left no genetic trace in the Americas and there may have been some passing Inuit contacts with the Norse a century or two later after the Norse colonies were abandoned.

There is botanic evidence of at least one round trip contact between the Pacific Coast of South America and the Austronesians which can be narrowed in time by the earliest archaeo-botanical evidence of domesticated South American plants in Oceania (the South American sweet potato had reached the Cook Islands by 1000 CE based on radiocarbon dating), and by the earliest dates at which Austronesian mariners had gotten within sailing distance of South America (probably not much earlier than ca. 300 CE, the earliest possible date for human settlement of Rapa Nui a.k.a. Easter Island).  This is corroborated by DNA evidence.  Easter Islanders have about 8% Native American DNA dated to roughly 700 years ago, and there is also evidence of genetically Polynesian individuals in Brazil who did not plausibly arrive there with European intervention. Also, Peruvian mummy was embalmed with materials only available in Oceania around 1200 CE.  In short, the case for sustained pre-Columbian contact between South American and the Polynesians from sometime after 300 CE is increasingly pretty strong.

There is some archaeological and legendary history evidence of a two hundred year long South American coastal city state dynasty (ca. 900-1100 CE) that may have involve a small number of Asian families that arrived in the pre-Columbian era. This would be more plausible if it was a Polynesian connection adding to the list of other contacts, rather than a truly Asian one, but only further investigation will get to the bottom of it.

Some of these contracts remain to be confirmed or ruled out, and a few other trace contacts may be discovered some day.

Another partially related question is that while first contact with Europeans undeniably led to pandemic illnesses that decimated pre-Columbian populations, often advancing ahead of the Europeans themselves, the lack of continuity between ancient DNA from people who lived in the Pacific Coast of South America and modern DNA in those same places still seems to great to explain by this means alone (citing this paper).

So, the question arises, were there additional processes that led to this apparent genetic discontinuity between ancient and post-Columbian populations of the Pacific Coast of South America, and if so, what were they? For example, could Polynesian or Asian mariners brought previous pandemics to the region upon their first contact? Or, did the genetic homogeneity of South America (which due to founder effects had even less diversity and a lower effective population size than North America), make it particularly vulnerable to pre-Columbian pandemics? 

6. Why did mass extinction occur in the Americas and Australasia, but not in Southeast Asia or Africa to nearly the same extent?

What about the Founding population of the Americas and the fauna there made mass extinction in the Americas so profound.

7. Are Missing Insights Hiding In The United States?

Native Americans in the United States have been highly skeptical of anthropological investigations and DNA studies.  As a result most data from New World indigenous populations is from outside the United States or is not widely available to non-specialists. There could be new insights to be found in these populations, such as novel private North American genetic signatures or evidence (or a lack of evidence) of population structure and/or language shifts.