DNA USA

by Bryan Sykes

  Then as now, the rivers of the western seaboard of America from Oregon to Alaska were conveyor belts of fish heading upstream to spawn. I once saw this spectacle myself on the Columbia River, where a fish ladder had been built to bypass a hydroelectric dam. Through the glass of the viewing gallery, where lampreys held on with their sucker mouths and edged up against the current, the silver shapes of Coho salmon appeared, paused for a few seconds, then continued their relentless journey to the spawning grounds with a powerful flick of the tail. These were a good two and a half feet long, weighing maybe seven or eight pounds. Then, suddenly, an enormous fish appeared, well over thirty pounds, I would guess. It stayed for a second or two and then it was gone, leaving me with the feeling that I had only imagined it. Then another came and went. And another. This was the King salmon, the monarch of the river.

  Such wealth as this sustained the Indians of the Pacific Northwest for at least four thousand years before they first encountered Europeans. The stability of the food supply, with dried fish filling the gap between annual spawning runs, made for the exuberant flowering of a culture for which the region is still known today. Relieved of the burden of being always on the move and having to carry everything from one camp to the other, their dwelling houses and their ceremonial totem poles became larger and more elaborate.

  It was one of these tribes, the Nuu-Chah-Nulth, which became the first to come under the scrutiny of DNA, and in so doing became the tribe against which all others would come to be measured. The Nuu-Chah-Nulth—the name means “all along the mountains”—live nowadays on the western coast of Vancouver Island and on the Olympic Peninsula of Washington State, a hundred miles south of Seattle. There are fifteen surviving groups within the Nuu-Chah-Nulth, far fewer than before European contact and its dreadful legacy of smallpox that so reduced the population. The first European to reach the Nuu-Chah-Nulth was the British navigator and explorer James Cook in 1778. He was on his way to try to find the Northwest Passage rumored to connect the Pacific and Atlantic oceans. Cook sailed on through the Bering Strait, but his way was blocked by sea ice. Ironically, had he tried the same voyage today, he might have got through thanks to the melting effects of the current phase of global warming.

  Today the Nuu-Chah-Nulth number around nine thousand. As well as being experienced fishermen, they have also in the past been skillful whalers. Along with other nearby tribes, they are enthusiastic participants in the potlatch ceremony, where hosts distribute lavish gifts in surroundings of celebration, music, and dance. Although potlatch was banned by the Canadian and U.S. governments in the late nineteenth century, the legislation was eventually repealed in 1951, and the ceremony soon regained its earlier popularity. Between 1984 and 1986 a large proportion of the Nuu-Chah-Nulth from Vancouver Island were recruited by scientists from the University of Utah as part a survey of blood groups and other biological markers. After the project had been completed, the blood serum was frozen, and it was this stored material that was used for the very first analysis of Native American mDNA sequences. It was also a minor technical triumph in that retrieving DNA from stored serum, which does not contain blood cells, had not been tried before. True, mitochondrial DNA had been analyzed five years earlier from a group of Pima Indians living along the Gila River valley near Phoenix, Arizona, but the data were confined to only a handful of genetic markers rather than DNA sequences themselves, where far more information resides.

  Sixty-three individuals were chosen who were known, from the extensive genealogical records collected by earlier researchers, not to be maternally related. Had they been, their mDNA would have been automatically identical and would have skewed the interpretation. What the Nuu-Chah-Nulth mDNA revealed was that there was far more genetic variability within the tribe than had been suspected. There were twenty-eight different sequences among the sixty-three Nuu-Chah-Nulth volunteers whose mDNA had been analyzed. At one extreme the most frequent of these was shared by nine individuals, while at the other, thirteen people had mDNA sequences found in nobody else in the group. Nonetheless there was a pattern. The Nuu-Chah-Nulth fell into four clusters within each of which their sequences were related to one another. The biggest of these clusters was made up of twenty-eight individuals, all of whom shared mutations at position 111 and 319, variations that were absent from the rest of the group. So they had to be related to each other, and they must have inherited these particular mutations from a common maternal ancestor. There were plenty of other mutations among the people in the cluster, but most of them were confined to a few individuals.

  The three other mDNA clusters had similarly shared sequences, and some differences in detail. In one cluster everyone shared a mutation at 278, while the mDNA from members of a third cluster all shared a variant at base number 325. All three of these clusters were united by the crucial mutation at 223, of which more later. The fourth cluster, however, did not have the variant at 223, but instead shared mutations at 189 and 217.

  When I read this paper soon after it was published in 1991,1 my eye caught this last detail. I had only recently returned from my first visit to Polynesia and was working through the analysis of the twenty Cook Islanders whose DNA I had brought back with me to Oxford. Nineteen of the twenty also shared these two mutations. This had to mean they were in some way maternally related to the Nuu-Chah-Nulth, but whether it meant that these Polynesians were descended from Pacific Northwest Indians, or the other way around, took some time to work out. The connection was confirmed a little later when the DNA from both the Cook Islanders and the Nuu-Chah-Nulth was shown also to lack a small section of DNA, nine bases long, which had been deleted from the mitochondrial circle. This deletion was extremely rare elsewhere, and combined with the sequence match, could lead to only one conclusion: They were definitely related.

  The message from the Nuu-Chah-Nulth paper was powerfully clear. This group of American Indians from the Pacific Northwest contained not one but four separate clusters of mDNA, each one of which contained masses of individual differences. This was not the DNA picture from a single recent migration of just a handful of people, as had been suggested. The sheer variety of sequences showed that the ancestors of the Nuu-Chah-Nulth were both numerous and ancient. But how old? Fortunately that was a question that could be answered, although exactly how to go about it took a few years to perfect. And the next question was this: If the ancestors of the Nuu-Chah-Nulth had migrated from somewhere else, where was the source? And what about Native Americans from other parts of the United States, and indeed from South and Central America as well? The rush was on to find out.

  Over the next few years scientists pursued the genetic origins of Native Americans with innocent enthusiasm, rarely if ever pausing to wonder what their subjects felt about it. But these were the golden years before such considerations became sophisticated and serious. The clouds of concern lay well beyond the horizon at the time. The work was technically straightforward, and as long as you could get your hands on some DNA, the rewards were both interesting and substantial. Peer-reviewed papers, the medium through which all modern scientists have come to be judged, flowed from the labs with regular ease. The extremely well-regarded American Journal of Human Genetics became almost the trade paper of the new breed of genetic anthropologists. Throughout the 1990s, papers on the genetic origins of Native Americans, along with other indigenous peoples from all over the world, were published in almost every issue—at least that’s how it felt at the time. I enjoyed publishing my own research papers on Polynesia and Europe in the American Journal of Human Genetics because, unlike Nature and Science, where articles were pared down to an absolute minimum, the American Journal (as it was known to all of us) gave you the space to present the detail of what you were doing and to discuss the conclusions at reasonable length.

  As far as the research on Native Americans went, before long the kind of study that had been done on the Nuu-Chah-Nulth had been replicated in populations throughout North and South America. From the Inuit of Ala
ska in the North to the Mapuche of Chile, dozens of papers containing hundreds of individual mDNA results were published, many of them in the American Journal. I remember eagerly awaiting each new issue, between the familiar green-and-gold covers, as the librarian put out the copies on display. What treasures did this month’s edition contain? There was almost always something.

  It was very soon clear that a pattern was emerging from all the effort being put into Native American DNA. First of all, the four clusters seen in the Nuu-Chah-Nulth were also found in virtually all the other studies, though their proportions varied considerably depending on where the DNA samples were from. By then a new vocabulary was emerging to identify these mitochondrial clusters and, for the simple reason that Native Americans were the first populations to be thoroughly studied using the new genetic tools, they were allotted the first four letters of the alphabet: A, B, C, and D. To emphasize that they were real women, and not just theoretical constructs, I gave them each names that, following the convention of The Seven Daughters of Eve, began with the initial of the cluster’s scientific notation. The clan mother of cluster A became Aiyana, Chochmingwu was the name I gave to the founder of cluster C, while Djigonese was the clan mother of cluster D. For reasons I will explain in a moment, the founder mother of clan B broke with this convention, and I called her Ina.

  There was reason to believe that three of these clusters were related to one another because they all possessed the important mutation at position 223. The exception was cluster B. The same logic of a shared maternal ancestry applies to the four American clusters just as it does to the seven European clans. Each cluster has just one woman as its matrilineal ancestor. In Europe I was able to estimate how long ago these seven clan mothers lived by extending the reasoning we saw earlier when comparing two individuals—that is, by averaging the mutations among all individuals within each cluster and multiplying this figure by the mDNA mutation rate. The result of this calculation showed that the seven native European clan mothers lived at different times between ten and forty-five thousand years ago. The same calculation applied to the Native American clusters gave clan ages of between just over thirty thousand years for cluster D and just under ten thousand for cluster C.

  A direct extrapolation from these genetic cluster dates to the timing of the first settlement of America appeared to support the idea that members of cluster D had arrived well before the time of the Clovis period around 11,500 years ago. Taken at face value, there was even a possibility from the genetic results that they had arrived as far back as the early and controversial lower levels at Monte Verde, which returned carbon dates of more than 30,000 years. However, things are not quite that simple. The additional consideration is that the only mutations that are relevant are those that happened after the mDNA got to America. Any mutations that occurred in matrilineal ancestors before they got to America might tell you something about what was happening in the ancestral homeland, but not in America itself.

  To illustrate the point, imagine trying a similar analysis on the Europeans who arrived in America within the last five hundred years. All seven European ancestral clans are represented, and there are lots of individual differences within the clans. But to apply the same method of averaging the mutations within each clan and multiplying them by the mutation rate would give ages for the clans that were pretty much the same as they were in Europe. It would be completely misleading to use these figures to conclude that Europeans arrived in America between ten and forty-five thousand years ago.

  After several years of collecting and analyzing Native American mitochondrial DNA, scientists eventually developed a way of distinguishing between ancestral founder sequences and derived sequences brought about by mutations that occurred in America after settlement. Only the derived sequences—mutations that had happened in situ—could be properly counted toward arrival time estimates. If a sequence could be found both among Native Americans and in people living today in an ancestral homeland, then it could be assumed that this was one of the founder sequences that should be discounted from the age estimates. If, on the other hand, a sequence had clearly evolved by mutation from a founder sequence, and was found in America but not in the ancestral source population, then it should be included in the age estimate. Founder analysis, as this method became known, is a tricky business that, for a start, requires that you know where the appropriate ancestral regions are.

  Ever since the Spanish scholar José de Acosta first proposed an Asian origin for Native Americans in 1596, historians and anthropologists have thought of northeast Asia as the most likely source of the first American settlers. The popularity of this hypothesis only increased when Vitus Bering showed how close the two landmasses really were when, in 1728, he sailed from Siberia to Alaska across the strait that now bears his name. Although there were other theories, as we shall see, there were a number of reasons for making a connection between Native Americans and northeast Asians, and these include some similarities in facial features and tooth shape. The blood-group frequencies of Native Americans and northeast Asians are also similar, so it came as no great surprise when collaborations between Russian and American scientists got under way as Cold War inhibitions thawed, that there were general similarities in the mDNA clusters found among the two groups. The genetic results from both continents were gathered together and put through a founder analysis, individually scrutinizing each Siberian and Native American sequence to identify, as far as possible, when an American mDNA had evolved in America rather than in Asia. This process had the effect of setting the genetic clock to zero so that the age since settlement could be calculated from mutations that had accumulated in situ rather than in the source population—which by its very nature must be older. As my research team2 and I discovered when doing something similar in Europe, founder analysis is a very laborious process. To do it properly a large amount of data are needed from the source population, far more than from the derived, so as to reduce the chances of missing founder sequences.

  Among Native Americans, calibrating the genetic arrival dates by founder analysis had the effect of making them more recent than the initial, uncalibrated estimates. The first attempt, in 1996, narrowed the range of genetic settlement dates in North America from 10,000 to 30,000 years to between 18,000 and 25,000 years.3 The most recent set of calibrated estimates was published in 2009.4 By then it had become almost routine to sequence the entire mitochondrial genome of more than sixteen thousand DNA bases rather than the short four-hundred-base control region. This extra sequencing certainly showed all possible mutations, but the richness of variation retained in the control region meant that the gain was not proportionate. The increase in mutation data and in the number of individuals involved ironed out the more extreme age estimates from the first founder analysis published thirteen years previously, and gave dates for the clusters ranging from 15,800 to 19,600 years (details are in the appendix). I think we will have to settle for that. In my view there is little scope for improving the accuracy of these estimates with further work.

  Insofar as genetic and carbon dates are remotely comparable, it seems to me that these calibrated mDNA dates suggest that the first Americans did arrive before Clovis. The same studies that identified founder sequences also established beyond reasonable doubt that Siberia was the original entry point for the majority of ancestors of today’s Native Americans. However the research also uncovered a significant surprise. While a Siberian origin for clusters A, C, and D is clear and the founder sequences identified, the same is not true for the fourth cluster, B.

  Cluster B is my favorite because it is predominant in Polynesia, where I began my work with mitochondrial DNA twenty years ago. I don’t have to look up the main Polynesian control region sequence—it is etched on my memory: 189 217 247 261. Believe it or not, even now, when I come across this sequence during my current research projects, I am back on the white sand beaches of Rarotonga, in the Cook Islands, gazing out to sea across the surf line as the Pacific Ocean
breaks on the reef. Three thousand years ago the very first Polynesians arrived here from the islands of Indonesia across hundreds of miles of open ocean. It was, I believe, the greatest feat of maritime exploration in the history of our species. I must be careful not to go on about this too much, and anyway I have already written extensively about my time in Polynesia in The Seven Daughters of Eve. Suffice it to say that the experience completely altered my view of the accomplishments of our ancestors. I no longer saw them as the savage equivalent of ourselves, but as fearless and resourceful pioneers who survived the many trials of raw nature that would defeat almost all of us living today.

  Tracking the origins of the Polynesians to Asia, and not to America as the Norwegian anthropologist Thor Heyerdahl famously argued, was fairly straightforward. I could see that the founder population was from Taiwan, where the sequences 189 217 and 189 217 261 were found among the aboriginal Taiwanese. You will forgive me, I hope, for breaking with convention and naming the clan mother of cluster B after the mythical Polynesian princess Ina, who was carried to her lover’s floating island by a shark and whose image adorns the Cook Islands banknotes.

  I scanned the mitochondrial sequences of Native Americans as they were published, eager to see if my heroines had managed to cross the entire Pacific Ocean and had taken part in the colonization of America, but I never did find the Polynesian motif. It was too far, and three thousand years ago was too late in the day, for the Polynesians to make any serious impact. We do know that they reached South America and returned with the sweet potato, a decidedly Andean crop, and also that large numbers of Polynesians were enslaved in the nineteenth century and taken by force to Peru.5 But they did not play a significant part in the settlement of America, with the notable exception of Hawaii, where the unmistakable Polynesian sequence is still to be found among the diminishing number of native Hawaiians.