by Martin Jones
Merriwether had not been alone in finding lineages that fell outside the classic four lineages of Wallace. One recurrent outsider became labelled ‘X’. It was characterized by mutations along the control region at positions 16,223 and 16,278. In one native American community, the Ojibwa, as many as one in four were of this rarer lineage. Like lineage B, it was absent from its presumed ancestral path from Siberia across to Alaska and Canada. What is more, this lineage seemed to be absent throughout Asia. The first geneticist to recognize this lineage, Antonio Torroni, had found the lineage, not in Asia, but in various parts of Europe. Could it be that some forgotten European travellers had carried the lineage across the Atlantic Ocean? Not according to Forster’s calculations using the molecular clock. These placed lineage X way back in prehistory, well before such a journey was feasible. Once again, the ancient DNA might cast light on this issue. Most ancient DNA studies of pre-Columbian Indians had targeted too small a sample to pick up rare lineages. There were however some important exceptions.
the pre-columbian ‘oneota’ of illinois
Around 200 years before Columbus arrived, a short-lived village community buried its dead within a mound on a bluff above the Illinois River. They were from a culture archaeologists describe as the ‘One-ota’, related to today’s Sioux. Most communities of the Oneota culture inhabited the Central Plains, further to the west. This particular village was an outlier, and perhaps suffered as a result. Many of the buried bodies had suffered a violent death, and their burial mound consequently conserves perhaps just a generation or two of their entire village. During the 1980s the burial mound was threatened by a highway development, and a rescue excavation was mounted by the Illinois State Museum. The 260 skeletons they uncovered were made available to Mark Stoneking and his student, Anne Stone. The work they did on the skeletons stands as a model study of the genetics of an ancient population.
Virtually all of Stone’s ancient Oneota population was found to belong to one of the four Wallace haplotypes. This is also true of the several hundred pre-Columbian burials that have now been typed, spanning the entire New World from Hudson Bay in the north to Tierra del Fuego in the south. What they reveal is a pattern and geography of pre-Columbian mitochondrial haplotypes that closely match those among living Native Americans.
Like Peter Forster, Anne Stone went beyond a study of well-known lineage markers to look in detail at variation in the mitochondrial control region. This is what allowed her to pick up extra lineages among her sample. She realized that two of her Oneota individuals came from an unusual lineage, which she first published as ‘lineage N’. On seeing her publication and reading the mutation site involved, Peter Forster realized what she had. These pre-Columbian individuals were of lineage X. The rare lineage possessed by 25 percent of the northern Ojibwa people, who live today just a few hundred kilometres to the north of where the Norris Farm mound was excavated, could also be found among that same Norris Farm population 700 years earlier.
Norris Farm may provide the most detailed glimpse of pre-Columbian lineage X, but it is not alone, nor is it the earliest example of the lineage. It has also been found among Brazilian clay pot burials which are 4,000 years old, and among the bodies recovered from the Windover Bog in Florida, which are 8,000 years old. These findings all support the genetic argument of Peter Forster and others, that the lineage is ancient. It entered America before a sea journey was plausible, and must have come across the northern land bridge. Lineages B and X illustrate how the great climatic fluctuations of the Quaternary Epoch not only determined the schedule of population movement, through exposing land bridges, but also wiped parts of the slate clean by eradicating lineages from the harsher environmental regions.
One cannot help comparing Greenberg’s three journeys in prehistory with Columbus’s three ships sailing several thousands of years later. In each account, there is a hazardous journey to be made between the Old World and the New, followed by rapid infilling of the continent by the journeyers’ descendants. Perhaps the concept of a perilous journey has been a distraction, one of several ‘great journeys’ that have rather over-simplified the human past. The data discussed above are not best explained by three small and discrete migrations, grabbing windows of opportunity to make a dry crossing of the Bering Straits. They fit better with a rather larger number of people coming across, from diverse ancestral homes. We have also had to rethink the timing and geography of any genetic bottleneck. From a modern-day, inter-glacial perspective, it is the crossing of the frozen north that constitutes the obvious bottleneck, the obvious challenge to expansion. However, from the standpoint of a late Pleistocene fisher-forager, whose family and dogs had lived for countless generations upon the windswept plains and grassy ranges of the vast land that now lies beneath the sea, there was no real sense of crossing from anywhere to anywhere. Instead that ancient population was fluctuating and responding to long-term climatic change. Some periods were ones of expansion, when lineages would be found further to the south. Other periods were ones of depletion, in which lineages would diminish and fade from some regions completely. The gradual population of what would later be transformed into a separate continent was imperceptible. In a sense, the molecular evidence has caused the concept of a great journey more or less to disappear.
ocean crossings
If the journey across the frozen north has been rewritten, what about the ancient seaward journeys that captured the imagination of Grafton Elliott Smith and many after him? The charters of arrows across the ancient map were absorbed by what to do about vast stretches of ocean. This was once an issue for anyone seeking connections between the Old World and the New, although we now know what a lowered sea level can do to the Bering Straits. It remains an issue for Australasia, which, even at the lowest sea levels, is still a seventy-kilometre sea journey from Asia. It became pivotal when envisaging the kind of intercontinental contact and world-wide island-hopping argued by followers of Smith’s global hypothesis.
It was with that in mind that Thor Heyerdahl set out in 1955 on one of several ocean voyages, this time headed for Easter Island in a remote part of the South Pacific Ocean. In the tradition of Smith, he was intrigued by the idea of a global diffusion of civilization with its roots in ancient Egypt. He wanted to know whether civilization could have been brought from Egypt by seafarers to Central America, whether it could have travelled from Central America southwards to Peru and, from there, across the oceans to the remote Pacific islands. He had established a few years earlier, in his famous Kon-Tiki voyage, that the second sea-leg was technically possible with a simple balsa-wood raft. Fifteen years later he would do the same thing for the first leg in the second Ra voyage. Now he was travelling to Easter Island in the relative comfort of a Norwegian trawler with a team of archaeologists hoping to recover evidence of the early voyagers themselves.
His team set to work on several parts of the island, surveying and excavating both the ancient remains of houses and the areas around the massive stone statues for which the island is famous. These were the monuments that had brought thoughts of prehistoric voyagers to mind, ‘Children of the Sun’ as Smith’s colleague, William J. Perry, had dubbed them. These stone statues and the stepped altars on which they stood each proved to be part of a monument of many phases. They began as simple altars in the first millennium AD, and ended several centuries later as well-worn ancient sites, with human burials incorporated into the rubble. These buried skeletons, and others that were excavated from separate graves nearby, provided a glimpse of the direct descendants of the first voyagers, before later Polynesians and Europeans had arrived. They were not the relatives of today’s Easter Islanders. Since Europeans discovered the islands in the eighteenth century, there has been a fair degree of population movement and replacement, and to the living islanders the great statues are as foreign and ancient as they are to us. By the time their families had reached the island, those who venerated the statues had died out. But the bones from Heyerdahl’s excav
ations belonged to people for whom the statues were part of their living world. These were the people for whom Grafton Elliott Smith’s diffusionism had suggested a South American ancestry, an ancestry that would save the notion of a single heartland of civilization. The bones, however, could neither speak nor offer any clue as to their ancestry–that is until their molecular tachometers were recovered and decoded.
Thirty-four years after the excavation, Erika Hagelberg was on her way to Chile’s Museum of Natural History where those human remains had been stored. Only a few months earlier Nature had published her groundbreaking paper, which presented evidence for the survival of DNA in ordinary archaeological bone. She was keen to take on an archaeological problem, and here was one. Did the ancestors of these ancient people travel in simple wooden rafts across the Pacific from America as Heyerdahl had surmised? If they had, then any remaining fragments of their mitochondrial DNA should trace their path. A decade had passed since the human mitochondrial genome had been mapped and, by now, a fair amount was known about regional variations in the setting of that sensitive evolutionary stopwatch, the mitochondrial control region.
Hagelberg succeeded in amplifying a stretch of the mitochondrial control region from some of these skeletons, and three particular base positions caught her eye. Among living people on the Asian side of the Pacific, one particular haplotype was recognizable by deviations from the Anderson sequence at positions 16,217, 16,247 and 16,261. This was not a haplotype found in the New World, and so provided a useful marker of ancestry. Sequences amplified from the ancient Easter Island skeletons displayed replacements at all three positions, sharing this pattern with populations found on only one mainland flank of the Pacific. Their ancestors had sailed from the direction of Asia, not America.
from balsa-wood boats to ‘fast trains’
Heyerdahl is a romantic at the tail end of the Smith and Perry tradition. He also knows a great deal about seafaring and the Pacific Ocean currents. With this knowledge, he was undoubtedly able to raise the debate about ocean crossings above what had been a poorly informed perusal of vast stretches of blue on a flat map. The path of the ‘Children of the Sun’ was, however, already fading from view. Around the time that Heyerdahl was making his sea voyages from America, others were piecing together the archaeology of the islands further to the west and, like Heyerdahl, making use of the newly developed method of radiocarbon dating. From New Zealand, Jack Golson was beginning to make sense of a particular style of stamped pottery recognized on a New Caledonian beach called ‘Lapita’, and found on other islands as far afield as Fiji and Tonga. Charcoal on the Lapita beach had provided dates for the pottery, placing it in the early first millennium BC. The geographical spread of the Lapita pottery fitted neatly at the heart of a larger region in which all the modern islanders speak related languages and have similar words for a range of plants and animals.
Here were the makings of a Lapita culture and an alternative set of ancient journeys into the Pacific, originating from the Old World in the west, rather than the New World in the east. As excavations of the settlements with Lapita-style stamped pottery got under way, fragments of their fishing, farming and gardening lives came to light. The islanders kept pigs and fowl, and repeatedly used their considerable seafaring skills for fishing and exchange with other distant islands, as well as to settle new islands. The growing number of carbon dates for these sites fell within a 1,000-year time band. On the islands just north-west of New Guinea, dates as early as 1500 BC were recorded. The carbon dates came out as progressively more recent as one moved to the south and east, indicating that by 500 BC the settlements had spread over 5,000 km into the heart of the Pacific Ocean. The pace of this expansion has attracted the nickname the ‘fast train to Polynesia’.
The significance of this prehistoric fast train from the west was enhanced when it was placed in the context of contemporary evidence of language and human genetics. The languages spoken in the Pacific islands cluster within the ‘Austronesian’ language family, a group whose greatest diversity is found in Taiwan. If a tree is built from variations in these languages and used in much the same way as a genetic tree, then it traces a clear arc. From its root in Taiwan, this arc branches southwards and eastwards towards the increasingly remote islands of the South Pacific. The genetic evidence is more complex, with a clear influence from island Melanesia as well as Taiwan and mainland Asia. Nevertheless, from blood proteins and DNA sequences alike, a restricted range of twigs on the Asian branch of the mitochondrial family tree can be traced as we move towards the remote islands of the Pacific. Modern languages and modern genetics described a sweeping arc across the Pacific, with, at its centre, a smaller arc derived from ancient evidence, the dated sites characterized by stamped Lapita pots and domestic animal bones. Peter Bellwood brought these together as evidence of an ancient suite of voyages by pioneer farming settlers. He suggested that they set out from Taiwan as much as 4,000 years ago and reached the most remote islands of all around 1,500 years ago. On the way, they laid the foundation of the Lapita culture in the central part of the arc.
Hagelberg’s ancient Easter Islander DNA gave a much better fit with this set of journeys. In a later comprehensive survey of the mitochondrial control region among modern islanders, Bryan Sykes found all three of her key base substitutions right across the arc, and two of the three substitutions could be traced back to Taiwan. Here was a pattern she could explore by extending her study of ancient bones on Easter Island to some even older bones elsewhere in Polynesia. With a more comprehensive range of ancient islanders and their DNA, she could really get to grips with this ‘fast train’ from Asia, or so she assumed.
Her work on the other Polynesian islands involved a characteristic mitochondrial feature that was playing a large part in piecing together the earlier journeys of colonization of the New World explored above. This was the deletion of nine base-pairs on the far side of the mitochondrial genome from the control region. It was the same deletion that Douglas Wallace had used to identify lineage B among the Native Americans. This deletion of nine base-pairs was also encountered in the islanders of the South Pacific, where it was extremely common. On the continental mainlands on either side this deletion was held by a significant minority of individuals, but here it was carried by a sizeable proportion of Pacific islanders. Of the modern Polynesians that Sykes surveyed, 94% carried the deletion. Hagelberg realized that she could use this deletion, together with the control region variations she hadfound at Easter Island, to explore the alternative journey and trace the path of pioneer farmers along the arc.
Ancient skeletal remains were collected from around a dozen Pacific islands, and DNA successfully amplified from twenty-one of them. Eleven of these were just a few centuries old, going back 700 years at most. Nine of those eleven bodies carried the nine base-pair deletion and, where the amplification was successful, the control region generally displayed the characteristic replacement of three bases. These ancient islanders were part of the same lineage as both modern islanders and the ancient Easter Islanders. However, none of these corresponded to the actual users of Lapita pottery. To explore the earliest part of Peter Bellwood’s story much older skeletons were needed. Lapita burials are not common, but Roger Green from New Zealand had found and excavated a small cemetery with eight adults on Watom Island. Hagelberg gathered a further ten skeletons, between 1,700 and 2,700 years old, from the cultural context in which Lapita pottery was used. They may have used the pottery, but their mitochondrial sequences were not what would be expected of the ‘fast train to Polynesia’.
None of the ten earlier skeletons carried the nine base-pair deletion. What is more, the three from which the control region was amplifiable displayed virtually no replacements on the characteristic sites. The people who actually used the Lapita pottery appeared on the face of it to be different from modern Polynesians. In spite of the geographical match between the ancient Lapita evidence and the modern variation, and some elegant archaeological
arguments in favour of continuity between the Lapita culture and more recent material culture traditions, the ancient DNA evidence suggested a break in genetic continuity.
Everyone, including Hagelberg, was cautious about making too much of a handful of skeletons. Six percent of modern islanders, like the Lapita burials, lack the nine base-pair deletion. Perhaps a rather larger sample of ancient sequences will modify the picture and bring the ancient islanders back into the genetic fold. However, even small samples are informative. Imagine tossing a coin a small number of times. If it falls head upwards two or three times in a row, that can be put down to chance. If it falls head up ten times in a row, then it is not unreasonable to suspect that the coin might be weighted. By analogy, the most straightforward explanation for the absence of the very common deletion from all ten skeletons is that the material culture traditions over-simplify the demographic history of the islands. There was a series of journeys, not all from the same starting point, that have accounted for the islands’ various populations through time. There may be some continuity between the material traditions identified on the Lapita beach and those of later times, but different human lineages with different ancestries became the agents of that continuity.
The Pacific islands provide an interesting example of how the retracing of ancient journeys has proceeded. Starting from evidence from the present day, be it of genetic traits, language or surviving monuments, projections back in time towards their origins lead us towards some fairly simple stories, stories that can easily be portrayed by sweeping arrows across the map. That is to some extent inevitable. Given a choice of conjectural paths that might conceivably lead to the present, it is good scientific practice to choose the simplest explanation. As scientific dates first, and then our molecular tachometer have come on the scene, we might hope they would select which simple explanation is the correct one. What they have instead tended to show is that real life is more complicated than that, and simplicity is not a particularly relevant criterion in tracking real human pasts. Hagelberg’s work on ancient DNA provided a neat fit for neither the eastward nor the westward journey in their simplest forms. There were other things going on, and perhaps rather more journeys involved, something borne out by Bryan Sykes’s extensive survey of the modern DNA.
