Who We Are and How We Got Here

by David Reich

  There is a general issue here about the ethical responsibilities of genetic research. When I examine an individual’s genome, I learn not only about the genome of the individual, but also about those of his or her family, and ancestors. I also learn about other members of the community—other descendants of those same ancestors. What are my responsibilities here? What do I owe not only to close relatives of the individual I study, but also to other more distantly related members of their family, to their population, and to our species as a whole? An extreme position that everyone needs to be consulted would make scientific progress in human genetics (including genetic medicine) nearly impossible. There would not be enough time for scientists in modest-sized laboratories like mine to talk with every tribal group that might be interested in the work.

  My own perspective is that we need as a scientific community to arrive at a middle ground, an approach that does not require obtaining permission from every possible interested group or tribe. On the other hand, given the well-founded concerns of tribal communities in North America, which have developed as a result of a persistent history of exploitation, we scientists should aspire to carry out meaningful outreach when we study Native American population history to ensure that any manuscripts we write are sensitive to indigenous perspectives. The details of how to achieve such consultation need to be worked out, and it seems to me that there will never be a solution that everyone will find comfortable. But we need to try to make progress beyond the situation we are facing right now, in which many researchers are reluctant to undertake any studies of Native American genetic variation for fear of criticism, and because of the extraordinary time commitment that would be required in order to accomplish all the consultations that some tribal representatives and scholars have recommended. This has had the effect of putting research into genetic variation among Native Americans into a deep chill—with far less research in this area going on than anyone but the people most hostile to scientific research would like.

  Disputes over Bones

  Ancient DNA studies of population history are mostly not as fraught as studies of present-day people. However, in 1990, the U.S. Congress passed the Native American Graves Protection and Repatriation Act (NAGPRA), which requires institutions that receive U.S. funding to contact Native American tribes and offer to return cultural artifacts, including bones that are from groups to which Native Americans can prove a biological or cultural connection. This has meant that Native American remains are being returned to Native American tribes and the opportunity to carry out ancient DNA analysis on many of the samples is disappearing. NAGPRA has had its greatest impact on archaeological remains dating to within the last thousand years, for which a relatively strong case can be made for cultural connections with living Native American tribes. The case for cultural connection is harder to make for very old remains, such as the approximately eighty-five-hundred-year-old Kennewick Man found on U.S. lands in Washington State in 1996.

  Kennewick Man’s skeleton was initially slated for return to five Native American tribes that claimed him as an ancestor, but was made available for scientific study instead after courts found that there was no good scientific evidence that he was Native American under the rules of NAGPRA. To win their case, the scientists who challenged the tribal claims pointed to analyses of skeletal morphology that suggested that his skeleton was closer to Pacific Rim Asian and Pacific islander populations than to present-day Native Americans.18 In 2015, though, Eske Willerslev and his colleagues extracted and studied ancient DNA from Kennewick Man, which showed that these conclusions from the morphological studies were wrong.19 Kennewick Man is in fact derived from the same broad ancestral population as most other Native Americans.

  Ancient DNA trumps morphological analysis whenever it is possible to compare the two types of data. The reason is simple. Morphological studies of skeletons can only examine a handful of traits that are variable among individuals, and thus can usually support only uncertain population assignment. In contrast, genetic analyses of tens of thousands of independent positions allow exact population assignment. Thus, the characterization of the ancestry of a single sample (like Kennewick Man) based on a small number of morphological traits cannot convincingly distinguish between Native American and Pacific Rim ancestry. Genetic data can.

  While the ancient DNA study produced clear proof of the Native American ancestry of Kennewick Man, it was not so clear whether he bears a particularly strong relationship to the Washington State Native American populations that made claims on his remains. The paper reporting the Kennewick Man genome sampled DNA from the Colville tribe, one of the five tribes staking a claim of relationship to him, and argued that the data were consistent with a direct link. However, the Colville was the only tribe from the lower forty-eight states of the United States that the scientists analyzed, and a close look at the details of the paper provides no compelling case that Kennewick Man is more closely related to the Colville tribe than he is to Native Americans as far away as South America.20 The Colville data are also not available to the scientific community for independent analysis—they were not provided to my group on request despite the fact that the journal in which they were published requires sharing of data as a condition for publication.

  Wishful interpretation of genetic data is not limited to Kennewick Man. In 2017, a study of an approximately 10,300-year-old skeleton excavated from an island off the Pacific coast of present-day Canada, claimed evidence for an unbroken presence of a lineage of Native Americans in the same region from his time until to the present day.21 But an examination of the analyses presented in the paper showed that this individual, too, was no more closely related to local people than to Native Americans in South America.

  These are just two examples of how the ancient DNA literature is beginning to fill up with unsubstantiated claims of direct ancestral links between ancient skeletons and groups living today, a problem that is not limited to the Americas. Scientists working with indigenous people have an incentive to make such claims, as claims like this are often welcomed by local groups, and open the door to sampling. The normal scientific process, in which scientists point out claims that are not compellingly supported by data, is also not working as it should. A concern is that when members of groups are directly engaged in scientific investigation of their own history, people’s wish that certain things should be true often colors presentation of the findings. And scientists not involved in the work are often too anxious about repercussions to point out problems.

  The Kennewick Man case was contentious and played itself out in court, engendering hostility between academics and Native American tribes. It has had consequences for scientists interested in Native American population history, and it has made such research far more difficult. From my experience interacting with archaeologists, anthropologists, and museum directors who focus on Native American prehistory, it is clear to me that many feel a deep sense of loss about the return of collections of scientifically important bones, and wish to keep them in the possession of museums while acknowledging the dubious ways in which many of these collections were assembled in the course of U.S. expropriation of Native American lands.22 Balanced against this is the sense of loss that many Native Americans feel about having ancestors’ remains disturbed. To navigate these competing interests and the law, many museums employ “NAGPRA officers” whose job it is to identify cultural and skeletal remains that can be associated with particular Native American tribes and to reach out to representatives of those tribes in order to return the items. But while the NAGPRA officers with whom I have interacted are dedicated to fulfilling the letter of the law and do so professionally, they are also careful to not go beyond it. They feel distressed when, as in the case of Kennewick Man, remains are returned to tribes without the evidence of biological or cultural connection that NAGPRA regulations require.

  One geneticist who is breaking new ground in this area is Eske Willerslev. Not only with the Kennewick sample but also with ot
her indigenous skeletal remains from which he has assembled DNA, Willerslev has won the cooperation of indigenous communities in a way that is innovative and brilliant—even if it is not making everyone in the archaeological and museum community happy. He has realized that there can be shared interests between indigenous communities and geneticists because DNA studies can empower tribes to stake claims on remains. This happened in the case of the genome sequences extracted from an approximately one-hundred-year-old Australian Aboriginal hair sample,23 the almost thirteen-thousand-year-old Clovis skeleton,24 and the approximately eighty-five-hundred-year-old Kennewick skeleton.25 In all three cases, Willerslev approached tribes directly after obtaining DNA, instead of engaging them through an institutionally run process such as the ones that have been set up through NAGPRA.

  Although many in the archaeological community have been concerned about Willerslev’s approach of engaging tribes outside the formal institutional process, he has been successful in several ways. In Australia, his engagement with Aboriginal groups in the context of his work on the hundred-year-old hair sample generated goodwill and opened the door to a much more ambitious study of present-day Aboriginal populations published in 2016 by him and colleagues.26 Similarly, in the United States, Willerslev’s engagement with indigenous groups in the Clovis and Kennewick cases has helped generate goodwill and encouraged tribes to support ancient DNA analysis of other remains.

  A remarkable example of this progress is provided by remains found in Spirit Cave in Utah. In 2000, the U.S. Bureau of Land Management decided against returning these almost eleven-thousand-year-old remains to the Fallon Paiute-Shoshone tribe that requested them. The bureau’s basis for the decision was that there was no evidence of biological or cultural connection to that tribe. The tribe then sued, putting the remains into a legal limbo that allowed them to be investigated only for the purpose of studying their ancestry to determine whether they indeed might have a biological connection to the Fallon Paiute-Shoshone. In October 2015, after publication of the Kennewick paper, Willerslev was given access to the remains for ancient DNA analysis, and around a year later he delivered to the bureau a technical report showing that the individual had ancestry that was entirely from the same deep lineage as present-day Native Americans. On the basis of this report, the bureau decided to return the bones to the tribe.27

  This decision confused the NAGPRA officer I corresponded with about it, who noted that the interpretation was beyond the letter of the NAGPRA law, which required documentation of a connection to the Fallon Paiute-Shoshone more than to other groups, which Willerslev had apparently not demonstrated. But when I talked with Willerslev about returning samples to tribes, his view was that the letter of the NAGPRA law was not so important and that the community standard was changing even if the law had not yet caught up. In an article in the scientific journal Nature about the decision to return the Spirit Cave remains, the anthropologist Dennis O’Rourke was quoted as saying that the case set an example for how Native American groups could be engaged in using genetics to determine which remains to study and rebury. The anthropologist Kim TallBear pointed out how the Spirit Cave example showed that the relationship between tribes and scientists need not be antagonistic: “Tribes do not like having a scientific world view politically shoved down their throat…but there is interest in the science.”28

  Willerslev’s realization that ancient DNA data provide a type of evidence that can be used to establish claims on unaffiliated remains held in museum collections offers an unexpected opportunity to begin to break the logjam of poor relations that has built up between scholars and indigenous communities.

  There is also a second great area of unrealized common cause between Native Americans and geneticists—the potential to use ancient DNA to measure the sizes of populations that existed prior to 1492 by looking at variation within the genome of ancient samples. This is a critical issue for Native Americans, as there is evidence for about a tenfold collapse in population size in the Americas following the arrival of Europeans and the waves of epidemic disease that Europeans brought, leading to the dissolution of previously established complex societies. The relatively small population sizes that European colonialists encountered when they arrived in the Americas were used to provide moral justification for the annexation of Native American lands. The European colonialists had an interest in minimizing the estimates of Native American population sizes, of claiming that there were few if any civilizations or sophisticated populations in the Americas before Europeans came.29

  I hope that as the consequences of the genome revolution are more broadly realized, indigenous people will increasingly recognize how DNA can become a tool to connect present-day Native American people to their roots and to each other. This will not solve all the concerns that Native American ethicists and community leaders have articulated, but it may serve to reduce antagonism and promote greater understanding and even collaboration in the future.

  The Genetic Evidence of the First Americans

  The first genome-scale study of Native American population history came in 2012, when my laboratory published data on fifty-two diverse populations. A major limitation of the study was that we had no samples at all from the lower forty-eight states of the United States because of anxieties about genetic research on Native Americans. Nevertheless, the study sampled Native American diversity in much of the rest of the hemisphere, and provided new insights about the past.30

  Most of the individuals we studied derived small fractions of their genomes from African or European ancestors in the last five hundred years, reflecting the profound upheavals that have occurred since the arrival of European colonists. We carried out many analyses on individuals with no evidence of such mixture, but for some populations, especially in Canada, all the individuals we sampled had at least some non–Native American ancestry. Because we wanted to include these populations, we used a technique that allowed us to identify which sections of people’s genomes were of European or African origin. We did this by searching for extended genomic stretches in which individuals carried genetic variants at high frequency in Africans and Europeans but at low frequency in Native Americans. Masking out these sections of the genome helped us to peel back the history of five hundred years of admixture in the Americas to understand something about what the structure of Native American population relationships was like before European contact.

  We compared all possible pairs of Native American populations using the Four Population Test. We used this test to evaluate whether Eurasian populations—for instance, Han Chinese—shared more genetic mutations with one Native American population or another, testing all possible pairs of populations. For forty-seven of the fifty-two populations, we could not detect differences in their relatedness to Asians. This suggested to us that the vast majority of Native Americans today, including all those from Mexico southward as well as populations from eastern Canada, descend from a single common lineage. (Five remaining populations, all from the Arctic or from the Pacific Northwest coast of Alaska and Canada, also had evidence of ancestry from different lineages.) Thus the extraordinary physical differences among Native American groups today are due to evolution since splitting from a common ancestral population, not to immigration from different sources in Eurasia. We called this common ancestral population the “First Americans.”

  We hypothesized that the “First American” lineage that we had characterized represented the descendants of the first people to spread south of the ice sheets, whether via an ice-free corridor or along a coastal route. Genomic studies so far have not been able to determine how small this group was or how many generations it wandered. But whatever happened, we were arguing that this was a pioneer population of limited size that moved into a human vacuum, expanding dramatically wherever it arrived.

  The genetic data provide support for the correctness of this hypothesis in its broad outlines. As we applied the Four Population Test time and again, it became clear to us that the great
majority of Native Americans, from populations in northern North America down to southern South America, can be broadly described as branches of one tree, forming a sharp contrast to patterns of population relationships in Eurasia. Most populations branched cleanly off the central trunk with little subsequent mixture. The splits proceeded roughly in a north-to-south direction, consistent with the idea that as populations traveled south, groups peeled off and settled, remaining in approximately the same place ever since. The most striking exception to this pattern was the less than thirteen-thousand-year-old infant associated with the Clovis culture who was found in Montana very close to the present-day Canadian border. The Clovis infant came from a lineage different from that of present-day inhabitants of neighboring Canada, reflecting major population movements that must have happened later.

  In some places in the Americas, ancient DNA confirms the theory that populations have remained in the same region for thousands of years. According to analyses we and Lars Fehren-Schmitz have done of Peruvians dating up to nine thousand years ago, there has been broad continuity in Native American populations in this region. All the ancient genomes from Peru that we have studied are more closely related to each other and to present-day Native Americans from Peru who speak the Quechua and Aymara languages than they are to any other present-day South American populations. We have similar findings from Native American individuals from southern Argentina dating to around eight thousand years ago, and Native American individuals from southern Brazil dating to around ten thousand years ago. The same applies to Native Americans from the islands off British Columbia, who appear to have been part of a continuous population for around six thousand years, even if the local continuity does not clearly go back more than ten thousand years.31 All are more closely related to Native Americans who live in the same regions today than to Native Americans far away.