Japanese Ethnogenesis Arose From Three Components Not Two

Background

The conventional story of Japanese ethnogenesis is that the Jomon hunter-gatherer-fishing culture arrived in Japan in the Upper Paleolithic era, a few thousand years after the Last Glacial Maximum, until around 1000 BCE when the Yayoi wet rice farming, cavalry warrior people arrived from Korea and conquered the Jomon, receiving substantial genetic admixture from them but little linguistic or cultural influence. The admixture event was notable particularly for having more male admixture into Yayoi society than female admixture, which is the opposite of the usually pattern when one people conquer another.

The Main Findings Of A New Paper

While this story pieced together from archaeology, linguistics and ancient and modern population genetic analysis isn't wrong, it misses two key points which a new paper studying a small but representative set of ancient DNA samples with state of the art methods and large comparison data sets reveal.

The big point is that after the Yayoi conquered the Jomon and admixed with them ca. 1000 BCE, there was a second wave of migration to Japan during the Kofun period of Japanese history, ca. 300 CE to 538 CE, by a population similar to the modern Han Chinese people that has the source of more than 60% of the resulting populations autosomal DNA. Since that second migration event, there has been only some modest introgression of additional Han Chinese-like admixture into the Japanese gene pool.

As Wikipedia explains at the link above:

The Kofun period (古墳時代, Kofun jidai) is an era in the history of Japan from about 300 to 538 AD (the date of the introduction of Buddhism), following the Yayoi period. The Kofun and the subsequent Asuka periods are sometimes collectively called the Yamato period. This period is the earliest era of recorded history in Japan, but studies depend heavily on archaeology since the chronology of historical sources tends to be distorted.

It was a period of cultural import. Continuing from the Yayoi period, the Kofun period is characterized by a strong influence from the Korean Peninsula; archaeologists consider it a shared culture across the southern Korean Peninsula, Kyūshū and Honshū.  

The word kofun is Japanese for the type of burial mound dating from this era, and archaeology indicates that the mound tombs and material culture of the elite were similar throughout the region. 

From China, Buddhism and the Chinese writing system were introduced near the end of the period. The Kofun period recorded Japan's earliest political centralization, when the Yamato clan rose to power in southwestern Japan, established the Imperial House, and helped control trade routes across the region.

Incidentally, this influx of East Asians into Japan started at fall of Han dynasty (three kingdoms) during a chaotic time of northern barbarian uprisings/invasions when dynasties were being formed in the north and the south (in addition to the Yamato dynasty of Japan).

The second point is relatively minor. The Yayoi people had a Northeast Asian genetic affinity akin, for example, to Manchurians, rather than to Han Chinese people. This insight emerges from the big discovery that the non-Jomon component of Japanese population genetics can be broken down into two distinct waves of migrations, which, when separated, are clearly derived from genetically distinct populations.

The timing of the waves of migration is supported by genetic admixture data estimates. As the body text of the new paper explains:

[W]e find support for a two-pulse model from our dating of the admixture in the Kofun individuals by DATES.

A single admixture event with the intermediate population (i.e., YR_LBIA) is estimated to have occurred 1840 ± 213 years before the present (B.P.), which is much later than the onset of the Yayoi period (~3 ka ago).

In contrast, if two separate admixture events with two distinct sources are assumed, the resulting estimates reasonably fit the timings consistent with the beginning of the Yayoi and Kofun periods (3448 ± 825 years B.P. for the admixture between Jomon and Northeast Asian ancestry and 1748 ± 175 years B.P. for Jomon and East Asian ancestry). These genetic findings are further supported by both the archaeological evidence and the historical records, which document the arrival of new people from the continent during the period.

Historical Context

The background material in the introduction, is, as is often the case, very helpful:

The Japanese archipelago has been occupied by humans for at least 38,000 years. However, its most radical cultural transformations have only occurred within the past 3000 years, during which time its inhabitants quickly transitioned from foraging to widespread rice farming to a technologically advanced imperial state. These rapid changes, coupled with geographical isolation from continental Eurasia, make Japan a unique microcosm in which to study the migratory patterns that accompanied agricultural spread and economic intensification in Asia. Before the arrival of farming cultures, the archipelago was occupied by diverse hunter-gatherer-fisher groups belonging to the Jomon culture, characterized by their use of pottery. The Jomon period began during the Oldest Dryas that followed the Last Glacial Maximum (LGM), with the earliest pottery shards dating to ~16,500 years ago (ka ago), making these populations some of the oldest users of ceramics in the world. Jomon subsistence strategies varied and population densities fluctuated through space and time, with trends toward sedentism. This culture continued until the beginning of the Yayoi period (~3 ka ago), when the arrival of paddy field rice cultivation led to an agricultural revolution in the archipelago. This was followed by the Kofun period, starting ~1.7 ka ago, which saw the emergence of political centralization and the imperial reign that came to define the region. 

An enduring hypothesis on the origin of modern Japanese populations proposes a dual-structure model, in which Japanese populations are the admixed descendants of the indigenous Jomon and later arrivals from the East Eurasian continent during the Yayoi period. This hypothesis was originally proposed on the basis of morphological data but has been widely tested and evaluated across disciplines. Genetic studies have identified population stratifications within present-day Japanese populations, supporting at least two waves of migrations to the Japanese archipelago. Previous ancient DNA studies have also illustrated the genetic affinity of Jomon and Yayoi individuals to Japanese populations today. Still, the demographic origins and impact of the agricultural transition and later state formation phase are largely unknown. From a historical linguistic standpoint, the arrival of proto-Japonic language is theorized to map to the development of Yayoi culture and the spread of wet rice cultivation. However, archaeological contexts and their continental affiliations are distinct between the Yayoi and Kofun periods; whether the spread of knowledge and technology was accompanied by major genetic exchange remains elusive.

Significance 

The mechanism of this paper does a lot to explain why the Jomon weren't simply crushed demographically the way hunter-gatherers in other parts of the world were upon encountering farming societies. 

More generally, knowing that there were sequential Manchurian and Chinese contributions to the formation of the Japanese People, and their order and distance from each other in time, also helps in reconstructing puzzles about how Japanese culture came to be the way that it is now. 

The paper demonstrates that the current population genetic mix in Japan is very recent, on the order of 1600 years old. In contrast, the population genetic mix of Europe was largely fixed by the end of the Bronze Age about 3200 years ago.

This paper sheds far more specific light on the context in which the Japanese language arose which strengthens the Altaic hypothesis and more generally favors some theories about historical Japanese linguistics while disfavoring others.

The paper calls attention to a larger Tungusic linguistic and cultural area that was a social and political and military reality and peaked at a time on the boundary between history and prehistory, that is now almost forgotten.

The paper's findings regarding the relative homogeneity of Jomon genetics over thousands of years and the entire expanse of Japanese territory supports a hypothesis that the Jomon may have shared a single language family, presumably one close to Ainu, while disfavoring deep schisms adding structure that divides this population.

The Paper

The paper and its abstract are:

Prehistoric Japan underwent rapid transformations in the past 3000 years, first from foraging to wet rice farming and then to state formation. A long-standing hypothesis posits that mainland Japanese populations derive dual ancestry from indigenous Jomon hunter-gatherer-fishers and succeeding Yayoi farmers. However, the genomic impact of agricultural migration and subsequent sociocultural changes remains unclear. We report 12 ancient Japanese genomes from pre- and postfarming periods. Our analysis finds that the Jomon maintained a small effective population size of ~1000 over several millennia, with a deep divergence from continental populations dated to 20,000 to 15,000 years ago, a period that saw the insularization of Japan through rising sea levels. Rice cultivation was introduced by people with Northeast Asian ancestry. Unexpectedly, we identify a later influx of East Asian ancestry during the imperial Kofun period. These three ancestral components continue to characterize present-day populations, supporting a tripartite model of Japanese genomic origins.

Niall P. Cooke, et al., "Ancient genomics reveals tripartite origins of Japanese populations" 7(38) Science Advances (September 17, 2021) DOI: 10.1126/sciadv.abh2419

Genetic Details

The genetic data was summarized as follows in the body text with uniparental genetic haplogroups considered:

Mitochondrial haplogroups for all Jomon individuals belong to the N9b or M7a clades, which are strongly associated with this population and rare outside of Japan today. The three Jomon males belong to the Y chromosome haplogroup D1b1, which is present in modern Japanese populations but almost absent in other East Asians. 

In contrast, the Kofun individuals all belong to mitochondrial haplogroups that are common in present-day East Asians, while the single Kofun male has the O3a2c Y chromosome haplogroup, which is also found throughout East Asia, particularly in mainland China.

Further Analysis Of Each Of The Three Components

The discussion section of the paper spells out a narrative and clarifies the geographic origin and timing of each population genetic wave of migration to Japan.

The Jomon

The lineage ancestral to Jomon is proposed to have originated in Southeast Asia with a deep divergence from other ancient and present-day East Asians. The timing of this divergence was previously estimated to be between 18 and 38 ka ago; our modeling with the ROH profile of the 8.8-ka-old Jomon individual narrows this date to a lower limit within the range of 20 to 15 ka ago. The Japanese archipelago had become accessible through the Korean Peninsula at the beginning of the LGM (28 ka ago), enabling population movements between the continent and archipelago. The subsequent widening of the Korea Strait 17 to 16 ka ago due to rising sea levels may have led to the isolation of the Jomon lineage from the rest of the continent and also coincides with the oldest evidence of Jomon pottery production. Our ROH modeling also shows that the Jomon maintained a small effective population size of ~1000 during the Initial Jomon period, and we observe very little changes to their genomic profile in subsequent periods or across the different islands of the archipelago.

A TreeMix analysis places the Jomon as an offshoot of the Hoabinhian people (a Mesolithic wave of people in Southeast Asia and Southern China ca. 12,000 to 10,000 BCE), with the Kusunda people (who are hunter-gathers in Western Nepal who historically spoke a language that is an isolate and were animistic religiously) as an intermediate population. 

Y-DNA haplogroup D has a cryptic distribution found in isolated pockets across Asia including Siberia and Tibet that tends to favor a Northern route origin. 

The mtDNA haplogroups N9b and M7a also tell story so deep in history (both are very basal in the Eurasian mtDNA tree and derived from African  mtDNA haplogroup L3) that it is hard to reconstruct. Both mtDNA M and mtDNA N show distributions that tend to favor a Southeast Asian route to Japan, but perhaps this is because the northern bearers of this haplogroup went extinct, and were then almost fully replaced in the Last Glacial Maximum. 

See also this paper on Jomon and Ainu mtDNA, noting that by the Edo period, 29/94 of Ainu people in Hokkaido had characteristically Jomon mtDNA i.e. mtDNA N9b1, M7a2, G1b*; 33/94 has Okhotsk mtDNA (a NE Asian population with Y1 and C5a2b); 6/94 had Siberian mtDNA (D4o1, G1b1 and Z1a), and 26/94 has "mainland" Japan mtDNA (D4xD4o1, M7b1a1a1, F1b1a, N9a, M7a1a7, A5a, and A5c) (a classification I'm not inclined to fully agree with).

The Yayoi 

The Yayoi described as a distinct wave tends to support the hypothesis that Japanese is part of an Altaic macro-language family, with the Korean and Japanese languages probably having the most close affinity to the Tungusic languages of Eastern Siberia and Manchuria (shown in red below) prior to heavy borrowing from Chinese (each of the three maps below are from Wikipedia). As the Altaic language link above explains:

With fewer speakers than Mongolic or Turkic languages, Tungusic languages are distributed across most of Eastern Siberia (including the Sakhalin Island), northern Manchuria and extending into some parts of Xinjiang and Mongolia. Some Tungusic languages are extinct or endangered languages as a consequence of language shift to Chinese and Russian. In China, where the Tungusic population is over 10 million, just 46,000 still retain knowledge of their ethnic languages.

Scholars have yet to reach agreement on how to classify the Tungusic languages but two subfamilies have been proposed: South Tungusic (or Manchu) and North Tungusic (Tungus). Jurchen (now extinct; Da Jin 大金), Manchu (critically endangered; Da Qing 大清), Sibe (Xibo 锡伯) and other minor languages comprise the Manchu group.

The Northern Tungusic languages can be reclassified even further into the Siberian Tungusic languages (Evenki, Lamut, Solon and Negidal) and the Lower Amur Tungusic languages (Nanai, Ulcha, Orok to name a few).

Significant disagreements remain, not only about the linguistic sub-classifications but also some controversy around the Chinese names of some ethnic groups, like the use of Hezhe (赫哲) for the Nanai people.

The spread of agriculture is often marked by population replacement, as documented in the Neolithic transition throughout most of Europe, with only minimal contributions from hunter-gatherer populations observed in many regions. However, we find genetic evidence that the agricultural transition in prehistoric Japan involved the process of assimilation, rather than replacement, with almost equal genetic contributions from the indigenous Jomon and new immigrants at the Kyushu site. This implies that at least some parts of the archipelago supported a Jomon population of comparable size to the agricultural immigrants at the beginning of the Yayoi period, as it is reflected in the high degree of sedentism practiced by some Jomon communities. 

The continental component inherited by the Yayoi is best represented in our dataset by the Middle Neolithic and Bronze Age individuals from the West Liao River basin with a high level of Amur River ancestry (i.e., WRL_BA_o and HMMH_MN). Populations from this region are genetically heterogeneous in time and space. The Middle-to-Late Neolithic transition (i.e., between 6.5 and 3.5 ka ago) is characterized with an increase in Yellow River ancestry from 25 to 92% but a decrease in Amur River ancestry from 75 to 8% over time, which can be linked to an intensification of millet farming. However, the population structure changes again in the Bronze Age, which started around 3.5 ka ago, due to an apparent influx of people from the Amur River basin. This coincides with the beginning of intensive language borrowing between Transeurasian and Sinitic linguistic subgroups. Excess affinity to the Yayoi is observable in the individuals who are genetically close to ancient Amur River populations or present-day Tunguisic-speaking populations. Our findings imply that wet rice farming was introduced to the archipelago by people who lived somewhere around the Liaodong Peninsula but who derive a major component of their ancestry from populations further north, although the spread of rice agriculture originated south of the West Liao River basin. 

Further linguistic analysis can be found in another recent paper referenced in this one, which argues that the original macro-Altaic homeland was an early Neolithic one in the West Liao River basin.

The Kofun

The Kofun wave of migration was Han Chinese-like and from the Southern Korean peninsula. 

The most noticeable archaeological characteristic of Kofun culture is the custom of burying the elite in keyhole-shaped mounds, the size of which reflect hierarchical rank and political power. The three Kofun individuals sequenced in this study were not buried in those tumuli, which suggests that they were lower-ranking people. Their genomes document the arrival of people with majority East Asian ancestry to Japan and their admixture with the Yayoi population. This additional ancestry is best represented in our analysis by Han, who have multiple ancestral components. A recent study has reported that people became morphologically homogeneous in the continent from the Neolithic onward, which implies that migrants during the Kofun period were already highly admixed. 

Several lines of archaeological evidence support the introduction of new large settlements to Japan, most likely from the southern Korean peninsula, during the Yayoi-Kofun transition. Strong cultural and political affinity between Japan, Korea, and China is also observable from several imports, including Chinese mirrors and coins, Korean raw materials for iron production, and Chinese characters inscribed on metal implements (e.g., swords). Access to these resources from overseas brought about intensive competition between communities within the archipelago; this facilitated political contact with polities in the continent, such as the Yellow Sea coast, for dominance. Therefore, continuous migration and continental impacts are evident throughout the Kofun period. Our findings provide strong support for the genetic exchange involved in the appearance of new social, cultural, and political traits in this state formation phase.

The Ancient DNA Samples Used

The paper looks at nine new Jomon ancient DNA samples in addition to three old Jomon ancient DNA samples (in all four men and eight women from ca. 6819 BCE to 569 BCE), two old Yayoi ancient DNA samples (a man from ca. 44 CE and a woman), and three new Kofun ancient DNA samples (a man and two women from ca. 622 CE to 675 CE) (although the dating would suggest that they are actually from the Asuka era).

The paper further explains its samples as follows:

Here, we report 12 newly sequenced ancient Japanese genomes spanning 8000 years of the archipelago’s pre- and protohistory. To our knowledge, this is the largest set of time-stamped genomes from the archipelago, including the oldest Jomon individual and the first genomic data from the imperial Kofun period. We also include five published prehistoric Japanese genomes in our analysis: three Jomon individuals (F5 and F23 from the Late Jomon period and IK002 from the Final Jomon period), as well as two 2000-year-old individuals associated with the Yayoi culture from the northwestern part of Kyushu Island, where skeletal remains exhibit Jomon-like characters rather than immigrant types but other archaeological materials clearly support their association with the Yayoi culture. Despite this morphological assessment, these two Yayoi individuals show an increased genetic affinity to present-day Japanese populations compared with the Jomon, implying that admixture with continental groups was already advanced by the Late Yayoi period.

The paper later notes that:

Our kinship analysis confirms that all pairs of individuals are unrelated. 

The conclusion of the paper notes that the main limitation of its findings is the small size of its ancient DNA samples for each period which may not reflect geographic diversity and population structure in each time period (especially the Yayoi and Kofun periods).