For  a long while, many writers have theorized that Austronesian-speaking migrants must have reached  Japan, but attempts to clearly identify Austronesian genomic segments of the Japanese or ancient Jomon as Austronesian have not been successful so far.

Below we set out a series of excerpts of articles and materials that present possible sources of genetic material for Austronesian populations and their homeland, including Japan.

A number of things become clear from reading the literature: There is no consensus as to whether there was a large-scale Austronesian expansion, or even whether it was an agriculture-related dispersal or if it was a seafaring-related or inland vs. coastal one. While there is an agreement generally on the existence of the Austronesian language, a broad suite of characteristic Austronesian technology, there is no agreement as to the routes of dispersal/expansion, or of the chronological timeline of arrivals, nor of the origins/genetic makeup of the Austronesians themselves. Writers also appear to have great difficulty disentangling the genetic components between Austroasiatic and Austronesian people given that most people today are admixtures of both.

There are no conclusive or easy answers to the questions of if, who, or when the Austronesian people arrived in Japan. It is plausible and easiest perhaps to follow Hudson and Sagart’s theories — the former could characterize the arrival of some Austronesian (or Proto-Austronesian?) people south of the Ryukyu during the Jomon period (B4e(Mtdna) was found at the Shiraho-Saonetsu site with a  Taiwan or Southeast Asian signal, as well as a lone Urawa y-DNA E1b1b among the Honshu Jomon); The latter could see the O3c(ydna) dispersals into Mainland Japan as a significant late Austronesian-related dispersal, although these do not arrive out-of-Taiwan.

1. This paper by Mark J. Hudson, The Ryukyu Islands and the Northern Frontier of Prehistoric Austronesian Settlement, Research Gate Feb 12, 2018., argues that the southern Ryukyu Islands should be included within the Austronesian cultural sphere..

2. Peter Bellwood’s Austronesian Prehistory in Southeast Asia Homeland, Expansion, and Transformation looks at the question of the Ultimate Homeland of the Austronesians as well as where the Proto-Austronesians might have arisen. Making the connection between the expansion of rice-agriculture and Austronesians, he writes:

”One has to consider very seriously the possibility that the initial expansions of Austronesian and Tai-Kadai languages (and probably also Austroasiatic) began among Neolithic rice-cultivating communities in China south of the Yangzi. The archaeological record agrees very well and provides a date range for initial developments between 5000 and 4000 BC.
Moving beyond Austro- Tai into Austronesian proper, the reconstruction of linguistic prehistory which is most widely used today is that postulated by Robert Blust (1984-5). This is based on a “family tree” of subgroups and a hierarchy of proto-languages extending from Proto-Austronesian (PAn) forwards in time. Reduced to its essentials, Blust’s reconstruction favours a geographical expansion beginning in Taiwan (the location of the oldest Austronesian languages…”

3. The Ami and Atayal are identified as the ancestral Austronesians. Two routes for expansions out of southern China or East Asia (with the coastal one being the AN dispersal), are proposed by Xia et al., Inland-coastal bifurcation of southern East Asians revealed by Hmong-Mien genomic history  available under a CC-BY-NC-ND 4.0 International license.

ABSTRACT
The early history of the Hmong-Mien language family and its speakers is elusive. A good variety of Hmong-Mien-speaking groups distribute in Central China. Here, we report 903 high-resolution Y-chromosomal, 624 full-sequencing mitochondrial, and 415 autosomal samples from 20 populations in Central China, mainly Húnán Province. We identify an autosomal component which is commonly seen in all the Hmong-Mien-speaking populations, with nearly unmixed composition in Pahng. In contrast, Hmong and Mien respectively demonstrate additional genomic affinity to Tibeto-Burman and Kra-Dai speakers. We also discover two prevalent uniparental lineages of Hmong-Mien speakers. Y-chromosomal haplogroup O2a2a1b1a1b-N5 diverged ~2,330 years before present (BP), approximately coinciding with the estimated time of Proto-Hmong-Mien (~2,500 BP), whereas mitochondrial haplogroup B5a1c1a significantly correlates with Pahng and Mien. All the evidence indicates a founding population substantially contributing to present-day Hmong-Mien speakers. Consistent with the two distinct routes of agricultural expansion from southern China, this Hmong-Mien founding ancestry is phylogenetically closer to the founding ancestry of Neolithic Mainland Southeast Asians and present-day isolated Austroasiatic-speaking populations than Austronesians. The spatial and temporal distribution of the southern East Asian lineage is also compatible with the scenario of out-of-southern-China farming dispersal. Thus, our finding reveals an inland-coastal genetic discrepancy related to the farming pioneers in southern China and supports an inland southern China origin of an ancestral meta-population contributing to both Hmong-Mien and Austroasiatic speaker

Introduction

Starting ~9,000 years before present (BP), China is the second earliest birthplace of agriculture, following the Near East1,2. Subsequently, farming dispersed into adjacent areas of East Asia, especially Southeast Asia, the Korean Peninsula, and the Japanese archipelago. However, it is still not fully settled whether and to what extent human migration propelled the dispersal of agriculture. Particularly for Southeast Asia, recent archaeogenetic studies support that the East Asian ancestry of the first farmers in Southeast Asia can be traced to southern China3,4, raising the further issue on the deeper demographic prehistory of both regions….

Currently, Hmong-Mien speakers mainly dwell in South Central and Southwest China, with sporadic distribution in Mainland Southeast Asia (MSEA) due to the migration within the recent centuries7. The linguistic phylogeny of the Hmong-Mien languages remains controversial, while a common hypothesis categorizes them into two sub-branches: Hmongic and Mienic. Whereas the Hmongic languages receive more influence from Tibeto-Burman languages, the Mienic languages show more impact from Chinese. In Húnán Province of Central China, both clades distribute, including two Hmongic languages: Hmong (to be specific, Qo Xiong) in the northwest Húnán and Pahng (Hm Nai) in the southwest, as well as a Mienic language (Mien, or specifically, Iu Mien) in the southeast.

Approximately synchronous with the time estimation of Proto-Hmong-Mien (~2,500 BP)19, we estimate the most recent common ancestor (TMRCA) of O2a2a1b1a1b-N5 as ~2,330 BP (Fig. 5B), consistent with slight precedence of the divergence of patrilineage than the dispersal of language and ancestral population. By phylogenetic analysis (Fig. 5A), most of the Pahng (with one exception) and Mien (with two exceptions) samples affiliated to mitochondrial haplogroup B5 are identified as the subclade B5a1c1a (TMRCA in ~9,800 BP). Given the substantially higher frequency of Hmong-Mien speakers (63.6%, 21 out of 33) than the others in B5a1c1a, we propose a Hmong-Mien origin of this subclade.
In summary, we find strong evidence from genomic and uniparental analyses for a founding population who is responsible for the dispersal of Proto-Hmong-Mien and thus has a substantial genetic impact on present-day Hmong-Mien speakers, hence naming it as “Ancestral Hmong-Mien” (AHM).
Inland-coastal bifurcation of southern East Asians…

One of the core issues regarding the history of the Hmong-Mien language family is its place of origin. Whereas linguistic evidence, such as the reconstructed vocabulary related to wet-rice cultivation, supports a Yangtze Basin origin of Hmong-Mien7, archaeological record reveals a quite more complex scenario. During the Neolithic, there were two independent agricultural centers in southern China: Yangtze Delta and the middle Yangtze1,20 (here we definite the northern boundary of “southern China” as Qín Mountains and Huái River). The former partly explains the coastal route for farming expansion as far as Taiwan and Luzon, while the latter accounts for another inland expansion of agriculture into Southwest China and MSEA21. In particular, there is a strong bond between the material culture of Neolithic Yangtze Delta and later Austronesian Pacific islands previously described22. Nevertheless, it is still ambiguous about the relationship between Proto-Hmong-Mien and these agricultural centers. Therefore, to determine the genomic relationship between AHM and other populations is pivotal to the resolution of the issue. …

Besides AHM, previous genetic studies have already distinguished two other ancestral populations highly associated with the farming dispersal from southern China3,4,23. One of them is represented by the East Asian ancestry of Neolithic farmers in MSEA, as well as present-day isolated Austroasiatic-speaking groups (e.g., Htin and Mlabri), who largely maintain genetic continuity with the former4, hence responsible for Austroasiatic expansion and named as “Ancestral Austroasiatic” (AAA). The other is represented by Austronesian-speakers in Taiwan (e.g., Amis and Atayal) in nearly unmixed form23 and accounts for the dissemination of Austronesian languages, hence addressed as “Ancestral Austronesian” (AAN).
Given that, we computed a series of D-statistics16 to address the genomic relationship between AHM and other populations (Table 2, Extended Data Table 2 & 3). Characterized by significant negative D (X, Mbuti; Devil’s Cave, Amis) [X = Hàn, Hmong (representing AHM), Htin/Mlabri (representing AAA); Z 0, Z = 7.0, Table 2], and Hàn receives slightly more genomic influence from northern East Asians than Hmong [D (X, Mbuti; Hmong, Hàn) < 0, Z = −2.9 for Nganasan, −2.7 for Devil’s Cave]. Devil’s Cave inclines towards Austronesian in Taiwan than isolated Austroasiatics in D-statistics [D (Devil’s Cave, Mbuti; Htin/Mlabri, Amis/Atayal) < 0, Z < −7.9], compatible with a deep ancestry from Australasian lineage harbored by isolated Austroasiatics as previously proposed4.
Revealed by the direction of D-statistics, different modern and ancient populations in Southeast Asia do not equally relate to Hmong, Hàn, and Amis, and can primarily categorize into three groups (Table 2). Most of Neolithic and Bronze Age samples in MSEA and isolated Austroasiatics incline towards Hmong than Hàn but symmetrically relate to Hmong and Amis. Austronesians in Luzon and Visayas highly deviate to Amis from Hmong, but symmetrically relate to Hmong and Hàn. Present-day Kra-Dai speakers and Khmer possess increase affinity both to Hmong than Hàn and to Amis than Hmong. The results above imply an intricate internal structure within the southern East Asian lineage

Given the current and traceable historical distribution of relevant populations, we provisionally call the lineage constructed by AHM and AAA as an “inland lineage” and the one represented by AAN as a “coastal lineage”. …

Considering the arrival of such an AAN-like lineage in MSEA, we find that the signal first occurred in the Bronze Age Vietnamese (dated to ~2,000 BP) who shows an increasing affinity to Amis and a decreasing affinity to Hmong (Table 3). Given the Austro-Tai hypothesis24 that proposes a genealogical relationship between Kra-Dai and Austronesian languages, the occurrence of the coastal lineage in MSEA is appropriate to be explained as a southwestward migration wave related to Kra-Dai expansion3 after the presence of AAA in this region. Distinct from other Austronesians, both Malay and Dusun show an increasing affinity to Hmong than Han…

All the evidence indicates an earlier occurrence of the inland  lineage in inland Southwest China and MSEA and supports an inland-coastal bifurcation of the southern East Asians, corresponding to the two distinct routes for agricultural expansion from southern China.Given the current and traceable historical distribution of relevant populations, we provisionally call the lineage constructed by AHM and AAA as an “inland lineage” and the one represented by AAN as a “coastal lineage”.

[O2a2a1b1-M209 [O2a2a1b1a1b-N5 for Hmong-Mien speakers; O2a2a1b1a1*-F2309(xM113, xN5) for Kinh17;

O2a2a1b2-F5511 for Amis;

mitochondrial haplogroup B5a1 [B5a1c1a for Hmong-Mien speakers; B5a1a for Vietnam_N and Thailand_LN_IA4; B5a1c for Thailand_LN_IA4]]

Diffusion of the southern East Asian lineage
Neolithic agriculture in Yangtze Basin is famous for its earliest domestication of sinica/japonica rice (Oryza japonica), which later spread into the vast tract of land ranging from Japan to Madagascar1,30. To test whether the agricultural diffusion from southern China is concomitant with the emigration from the same region, we used the qpAdm31 to model the spatial and temporal distribution of the southern East Asian lineage, comprising both the inland (i.e., AHM and AAA) and coastal (i.e., AAN) lineages (Fig. 7 & Extended Data Table 4).
We started from the two-way model of Amis (representing the southern East Asian lineage) and Devil’s Cave (representing the northern East Asian lineage), adding additional reference populations (Hòabìnhian, Afanasievo, Namazga, and Kolyma) when the initial model fails (see Method). In addition to MSEA populations (9.5–30.0%)3,4, South Asia32,33 (Kharia, 61.6 ±2.2%), and the Jōmon individual (58.1 ±2.7%) dated to ~2,600 BP3, Hòabìnhian-like component also occurs in present-day Tibetans (32.3 ±2.0%), suggesting a once wide-ranging distribution of the Australasian lineage in these regions prior to the agricultural expansion in East Asia. Regarding the East Asian lineage, there is no significant difference in fitness of Jōmon using either Devil’s Cave (p =0.619) or Amis (p =0.558) as the proxy of East Asian, suggesting that the East Asian ancestry of Jōmon may phylogenetically be basal to the mainstream northern or southern East Asians. By contrast, the East Asian ancestry of Tibetan, which is probably related to the Tibeto-Burman dissemination in Tibetan Plateau, can be represented by Devil’s Cave (67.7 ± 2.0%, p = 0.178) rather than Amis (p = 1.24×10-4), consistent with the North China origin of the Sino-Tibetan language family34,35.
Fitting the scenario of the expansion of Neolithic farmers in southern China, southern East Asian lineage represented by Amis predominates in Hàn (69.3 ±2.8%), other populations in southern China (69.2–78.6%), and populations in MSEA (54.3–76.7%). Hàn and Tibeto-Burman-speaking populations (Nàxī, Yí, and Tǔjiā) tend to have more Devil’s Cave ancestry (23.5–30.7%) than their neighbors, consistent with our previous inference. Compared with isolated Austroasiatics, Kinh and Thai/Khmer respectively harbor additional Devil’s Cave (18.7 ±7.5%) and Namazga (8.4 ±1.0% for Thai, 3.7 ±1.1% for Khmer) ancestry, likely reflecting immigration from Hàn Chinese and South Asians in the historical period, accordingly. Apart from MSEA, Austroasiatic-speaking Kharia in Central India is also estimated with Ami-like component (18.5 ±3.0%), indicating a southern East Asian origin of the Muṇḍā branch.
Although the Devil’s Cave ancestry is generally the predominant East Asian lineage in North Asia and adjacent areas, there is an intriguing discrepancy between the eastern [Korean, Japanese, Tungusic (except northernmost Oroqen), and Mongolic (except westernmost Kalmyk) speakers] and the western part [West Xiōngnú (~2,150 BP)36, Tiānshān Hun (~1,500 BP)36, Turkic-speaking Karakhanid (~1,000 BP)36 and Tuva, and Kalmyk]. Whereas the East Asian ancestry of populations in the western part has entirely belonged to the Devil’s Cave lineage till now, populations in the eastern part have received the genomic influence from an Amis-related lineage (17.4–52.1%) posterior to the presence of the Devil’s Cave population roughly in the same region (~7,600 BP)12. Analogically, archaeological record has documented the transmission of wet-rice cultivation from coastal China (Shāndōng and/or Liáoníng Peninsula) to Northeast Asia, notably the Korean Peninsula (Mumun pottery period, since ~3,500 BP) and the Japanese archipelago (Yayoi period, since ~2,900 BP)2. Especially for Japanese, the Austronesian-related linguistic influence in Japanese37 may indicate a potential contact between the Proto-Japonic speakers and population(s) affiliating to the coastal lineage. Thus, our results imply that a southern-East-Asian-related lineage could be arguably associated with the dispersal of wet-rice agriculture in Northeast Asia at least to some extent.

DISCUSSION
In this study, we discover an explicit genomic pattern regarding Hmong-Mien speakers and other southern East Asians, which has been established in Neolithic and is still detectable in present-day populations. After the separation from the northern East Asians, southern East Asians further divided into two lineages genetically, concordant with archaeological and linguistic evidence. While the inland lineage is associated with the farming dispersal from the middle Yangtze and the dissemination of the Austroasiatic and Hmong-Mien language family, the coastal lineage is more correlated with the agriculture expansion out of Yangtze Delta and the diffusion of the Austronesian and (at least partially) Kra-Dai language family. 

…

Further  excerpts:

Comparing plant materials from Shandong, lower Yangtze and Taiwan neolithic sites
To further illustrate the differences between the NES and LY Neolithic, we compare the domesticated and non-domesticated plants found in Shandong, Taiwan and Hangzhou Bay area neolithic sites (Table 4). Foxtail millet and broomcorn millet were present in Shandong and Taiwan but have not been found in Lower Yangtze/Hangzhou Bay sites. Aquatic nuts (Trapa spp., Euryale ferox) formed an important part of the subsist- ence in the Hangzhou Bay Neolithic (Deng et al. 2015) but are virtually unknown in early Neolithic sites in Taiwan and are rare in the Shandong Houli and Beixin/ Dawenkou cultures. Wild barnyard grasses (Echinochloa spp.) were harvested and consumed before 5000 BCE in the Hangzhou Bay area (Yang et al. 2015) but have not been reported as a significant source of food in either Taiwan or the Houli and Beixin/Dawenkou cultures of Shandong. Finally, ritual tooth ablation, present in the Houli and Beixin/Dawenkou cultures of Shandong, in the early Formosan Neolithic and in scattered locations between Shandong and Taiwan, has not been reported in the main Hangzhou Bay sites.

The hypothesis of a Shandong origin of the Formosan neolithic
To recapitulate, the presence in Shandong well before the onset of the Formosan neolithic of an agricultural system associating foxtail, broomcorn and small quantities of rice, accompanied by ritual tooth ablation, make Shandong the stronger candidate precursor of the Formosan Neolithic (Ko et al. 2014; Sagart 1995; Fuller et al. 2010; Stevens et al. 2016; Sagart 2008).
The population expansion signal detected at c. 6000– 8000 BCE in the Austronesian mtDNA E haplogroup by geneticists (Ko et al. 2014) may represent millet-fueled population growth c. 8000 BCE preceding and during the early Houli culture, followed at c. 6000 BCE by the addition of rice to the original repertoire. Population growth stimulated by diversified cereal agriculture led groups in north Shandong to expand south (since during the climatic optimum, Shandong was the northern limit of rice cultivation) shortly afterwards, their expansion materialized by the southward progress of tooth ablation. We suggest that in the late 4th millennium BCE, these groups, some of whose members carried the mtDNA M9E haplogroup and/or the Y chromosome O3a2b2-N6 haplogroup, introduced to Taiwan the Proto- Austronesian language; a mixed farming, fishing and inter- tidal foraging subsistence strategy; domesticated landraces of foxtail millet, broomcorn millet and japonica rice; a be- lief in the sacredness of foxtail millet; ritual ablation of the upper incisors in adolescents of both sexes; domesticated dogs; and a technological package including inter alia houses, nautical technology, and loom weaving. Better than other models, the hypothesis of a southward demic expansion out of Shandong provides a credible account of the Austronesian settlement of Taiwan.

Conclusion
Our botanically informed linguistic fieldwork converges with earlier findings in archaeology and genetics to assign a lesser role for rice than was earlier thought, and a more important one for the millets. Our study of domestication genes in a collection of traditional rice landraces maintained by the Formosan aborigines shows that early Taiwanese rices were introduced to the island in already domesticated form. We argue that domesticated rice and millets were brought to Taiwan by a population having expanded south along the coast from Shandong after c. 5000 BCE, reaching western Taiwan in the second half of the 4th millennium BCE.

…

5. Van Driem’s Rice and the Austroasiatic and Hmong-Mien homelands posit that y-chromosome O1a is likely the genetic imprint of the Austronesians:

”the currently emerging Y chromosomal picture based on single nucleotide polymorphisms suggests that Kradai peoples could descend mainly from ancient Hmong-Mien and Austroasiatic language communities which were linguistically assimilated by ancient Austronesian remigrants to the East Asian mainland. The Y chromosomal haplogroup M119-O1 (O1a) occurs at a high frequency amongst the Austronesian aboriginal peoples of Formosa and also, albeit in a much lower frequency, in the Philippines and southeastern China, especially in Kradai language communities (Abdulla et al. 2009). The paternal genetic imprint of the ancient Austronesians is but faint in comparison to their linguistic and cultural impact, which has extended from Formosa across half the planet.11”

6. Soares et al., Resolving the ancestry of Austronesian-speaking populations consider the range of putative ancestral Austronesian haplogroups and haplotypes and find the timeline of arrivals in ISEA of most of them to be either too early, or spotty to have formed the “Out-of-Taiwan” hypothesis. The best candidate they find is M7c3, but they conclude there is no large-scale migration, nor a rice-agriculture-related migration:

“MSEA is a source region in this analysis, so this value in the founder analy- sis corresponds to ancient lineages private to ISEA only. In the mtDNA analysis, lineages descending directly from the haplogroups carried by the first settlers correspond to M*, N*, R* and possibly haplogroup F3 (Fig. 1d). Although a recently published ancient mtDNA haplogroup E sequence (Ko et al. 2014) was used to suggest a Taiwanese source for this clade, an early origin in ISEA (Soares et al. 2008) remains more likely, as discussed below. At this ancient time-frame, Y-chromosome lineages (with STR ρ dating) are uninformative due to saturation, but haplogroups K* and even C may date to the first colonization at that time. These are above 30 % in the Y-chromosome analysis.
Overall, the migration at ~8 ka contributes the most line- ages to the current gene pool of ISEA with a fraction of ~40–50 % in both mtDNA and Y-chromosome variation (Fig. 1c). We stress again that, statistically, this migration time could include lineages entering ISEA throughout the period of sea-level rises, from 14 to 8 ka, covering all three flooding episodes (Pelejero et al. 1999). This partition prob- abilistically includes major and well-studied haplogroups such as B4a1a (Soares et al. 2011), subclades of haplo- group E (Soares et al. 2008), F1a*, and subclades of hap- logroup M shared between ISEA and MSEA, with B4a1a and E the major contributors. In Y-chromosome variation, this migration includes most clusters within haplogroups O2a1 and O3 and a subclade of O1a (Fig. S4), matching to some extent the results of Karafet et al. (2010) indicating that O1a* entered ISEA before the Neolithic. We should note that in our recent Y-chromosome survey (Trejaut et al. 2014), O2 and O3 clades declined in frequency moving north from ISEA towards Taiwan, the opposite of what one might expect from an “out-of-Taiwan” movement. A pre- vious survey (Karafet et al. 2010) also suggested that O3, O2a1 and O1a* entered ISEA from the mainland before the Neolithic period.
The contribution at the time of the Neolithic, at 4–5 ka, varied with the criterion and the genetic system, but 25–35 % is probably the best estimate (Fig. 1c). (The f1 criterion in mtDNA probably overestimates recent migra- tion due to the large size of the source sample used.) Only one major founder presented significant differences between the analyses: B4b appears Neolithic in f1 crite- rion and part of the postglacial migration in the f2 criterion (Fig. 1d). This haplogroup deserves further attention in the future. The widely held model for the spread of the Neo- lithic in ISEA implicates expanding pre-Austronesian/Aus- tronesian speakers from South China/Taiwan (Bellwood 1997); but in fact not all of the Neolithic founders we iden- tify support this hypothetical “out-of-Taiwan” dispersal. A large fraction of Neolithic mtDNA founder clusters from haplogroups B5a1 and F1a1a (~10 % out of the 25–35 % Neolithic lineages in the analysis) appear to have originated in MSEA, and are rare or absent in either Taiwan or the Philippines.
Our results therefore suggest that mid-Holocene Neo- lithic immigration into ISEA was in part via MSEA, temporally associated with spread of basket-marked and carved paddle-impressed pottery, which appeared across MSEA as early as red-slipped pottery appeared in Tai- wan (Bulbeck 2011), and possibly involving speakers of Austroasiatic languages (i.e. Anderson’s “Neolithic I”) (Anderson 2005). The mtDNA haplogroups M7c3c, Y2, F1a4a, B4c1c and possibly B4b (which shows contrast- ing patterns under the two criteria) may, however, rep- resent genuine “out-of-Taiwan” clades in ISEA. These founders are all derived from Chinese-mainland source haplogroups, and within Austronesian-speaking popula- tions they have a higher overall frequency in Taiwan and the Philippines (Fig. 2a). This input, at ~20 %, lends sup- port to a modified, small-scale “out-of-Taiwan” model [Anderson’s “Neolithic II” (Anderson 2005; Donohue and Denham 2010)], proposed to explain the appearance of red-slipped pottery in relation to the early dispersal of Austronesian languages.
On the male line of descent, the Neolithic contribution is lower (15–20 %) but, since MSEA is not represented in the Y-chromosome dataset, all these Neolithic founders are likely to represent the putative “out-of-Taiwan” dispersal, mirroring closely the ~20 % “out-of-Taiwan” founders for mtDNA. Most of O1a and all of O1a2 likely represent sig- nals of Neolithic migrants from Taiwan, confirming earlier suggestions (Karafet et al. 2010; Trejaut et al. 2014). A por- tion of O3a (~10 % in the f1 criterion) was also partitioned into the Neolithic in our analysis….”

“Haplogroup M7 dates to just over 50 ka. An overall mainland Eastern Asian distribution is clear for the M7 phylogeny (Fig. 4; full tree in Supplementary Material 2). There are two basal branches, M7a, which displays a strong Northeast Asian ancestry centred on Japan and Korea, and a second major clade encompassing M7b, M7c, M7d, M7e, M7f and M7g, which we refer to as M7b′c′d′e′f′g. This splits into two further major subclades, M7b′d′g and M7c′e′f both with an East Asian ancestry.
The overall phylogenetic and phylogeographic pattern is strikingly clear: both aboriginal Taiwanese and Island Southeast Asian-specific lineages are close to the tips of an overall mainland Eastern Asian distribution. The major subclade of M7b3, M7b3a, is only present in Taiwan and ISEA. It is frequent in Taiwan (at ~10 %) and considering its age (~6 ka) seems likely to have arrived in Taiwan with the rice Neolithic from South China; but it is vanish- ingly infrequent across ISEA. In M7b1, M7b1d3 is also restricted to Taiwan, and with a similar age may also have arrived from China with the Neolithic, but again it is virtu- ally absent from ISEA.
In M7c′e′f, the three subclades branch from a single node and all show evidence of East Asian ancestry. Within M7c, M7c3 is by far the most frequent and the only one to disperse significantly into Taiwan and ISEA. This clade probably had an origin in South China, with several sub- clades also present in Taiwan. Its major subclade, M7c3c [M7c1c in Hill et al. (2007)], here re-dated with whole- mtDNA genomes to ~5 ka, is restricted to Austronesian- speaking populations (both Taiwan and ISEA). Given the presence of other subclades of M7c3 in Taiwan and South China, the most probable source for M7c3c is in Taiwan (amongst M7c3 arrivals from China, again perhaps with the rice Neolithic), with subsequent dispersal into ISEA. Several subclades of M7c3c exist throughout Taiwan and ISEA, and there is also one in the Pacific (M7c3c2, found in both Micronesia and the Solomon Islands), dating to less than 3 ka. This pattern confirms M7c3c as a strong candidate for an “out-of-Taiwan” marker, as indicated by the HVS-I founder analysis.
We can contrast this distinctive pattern with the distribution of haplogroups B4a1a and E, both of which are—like M7c3c—largely restricted to insular, Austronesian-speaking populations. For that reason they have been proposed as candidates for “out-of-Taiwan” markers, but neither shows a direct ancestry in South China …

Here we show that two Neolithic waves entered ISEA, as previously suggested on the basis of pot- tery comparisons (Anderson 2005) and recently from auto- somal analyses (Lipson et al. 2014), but that both were small-scale affairs.
The first Neolithic migration, from MSEA [“Neolithic I” in the scheme of Anderson (2005)], reflected in the distribution of haplogroups B5a1 and F1a1a and the “pale green” genome-wide component, took place ~4.5 ka and affected mainly Western Indonesia/Borneo—although it extended as far as Eastern Indonesia, particularly in the south, even reaching regions of contact with Papuan populations. A signal for this dispersal was also recently proposed by Lipson et al. (2014), although they favoured admixture with Austronesian agriculturists dispersing around the coasts of MSEA as an explanation, which our results render unlikely.
The second Neolithic wave [“out-of-Taiwan” or Anderson’s “Neolithic II” (Anderson 2005)] is marked by the appearance of red-slipped pottery ~4 ka (Spriggs 2007, 2011) and impacted strongly on the Philippines (accounting for 30–40 % of current genetic diversity), where domesti- cated rice does indeed appear relatively early in the archaeological record (Paz 2002). However, for the rest of ISEA (the Indo-Malaysian archipelago), the demographic impact was much lower—often negligible. The overall fractions of “out-of-Taiwan” immigrants in the founder analysis for both mtDNA and Y-chromosome variation are very similar at ~15 to 20 %, suggesting that previous models inferring highly divergent male and female contributions are incor- rect (similarly to the Pacific). The mtDNA haplogroup M7c3c, in particular, closely matches the expected pattern for an “out-of-Taiwan” marker.
Thus, although the Neolithic dispersal from Taiwan suggested by red-slipped pottery proves not to have been a large-scale demographic event (at least, beyond the Philippines), it did indeed occur, and followed an expansion into Taiwan from South China, as one archaeological model predicts (Bellwood 1997). However, we must be careful what we mean by the term “Neolithic”, since the archae- ological record for most of ISEA primarily indicates the appearance of various novel ceramics, and provides little or no evidence for large changes in the subsistence base. The low level of settlement across ISEA at this time accords not with large-scale demic diffusion fuelled by rice agricul- ture, but with more with archaeological views that stress the transition from grain cultivation to the root and arboreal crops that dominate agricultural systems in the west- ern Pacific (Donohue and Denham 2010; Paz 2002). It is clearly parsimonious to conclude that these sea-faring settlers spoke Austronesian and spread their languages across ISEA, but they may have had rather little to do with either rice farming or arboriculture/vegeculture (aspects of which originated much earlier, in part diffusing from Near Oce- ania (Barker and Richards 2013; Blench 2012).
The low scale of the migrations overall concurs with recent archaeological evaluations (e.g. Spriggs 2011), but contrasts sharply with the recent interpretation of Lip-son et al. (2014). However, their assumption that aborigi- nal Taiwanese represent the source for ISEA, their use of only three autosomal source clusters and their extremely recent age estimates for admixture times (within the last 2200 years) compromise their conclusions. Our analysis supports a scenario in which language shift played the major role, rather than large-scale population replacement (Donohue and Denham 2010, 2015).

The genetic situation further east seems to require a model where language was transmitted mostly horizontally across the north coast of New Guinea. Curiously, M7c3c (most or all probably belonging to the subclade M7c3c2 dating to ~2.6 ka) and some other putative “out- of-Taiwan” subclades (like B4b1) are detected at relatively high frequencies in Eastern Micronesia/Northwest Polynesia.  These lineages may have been carried directly through Western Micronesia from the vicinity of the Philippines (Fitzpatrick and Callaghan 2013; Hung et al. 2011). This migration was, however, distinct from the primary spread of the Austronesian languages into the Pacific, and would be expected to have affected mainly the Marianas.
Otherwise, whilst languages may have moved alongside other lineages integrated within ISEA, “out-of-Taiwan” haplogroups are virtually undetected across the north coast of New Guinea, the Bismarck Archipelago or the Solo- mon Islands. Minor exceptions include 1.4 % M7c3c in the Admiralty Islands (Kayser et al. 2008a), <0.2 % in New Britain (Friedlaender et al. 2007) and two closely related whole-mtDNA M7c3c sequences (~2 %) in the Solomon Islands (Duggan et al. 2014). M7c3c sequences, all within M7c3c2, are also seen in Ontong Java, a Polynesian outlier in the north Solomons. M7c3c and the other probable “out- of-Taiwan” clades have not been detected in Vanuatu, Fiji or Samoa, despite very extensive sampling.
Most of the present-day diversity in Near and Remote Oceania was established in New Guinea by ~10 ka (Soares et al. 2011), a fraction of which was carried by Austrone- sian speakers into the Remote Pacific.

Powerful, long- established spheres of interaction may have facilitated the spread of the Austronesian languages in the south (Bulbeck 2008; Terrell and Welsch 1997; Torrence and Swadling 2008). They may thus have spread stepwise from the north and west via small-scale interactions and waves of accul- turation. There appears to have been no “Austronesian farming-dispersal” in any meaningful sense across ISEA— early Austronesian speakers were more likely fisher–foragers—opening up the discussion to a range of innovative archaeological and linguistic models (Barker and Richards 2013; Donohue and Denham 2010, 2015). As both archaeologists and linguists have suggested, alluding to the spread of the early Metal Age in Europe, it may be that what began to spread across ISEA around 4000 years ago was primarily a new way of thinking—the adoption of a new ideol- ogy and perhaps even a new religion.”

The authors also caution us that:

”Frequencies of the latter [The Chinese/Northeast Asian genetic component] in Taiwan (~30 %) and Southeast Asia (5–30 %) match the mtDNA picture of Neolithic-age Chinese gene flow into ISEA (Fig. S8; cf. Fig. 2a). It is, however, difficult to directly connect a given component in ancestry analysis with a given demographic occurrence. One could calculate the time of admixture, but admixture ages are not necessarily indicative of time of migration (Lipson et al. 2014). In addition, the ages calculated are sometimes dubious and under-estimated ”

…

7. Lipson’s study, published in 2014 in Nature Communications, uses genetic data to reconstruct the population history of Austronesians and sheds light on population mixing and past migrations in this region:

Lipson, M. et al., Reconstructing Austronesian population history in Island Southeast Asia Nat. Commun. 5:4689 doi: 10.1038/5689 (2014)

“The Austronesian-speaking populations of Island Southeast Asia have been shown to be more closely related to aboriginal Taiwanese people than modern day populations inhabiting mainland Southeast Asia. …

The expansion of Austronesian-speaking populations is estimated to have begun approximately four to five thousand years ago and is thought to have roots in Taiwan. However, the ancestry of present day Austronesian populations remains largely unresolved.

David Reich and colleagues have analysed genome-wide genetic markers from individuals spanning 56 populations from Island Southeast Asia in order to examine the exchange of genes that occurred between genetically distinct populations; a process known as admixture.

The authors find that the genetic component of Austronesian speakers can be attributed to four discreet sources, including aboriginal Taiwanese, Thai and New Guinean populations. While a large proportion of this genetic component can be traced back to Taiwan, Western Austronesians were shown to have a strong Austro-Asiatic component. This suggests that Austronesian speakers may have migrated through Vietnam or the Malaysian peninsula, where they mixed with Austro-Asiatic populations on Mainland Southeast Asia before settling in Western Indonesia.

…

8. 2010, MS Peng  et al., Tracing the Austronesian Footprint in Mainland Southeast Asia

…diffusion in MSEA cannot be simply explained as a demic diffusion. Given the fact that the Mon–Khmer speakers had already occupied the middle and southern part of Vietnam before the arrival of the Austronesian immigrants (Bellwood 2006, 2007), contact between both populations might have involved extensive genetic admixture. During the process of admixture, one expanding Austronesian language (proto-Chamic) was imposed on or adopted by certain immigrants from ISEA with only a relative minor genetic contribution from the expanding Austronesian people. Conversely, the indigenous Mon–Khmers—the major genetic donors to the Chams, only contributed minor linguistic components as loan words to the Chamic language (Thurgood 1999; Southworth 2004). Thus, the Austronesian diffusion in MSEA was mainly a process of language shift (Cavalli-Sforza et al. 1994; Diamond and Bellwood 2003) by indigenous populations.
When we focused on the potential recent Austronesian components in the Chams that were shared with the ISEA populations (fig. 5 and table 2), similar DHS values were observed across several Austronesian populations from ISEA. However, the populations (Kota Kinabalu and Banjarma- sin) from Borneo—the potential source of Chamic (Blust 1994)—were relatively distant (DHS values: 0.923 and 0.931, respectively) from the Chams. Consequently, the Austronesian homelands of the Chams as well as the possible migration route of pioneer Austronesian people (and the language) to southern Vietnam cannot be pinpointed for the time being, although the bias caused by incomplete sampling of the ‘‘correct’’ source populations cannot be excluded. Taken together, our results support that cultural diffusion had played a dominate role during the spread of the Austronesian language into MSEA as suggested by the NMTCN hypothesis (Solheim et al. 2007). …”

9. Robert Blust writes that “The Austronesian language family is the second largest on Earth in number of languages, and was the largest in geographical extent before the European colonial expansions of the past five centuries. This alone makes the determination of its homeland a research question of the first order. There is now near-universal agreement among both linguists and archaeologists that the Austronesian expansion began from Taiwan, somewhat more than a millennium after it was settled by Neolithic rice and millet farmers from Southeast China. The first “long pause,” between the settlement of Taiwan and of the northern Philippines, may have been due to inadequate sailing technology, an obstacle that was overcome by the invention of the outrigger canoe complex. The second “long pause,” between the settlement of Fiji–Western Polynesia and of the rest of Triangle Polynesia, may also have been due to inadequate sailing technology, an obstacle that was overcome by the invention of the double-hulled canoe.” See…

Blust, Robert, The Austronesian Homeland and Dispersal (January 2019). Annual Review of Linguistics, Vol. 5, Issue 1, pp. 417-434, 2019. Available at SSRN:  https://ssrn.com/abstract=3335700  or  http://dx.doi.org/10.1146/annurev-linguistics-011718-012440

10. Focus on the Sa Huynh, a culture associated with the Austronesian-speaking Chamic peoplethe Austronesian migration to Central Vietnam crossing over the Iron Age Southeast Asian Sea

“Pleistocene seafarers could cross large expanses of open sea, as evidenced by the movement of early anatomically modern humans across Wallacea to Australia as far back as 50,000 years ago. Much later, Austronesian speakers (referred to simply as Austronesians from here on) based in Island Southeast Asia (ISEA) colonised the far-flung islands of the Indian Ocean and Oceania from Madagascar in the west to Easter Island in the east. Blust (1984–1985, 1996) and Bellwood (1997) reconstructed the overall linguistic and archaeological dispersal of Austronesians. A hypothesis on the homeland of Austronesian languages and the process of their dispersal was initially proposed by linguists and has significantly influenced the prehistoric archaeology of these regions. Bellwood has been working on verifying this hypothesis through archaeology and trying to reconstruct a complete history of the Austronesian dispersal that took place in various stages between 5,000 and 1,000 years ago (Bellwood 1997, 2004, 2005; Bellwood and Dizon 2005, 2008). His work suggests that the ancestors of the Austronesians originated in Southern China, and travelled to Taiwan – taking rice farming with them – probably by 5000 BP. Excavations on the Batanes Islands lying between Taiwan and Luzon identified possible evidence for the initial Neolithic dispersal out of Taiwan into the Northern Philippines by around 4000 BP (Bellwood and Dizon 2005, 2008, 2013). These Austronesians then proceeded to migrate south and east through the rest of the Philippine archipelago and into Sulawesi and Borneo before dispersing across the rest of ISEA and the Pacific.
Based on the linguistic work of Blust (1995), Bellwood described a branch of the Austronesian dispersal crossing over the South China Sea from Western Borneo to Vietnam around 2300 BP, perhaps made by people who spoke a language ancestral to Chamic (Bellwood 1997: 120–121). Chamic is actually the only Austronesian language spoken in Mainland Southeast Asia, being the language of the Chams, as well as some mountain dwellers presently living in Southern Vietnam and Cambodia.L
Bellwood thus inferred that the Iron Age Sa Huynh culture in Central Vietnam (Figure 19.1) was associated with an Austronesian-speaking (Chamic) population that had arrived in that region from either Peninsular Malaysia or Borneo, as documented archaeologically by the Sa Huynh culture itself (Bellwood 1997: 271–272).
A unique characteristic of the Sa Huynh culture is its mortuary customs, with lidded jar burials (Figure 19.2). These are associated with funerary accessory goods, such as pottery, iron and bronze implements, and earrings and beads made of agate, carnelian, jade (nephrite), and glass (see Nguyen Kim Dung, Chapter 18, this volume). The Sa Huynh cemeteries are often found on sand dunes extending along coasts or rivers on the alluvial plains. Although the Sa Huynh culture is recognised as having possibly fallen into decline in the latter half of the first century AD, the timing of its emergence still remains uncertain (Yamagata 2007a). Vietnamese archaeologists consider Champa to have emerged indigenously from local Sa Huynh (Ha 1999: 341), and that Sa Huynh populations like their Champa successors were Austronesian speakers (Ha 1984–1985).“

Because the jar burials are the most distinctive mortuary custom practised by the Vietnamese societies belonging to the Sa Huynh culture, and often found along the coasts on sand dunes, during the Iron Age, these hold great interest for the analogous parallels seen in the cultural package of Yayoi Japan. Particularly in the case of Doigahama site, where the burials were also in ceramics, and on the sand dunes, and where ritual tooth ablation was practiced.