Torroni and Wallace’s findings seemed to fit Greenberg’s three-entry model quite well. The immediate problem seemed to be the timescale. The young age of Group B even suggested that B might be independently signalling the post-glacial Clovis expansion in North America.37 The oldest ages of colonization suggested by the molecular clock, however, were supported by only minimal archaeological evidence and by very few archaeologists, but most of the younger dates were still older than Clovis. Taken as a whole, the new genetic evidence seemed to be moving even further away from a consensus with archaeologists and linguists.
As it turned out, the pre-Clovis genetic dates were the more durable conclusions of this major study. Subsequent re-estimates for the age of Group B in Amerinds have also tended to be pre-Clovis rather than post-Clovis. Research published since Torroni and Wallace’s 1993 results has produced similarly old genetic dates for the other three groups, A, C, and D, thus supporting a growing genetic perspective that all the main mtDNA lines had already entered America before the LGM. The older genetic dates have even raised the possibility that the Americas could have been colonized during a previous glaciation 30,000 or even 40,000 years ago.38
What has changed our perspective of American colonization more is detailed knowledge of the branching coming out of those founding branches, and advances in the methods used to date them. The insights and new methods that developed from this extensive and expensive enquiry into American genetic prehistory laid the ground for all the analytically much more difficult studies on Africa, Asia, and Europe that have provided much of the new body of knowledge of ancient migrations discussed in this book.
Greenberg’s three groups are related to one another within America
As more markers came to be used, the inner structure of the American mtDNA branches revealed a genetic link between the Na-Dene, the Inuit, and the Amerinds that Torroni and Wallace did not uncover in their earlier work. Initially they had thought that both Inuit-Aleut and Na-Dene speakers had only Group A markers and that this somehow indicated that they were descended from a separate migration event. It was also soon realized that Inuit-Aleut speakers had both A and D groups. Their D group was identified as a new subgroup of Asian D which had not previously been recorded in Asia or America: D2. The main American founder D type, D1, was partly related to D2 but unique to Amerinds.39
A much more convincing argument that Na-Dene and Inuit-Aleut speakers shared a common source with Amerinds came with the discovery that the American A cluster was actually composed of a unique founder, A1, and her daughter, A2 (and other daughters). A1 was a first-generation daughter of the Asian root A type, and A2 was A1’s daughter by one mutational event. While A1 was found only in Greenberg’s Amerind group, A2 was found in all three language divisions but was the only type in the Na-Dene and Inuit-Aleut. These complex inclusive and exclusive founder relationships carried a simple message: they bound together the three Greenberg groups as a ‘family’ in that they had at least one founding genetic source in common – the American founding line A. Neither A1 nor A2 are found in Asia, with the exception of Siberian Inuit and Siberian Chukchi. (The Chukchi language belongs to Chukchi-Kamchatkan, a unique group of Paleo-Siberian languages spoken in the far eastern tip of Siberia and the Kamchatka Peninsula.)40
These two Siberian peoples share a further unique subgroup of A2 with all other Inuit populations in both North America and Greenland, thus still keeping their family together, as it were by a network of related A types. This again seems to point to a single entry into the Americas. A similar network of subgroups related to one founder seemed at first to bind D subgroups throughout America. D2 was a feature of the Inuit-Aleut, while another related but unique subgroup, D1, was found almost only in Amerinds, who constitute the majority of Native Americans.41
These genetic findings appeared to strengthen the links between the three Native American groups, the two in the Arctic and Subarctic and the rest (Greenberg’s Amerind group) farther south. This seemed to be leading back to the reductionist idea of a single migration, but several puzzling questions remained. For a start, the most convincing link between the three language divisions, the A1/A2 family network, although it arose after the exit from Asia, still shows a very deep and ancient division. This split separates the Arctic and Subarctic groups from the rest of the Americans to the south. Not only that, but the A2 subgroup in the Inuit-Aleut and Na-Dene is much younger than in the rest of America.
Why so little diversity in the north? Must be the ice age
Another continuing difficulty that faces those trying to link the colonizations of the American Arctic and Subarctic with the peopling of the rest of the two continents, as a single migration from Asia, is the imbalance in distribution of the founder groups. According to the principles of founder analysis – and an evolutionary rule of thumb – the nearer you get to the source of a migration, the more likely you are to find all the founders. Conversely, the farther you travel from the start of migration, the lower the diversity becomes. In the Americas, we have just the reverse. Up in north-west Alaska, where the migration into the Americas must have started, there is now only one of the founder lines, A, present only as the derived type A2. Furthermore, B and C are completely missing from the whole of the Arctic and Subarctic. Conversely, in the rest of America, the founding Groups A to D are widely distributed and diverse. This seems all wrong. How could A, B, C, and D all be at large in the Americas when all we can find at the portal of entry in Alaska and neighbouring parts of Canada is a rather young A2 maternal clan?
In many ways, the high genetic diversity in the southern half of North America, even more diversity in South America, and the lowest diversity in the extreme north (Canada and Alaska) mirrors the upside-down distribution of language numbers. If Clovis-first was correct, and the first colonization occurred after the LGM, we would find the opposite. As I argued for the languages, if the primary colonization had occurred before the ice age, extensive parts of North America would have been depopulated during the LGM. If the North American population had re-expanded during the ‘big melt’, northern peoples would show lower diversity and consequently younger time depths than those farther south.
The idea of an expansion phase in North America after the ice age, which would explain some of the upside-down genetic findings, had already been put forward in 1993.42 Alaskan geneticist Gerald Shields and colleagues had noticed that all the circum-Arctic populations in Asia and in the Americas appeared to be genetically young, similar to one another, and different from populations farther south in both continents. In particular, the Group B was completely absent north of the 55th parallel. They argued that the circum-Arctic peoples could have originated in a more recent expansion of a single northern population lacking in genetic diversity. They also argued that this expansion had happened after the Americas had been colonized by a more diverse population. This view of populations with low genetic diversity expanding in Canada and Alaska, after the ice age and after a previous colonization of the Americas, appeared to go a long way to answer some of the genetic and linguistic conundrums posed by those regions.
A refuge on the lost continent of Beringia
There was still the problem of where these re-expanding far-northern groups came from – was it Asia or America? The new genetic evidence of the A1/A2 link seemed to point to America, while the presence of A2 in Siberian Inuit and Chukchi suggested a possible Asian source. Anyway, if the Na-Dene and Inuit-Aleut were from the same original genetic stock as all other Americans, how come there was such a difference in age, and such a deep genetic, physical, and linguistic split?
In 1996, the Anglo-German geneticist Peter Forster and an international team drew these problem strands into a coherent explanation. It was like a chess problem which had had a number of people puzzling over it for a long time: the answer, when it came, was so simple. Forster’s collaborators included Antonio Torroni and a German mathematician/polymath, Hans-Jurgen Bandelt, who has been the creator a
nd inspiration for much of the analysis necessary to create the real gene trees I have used in this book. The answer to the Americas conundrum was that the ice-age homeland of the northerners was neither Asia nor America, but another continent – Beringia.43 (Figure 7.1.)
Between 11,000 and 25,000 years ago, sea-levels were so low that the Bering Strait was then a land bridge between Asia and North America. But Beringia was more than that: it was a huge continent in its own right, 1.3 million square kilometres (500,000 square miles) at its maximum extent. Not only was it ice-free, but its grassy tundra supported herds of herbivorous mammals. The summers were for sure cooler than today, but the winters were, paradoxically, milder. For most of the period in which the land bridge existed, the ice caps persisted farther south and, as we saw earlier, the ice corridor was closed between 15,000 and 22,000 years ago, presumably preventing contact between the far north and the rest of America. The Siberian hinterland could hardly have been more inviting than North America, being an Arctic desert at this time, so Beringia and the western part of Alaska effectively became an ice-age refuge, cut off from both continents (Figure 7.4). With the Beringian refuge holding remnants of the original genetic founders of America, we can now see why the Na-Dene and Inuit-Aleut who, with their low diversity, seemed so different from the rest of the Americans, were yet linked to America through their A1/A2 group gene tree.44
Do the dates fit entry before the ice age?
So far, so good. The high-resolution mtDNA analysis seemed to be saying that a collection of founding mitochondrial lines had arrived in Beringia before the LGM and spread throughout the Americas. Then along came the ice age, cutting off the Canadians and Alaskans from the rest of the population farther south. In the process, the genetic diversity of those far northerners was severely reduced by the extreme privations and near-extinction resulting from trying to survive the ice age in Beringia. So, what about the genetic dates? Do they fit this new theory of one entry to the Americas before the ice age, with secondary re-expansion afterwards?
The answer is that they do, and surprisingly well. For the regions south of the ice caps, Forster and colleagues calculated the ages of A1, A2, B, C, and D1 in Amerinds from North America, Central America, and South America. Ten of the twelve groups analysed were all older than Clovis. To increase the precision of the estimate they then averaged the ages of all four groups (A–D) together in each of the three regions. North American Amerinds came out at 23,000 years, Central Americans at 16,000 years, and South Americans at 21,000 years. So there was a case for the founding lineages A1, A2, B, C, and D1 arriving before 21,000–22,000 years ago, the time when the ice corridor closed. When Forster and his team looked only at the derived A2 types in the Inuit-Aleut and Na-Dene, the expansion age came out at 11,300 years, suggesting a re-expansion around the time of Clovis (see Figure 5.5). The other insights these studies give us add to the strong impression that the Inuit cultures and peoples are a relatively recent phenomenon, but that they still owe their genetic heritage to the ancient founders of America.45
Figure 7.4 North America during the last glacial maximum. At the LGM, no further movement was possible and the ancestors of sub-arctic Americans were locked into a Beringian refuge with drift down to one line A2. South of the ice, people carrying groups A, B, C and D could continue to develop diversity in language, culture and genes.
Figure 7.5 Post-glacial mtDNA re-expansions in North America starting from 13,000 years ago. In Beringia, the tundra area has increased and access to the Aleutians and south along the west coast has opened, allowing group A2 to reexpand among the Inuit-Aleuts and Na-Dene. The ice corridor is not quite open. Group B can also expand along further down the west coast and into the southwest.
For my money, the results obtained by Forster and colleagues are among the most important in clarifying dates in the complex story of American colonization. They also contained the first clear genetic evidence for the Beringian re-expansion as previously suggested by archaeologists. It is worth summarizing the conclusions. The first is confirmation that there were four main American founder mtDNA types arising from the four Asian groups A–D identified by Antonio Torroni and colleagues in 1993: A1 (leading to A2), B, C, and D1. Second, these four founders probably arrived in America before Clovis and before the LGM 18,000 years ago. Third, the homeland for the movement of these founders into the Americas was Beringia, not Asia.
This last point may sound like splitting hairs (or continents), since Beringia must ultimately have received its people from Asia, but this is not a fine point. What Forster and his colleagues are saying in effect is that from some time after it was first colonized, before the last ice age, until maybe the last thousand years of its existence, Beringia was effectively cut off from Asia and received no further Asian founders. The importance of this distinction will become apparent shortly. Beringia was the source of American lines both before the LGM (Amerinds) and during the Arctic/Subarctic re-expansion by the Na-Dene and Inuit-Aleut. The post-glacial re-expansion from Beringia only included A2 presumably because the ice age caused a near-extinction in the north, and a so-called genetic bottleneck that resulted in only one genetic line surviving.
Others have since come up with essentially similar scenarios or variations on the theme of an early migration(s) into the Americas taking in most if not all the founders before the last ice age.46
The X line
Before we look at what the evidence has to tell us about whether this preglacial invasion was single or multiple, there is one more maternal founder line to consider. The fifth founder had an enigmatic identity, which has only fully emerged in the last few years. Appropriately, the name chosen for this extra founding line was X.
In 1991, Ryk Ward came up with four clusters. Two years later, Torroni and colleagues identified their own groups A, B, and C in three of Ward’s clusters and resolved the remaining types partly into their new D group; but they left the rest of Ward’s fourth cluster in an unresolved ‘others’ bin. Over the next couple of years, more research groups came to see that this other group was, as Ward had originally suggested, another cluster with a single origin – in other words, a fifth American founder line. Among them were Argentinian geneticist Graciela Bailliet and collaborators, and also mathematician Hans Bandelt and colleagues, who noted the likelihood of a fifth founder in connection with his new method of genetic tree-building. Forster’s re-analysis of American founders in 1996 defined and renamed this new American clan the X group. This confirmed that the fifth American founder and the European X group which Torroni identified in the same year had the same common X ancestor.47
The American X group had a limited, northerly distribution, mainly in two separate Amerind-speaking cultural regions around the 50th parallel. The first of these cultural groups, in the Pacific north-west, were the Yakima of Washington State and the Nuu-Chah-Nulth. The latter is a fishing culture found on the western coast of Vancouver Island (and also on the Olympic Peninsula of Washington State) with a cultural continuity of 4,000 years. These groups had rates of 5 and 11–13 per cent X, respectively. The other northern groups holding X types were the Sioux and the Ojibwa of the eastern woodlands around the western edge of the Great Lakes. The latter had rates of 25 per cent.48
Could these X types in Native Americans have come from European female lines entering the tribal populations recently? Almost certainly not, since X was already present in pre-Columbian peoples of these regions. Ward’s original study had carefully excluded the possibility of significant European admixture in the Nuu-Chah-Nulth. In the case of the eastern woodlands Native Americans, direct evidence of X in pre-Columbian populations was provided by an analysis of ancient mtDNA from the prehistoric Oneota population of the nearby Illinois River region.49
Kennewick Man: a pre-Columbian European connection?
This transatlantic DNA cross-match between Europeans and Native Americans has helped to fuel a chronic, rather dubious debate on who got to America first. The bone of conte
ntion is the problem of skeletal resemblances between ancient Americans and Europeans. Skulls of ancient Americans dating from 5,000–11,000 years ago have been studied for decades, and are known to be different from the skulls of later inhabitants, but it is not clear why they are different.
In 1996, the same year as the X founder was given her name and was linked with Europeans, two young men found a skull in the Columbia River at Kennewick, Washington State. American physical anthropologist James Chatters was called in by the coroner and was able to retrieve more of the skeleton. Embedded in the pelvis was a stone spear tip typical of the location and the period 4,500–8,500 years ago. The obvious antiquity of the leaf-shaped serrated stone point prompted the coroner to order radiocarbon and DNA studies.50
From the start, it was obvious that the skull did not look like a typical modern Native American. This man, who had died aged between forty and fifty-five, had a long narrow skull with a narrow face with a jutting chin, and looked more like a European than the broad-faced, broad-headed individual that might have been expected. James Chatters’ professional reconstruction (see Plate 26) looks strikingly like Patrick Stewart of Star Trek fame, although maybe it is the baldness of the reconstruction that enhances this impression. The trouble is that Kennewick Man was not Caucasoid. Some of his features can be described as Caucasoid, and others as Asian. The features that make Kennewick Man look like a European were seized upon by the press – and by white supremacists eager to claim that America had been colonized by whites a long time ago.
Out of Eden: The Peopling of the World Page 30
