All  
Surface model of the Xiahe mandible after digital removal of the adhering carbonate crust. Source: Jean-Jacques Hublin, ResearchGate.

The Teeth Have It: Tracing the Denisovan Legacy Around the World

Print

A dental trait found in the Xiahe Denisovan jawbone as well as in some ancient and modern human populations now provides a rare opportunity to track the geographical reach and perhaps even the final fate of the Denisovans.

Confirmation of the discovery of the first Denisovan mandible (lower jawbone), found as far back as 1980 in a cave in Xiahe County, on the eastern edge of the Tibetan Plateau in China’s Ganzu Province, might now provide the key to understanding the geographical reach and perhaps even the final frontiers of this enigmatic pre-sapiens population. This follows the publication of a new study focusing on the mandible’s surviving teeth.

The study, headed by Shara E. Bailey, Associate Professor of Anthropology with the Center for the Study of Human Origins at New York University, notes that the Xiahe mandible has a mandibular molar (see fig. 1)—a molar in the lower jawbone—that displays three roots instead of the more usual two found in anatomical modern humans ( Homo sapiens ).

Reconstruction of the Xiahe mandible, now confirmed to be Denisovan in origin. (Jean-Jacques Hublin / Author Supplied).

Reconstruction of the Xiahe mandible, now confirmed to be Denisovan in origin. ( Jean-Jacques Hublin / Author Supplied).

This is the first time that this dental trait has been detected in a pre-sapiens population. What’s important about this finding is that this same dental abnormality has been recorded in a number of ancient and modern human populations, almost all of them of Asian or Native American extraction.

Since these same populations are now known to carry Denisovan ancestry Bailey and her team propose that this unique dental trait, referred to as 3RM1 (three-rooted lower first molars), is caused not by a gene mutation, as was generally assumed, but derives instead from introgression, that is interbreeding, either with Denisovans or with their hybrid descendants.

Some human populations, particularly those considered ‘ Mongoloid’ in ethnicity, such as Chinese, Aleuts, and Inuits, can have mandibular molars with an extra, supernumerary root in anything between 5 and 40 percent of the population.

Most usually this dental abnormality is seen in the first mandibular molar, where it is referred to as 3RM1. However, it can also affect the second or third molars, where it is known as 3RM2 and 3RM3 respectively.

Modern human mandible (lower jaw) showing the three molars and the first molar that can have an extra root, the term 3RM1 being given to this trait. (Author Supplied / Public Domain)

Modern human mandible (lower jaw) showing the three molars and the first molar that can have an extra root, the term 3RM1 being given to this trait. (Author Supplied / Public Domain)

Penghu 1

Indeed, it is in the second mandibular molar of the Xiahe jawbone that a third, supernumerary root was found, a feature found also in the large pre-sapiens mandible dredged out of the Penghu Channel between Taiwan and the nearby Penghu Islands. This was found by fishermen sometime before 2008 and is thought to be between 190,000 and 10,000 years old.

Known today as Penghu 1 , the jawbone (see fig. 3) has characteristics common to both the Xiahe mandible and the oversized Denisovan molars found in the Denisova Cave , located in the Altai Mountains of southern Siberia. This suggests that it too could belong to a Denisovan.

The Penghu 1 mandible found sometime prior to 2008 by fisherman dredging the channel between Taiwan and the Penghu Islands. (Animal Party / CC BY-SA 4.0)

The Penghu 1 mandible found sometime prior to 2008 by fisherman dredging the channel between Taiwan and the Penghu Islands. (Animal Party / CC BY-SA 4.0 )

If the Penghu 1 jawbone is Denisovan, then this not only adds weight to Bailey’s proposal that 3RM1 was inherited by modern humans via gene flow, but it also now permits us to better zero in on the geographical reach of the Denisovans through the continuance, arguably in the same regions in many instances, of their hybrid descendants .

The Philippines

To date, the oldest anatomically modern human remains to display 3RM1 is a mandible found in 1962, along with part of a skullcap, during excavations at the famous Tabon Cave at Quezon on the island of Palawan in the Philippines (see fig. 4).

The Tabon Cave in the Philippines showing the extraordinary simulacra of a human skull noted in 2014 by tourist Philip Maise. (Philip Maise / CC BY-SA 3.0)

The Tabon Cave in the Philippines showing the extraordinary simulacra of a human skull noted in 2014 by tourist Philip Maise. (Philip Maise / CC BY-SA 3.0 )

Dubbed Tabon Man, it is approximately 31,000 years old (+/-7000-8000 years). Not unconnected is the fact that a dental survey of human remains belonging to 171 individuals found in the Philippines and dating back almost 5,000 years found that no less than 17.4 percent displayed 3RM1.

This suggests that this trait might well have been present in the archipelago from the age of Tabon Man 31,000 years ago all the way down to 5,000 years ago, and arguably even beyond that to more recent times.

There is also convincing circumstantial evidence that 3RM1 might have been introduced to the archipelago either by Denisovans or by their hybrid descendants. With the sequencing of the Altai Denisovan genome in 2010, with higher coverage sequencing in 2012, it became apparent that some Austronesian indigenous peoples living on the island of Mindanao in the Philippines (see fig. 5), more specifically the Mamanwa and Manobo populations, possess noticeable traces of Denisovan ancestry.

Old photo of a Bagobo (Manobo) priest from the Philippines. The Manobo have been shown to possess noticeable traces of Denisovan ancestry. (Julian Felsenburgh / Public Domain)

Old photo of a Bagobo (Manobo) priest from the Philippines. The Manobo have been shown to possess noticeable traces of Denisovan ancestry. (Julian Felsenburgh / Public Domain )

The additional fact that a recent genetic survey of human populations from Indonesia and New Guinea, located south and southeast of the Philippines, uncovered compelling evidence that at least one branch of so-called Sunda Denisovans (see below) might have survived through until around 15,000 years ago tells us that the earliest peoples of the Philippines almost certainly encountered Denisovans. Where exactly is unclear, although most likely it was in an island in southeast Asia somewhere.

Thus it is becoming clear that islands of southeast Asia, which once formed part of a much larger landmass named Sunda that embraced Indonesia, Borneo, and the Malaysian peninsular, along with the Philippines and Taiwan—which was itself much larger in size at the end of the last ice age, the Penghu channel being then above water—can now be seen as potentially important territories of the Denisovans before their final disappearance perhaps as late as 15,000 years ago.

This we can deduce from the presence in the region of populations with both Denisovan DNA and a high frequency of 3RM1 (see fig. 6 showing world distribution of Denisovan ancestry among current populations and maps and fig. 7 showing the percentages of 3RM1 found in ancient and modern populations worldwide).

Map showing the distribution of Denisovan DNA in modern populations worldwide, along with sites associated with the discovery of archaic human remains associated with the Denisovans. (Andrew Collins / Author Supplied)

Map showing the distribution of Denisovan DNA in modern populations worldwide, along with sites associated with the discovery of archaic human remains associated with the Denisovans. (Andrew Collins / Author Supplied)

Map showing the percentage level of the dental trait 3RM1 in ancient and modern populations as recorded by a number of different studies as collated by Ahmet Cem Erkmana and Ferhat Kayab in 2014 (see Mediterranean Archaeology and Archaeometry 14, No. 1, 1-11). Note the high incidence of 3RM1 in eastern and southeastern Eurasia and also in North America. (Andrew Collins / Author Supplied)

Map showing the percentage level of the dental trait 3RM1 in ancient and modern populations as recorded by a number of different studies as collated by Ahmet Cem Erkmana and Ferhat Kayab in 2014 (see Mediterranean Archaeology and Archaeometry 14, No. 1, 1-11). Note the high incidence of 3RM1 in eastern and southeastern Eurasia and also in North America. (Andrew Collins / Author Supplied)

What is more, there are several other clear examples where modern populations have been found to possess both 3RM1 and Denisovan ancestry. They include the Chinese, Malay, Javanese, Japanese, Thai, and Vietnamese. All show not only a high incidence of 3RM1, but also traces of Denisovan ancestry.

The Inuit

Also striking is the case of the Inuit of Arctic Canada, Alaska, and Greenland. Genetic studies have shown that the Greenland Inuit not only possess Denisovan ancestry, but also carry two genes (TBX15 and WARS2) inherited from the Denisovans and enabling them to live permanently in extremely cold environments.

Thus it is interesting to note that both ancient and modern Greenland Inuits display 3RM1 in as much as 26.9 percent of the population , a figure very close to that of their Alaskan counterparts.

Africans and Europeans

In contrast, European and African human groups display little if any Denisovan ancestry. This is probably due to the fact that the Denisovans and their hybrid descendants would seem only to have ever occupied eastern Eurasia, along with Melanesia, Australia, and just maybe the American continent.

It therefore comes as no surprise to find that dental studies show a fairly low frequency of 3RM1 among European Caucasians. For instance, two separate studies of British Caucasians, one from 1899 and the other from 1973, both found that around 3.4 percent of the population display 3RM1. Another study of European Caucasians in general found 3RM1 in 3.2 percent of those examined, very similar to the results from Britain.

In Africa a study of the mandibular molars of 67 members of the Bantu population determined that not one individual possessed 3RM1. Another study of Nubian dental traits found 3RM1 in 2.6 percent of the population, with some groups displaying none whatsoever.

If found to be consistent among other populations of equatorial Africa, such findings would add weight to the idea that the Denisovans developed as an individual pre-sapiens species only after they reached the Eurasian continent and moved eastwards into central, eastern, southern, and southwestern Asia, leaving western Eurasia to their sister group, the Neanderthals.

The particularly low rate of 3RM1 among European Caucasians is usually seen as the result of a simple gene abnormality. This is in complete contrast with Asian populations, where the percentage of 3RM1 is so high that it can only have come from direct inheritance.

Originally it was thought that 3RM1 in Asian populations was the result of a genetic mutation in a past common ancestor. Yet with the discovery that a third root is present in lower molars belonging to both the Xiahe and Penghu 1 mandibles it is now becoming clear that a more suitable explanation would be to assume this dental trait in fact derived from introgression either with the Denisovans themselves or with their hybrid descendants.

So, the fact that between 3 to 3.4 percent of European Caucasians do carry 3RM1, as opposed to the zero rate detected in some equatorial African populations, could well imply that its presence may in fact be a very weak signal of past Denisovan introgression with the ancestors of some European populations.

Greg L. Little and the current author in new book Denisovan Origins provide new evidence that the Upper Paleolithic culture known as the Solutrean, which flourished in southwest Europe circa 20,000-15,000 BC, evolved from much earlier incursions into central Europe circa 30,000-25,000 years ago that might well have begun as far east as southern Siberia or Mongolia.

Solutrean tools, 22,000 – 17,000 BC. (World Imaging / CC BY-SA 3.0)

Solutrean tools, 22,000 – 17,000 BC. (World Imaging / CC BY-SA 3.0 )

If correct, then it is possible that this Proto-Solutrean or eastern Gravettian population possessed at least some North Asian ancestry, which perhaps included both Denisovan DNA and the 3RM1 dental trait. If so, then these could have persisted among the Solutreans of southwest Europe, before eventually being diluted through introgression with other Upper Paleolithic populations.

Moreover, the arrival in western Europe of Neolithic farmers inbound from the Near East circa 6000-5000 BC would have diluted this Denisovan ancestry and the 3RM1 dental trait still further, explaining why both are barely traceable in Europeans today.

Siberians & Mongolians

Such speculations also make sense of findings regarding the presence of 3RM1 in Siberian and Mongolian populations. For example, a dental study of 164 individuals who inhabited northeastern Siberia circa 200 BC found that 23.2 percent possessed 3RM1. Furthermore, an examination of 30 individuals who lived on the shores of Lake Baikal in southern-central Siberia during Neolithic times showed that 23.3 percent, nearly one quarter, possessed 3RM1.

Lake Baikal is 994 miles (1600 kilometers) east of the Denisova Cave, the type-site of the Denisovans, which they occupied through until around 45,000 years ago. The discovery during excavations at the cave of finely worked beads made of ostrich eggshell has led to speculation that the Denisovans had contact with either archaic human or modern human groups existing in the Lake Baikal region, since it was the Transbaikal, east and southeast of this great inland sea, that was the closest natural habitat of ostriches at this time.

The presence of 3RM1 in human groups living on Lake Baikal during Neolithic times was very possibly the result of introgression between their direct ancestors and either the Denisovans or their hybrid descendants. This too is an important realization, especially as hardly any Denisovan DNA has been found in the current populations of Siberia and Mongolia; any substantial Denisovan ancestry having been diluted and lost long ago.

The New World

Moving across to the New World we find that among the Aleut of the Aleutian Islands, off Alaska, the frequency of 3RM1 is a staggering 40.9 percent! In contrast, however, very little Denisovan ancestry has so far been detected among their population.

What has been found, however, in two recent genetic studies, is the presence of Australo-Melanesian ancestry among the Aleut. Most likely this originated in the Sunda region of southeastern Asia, where 3RM1 has been noted in 15% of the population.

The presence of 3RM1 among the Na Dene/Northwest Coast First Peoples of the Pacific Northwest has also been found to be around 15 percent, while in many other Native American populations the rate falls to between 5 and 7 percent. Some Denisovan ancestry has been traced among the First Peoples of North America , with the Tlingit of the Pacific Northwest and the Ojibwa and Cree of the Great Lakes/St. Lawrence River region, displaying the highest incidence of Denisovan DNA so far detected in any one group.

However, not enough information is known yet about Denisovan ancestry in Native Americans, so how this might relate to the presence of 3RM1 among the same populations remains unclear. There are, however, clues.

Sinodonty and Sundadonty

Repeatable patterns of dental traits and dental morphology have been noted among different Asian populations , and these can be broken down into two distinct groups, both of which display high frequencies of 3RM1. The first of these groups are described as sinodonty in nature (from sino, meaning ‘Chinese’, and dont, meaning ‘tooth-like’). Those who display a sinodonty dental morphology are mainly eastern Asian peoples such as Chinese and Japanese.

The second group is classified as sundadonty in nature, from Sunda, the name of the former great landmass embracing Indonesia, Borneo, and the Malaysian peninsular. Those displaying a sundadonty dental morphology have been found in skeletal remains belonging to the Jōmon people of Japan, along with modern-day populations of Taiwanese aborigines, Filipinos, Indonesians, Borneans, and Malaysians. This includes populations of Australo-Melanesian ethnicity, even though sundadonty is not seen in Australo-Melanesian skeletal remains older than 5,000 years.

This has led to speculation that an original Australo-Melanesian population was overrun at this time, most likely by Austronesians coming from Taiwan in the north as part of the so-called Austronesian Expansion, circa 6500-5000 years ago. This said sundadonty does, as its name suggests, probably originate in the Sunda region of southeastern Asia.

The pioneer in the study of sundadonty and sinodonty was American anthropologist Christy G. Turner II (1933-2013). In a comprehensive study published in 1990 he determined that sundadonty appeared to be more archaic, and thus far older, then sinodonty.

This is a very important find since it closely correlates with what we know about the two basic types of Denisovan—the Siberian Denisovans, who probably inhabited the more northerly territories of Siberia, Mongolia, northern China, and Tibet, and the Sunda Denisovans, who thrived farther south in mainland southeastern Asia, Island Southeast Asia, Melanesia and perhaps even the former Sahul landmass that once embraced Australia and New Guinea.

Both branches of Denisovan display their own unique genetic profiles, although it is the Sunda Denisovans who are now recognized as distinctly more archaic than their more northerly neighbors, the Siberian Denisovans. It is a situation that matches very well the evolution and geographical extent of sundadonty and sinodonty dental morphologies in Asian populations.

This seems unlikely to be coincidence, suggesting that the dental traits found in these two different types of population, one in eastern Asia, the other farther south among peoples primarily of Australo-Melanesian or Austronesian descent, may have been forged by introgression either with Siberian Denisovans or Sunda Denisovans, the latter having perhaps survived in one form or another down to around 15,000 years ago.

It was perhaps with the arrival of their hybrid descendants, carrying both sundadonty and sinodonty dental morphologies, that Denisovan ancestry and 3RM1 was introduced into both North America and South America. Whether these migrations included actual Denisovans is a vexing question unlikely to be answered any time soon.

A better understanding of both 3RM1 and Denisovan ancestry among the indigenous peoples of the Americas now seems crucial to better piece together this incredible story of human migration at the time of the last ice age.

Top image: Surface model of the Xiahe mandible after digital removal of the adhering carbonate crust. Source: Jean-Jacques Hublin , ResearchGate.

Denisovan Origins by Andrew Collins and Greg L. Little is available now from Amazon and Barnes & Noble .

By Andrew Collins

Next article