A few thousand years ago the islands of Japan were settled by a group of rice farmers, ushering in the Yayoi period. Prior to this Japan was home of the Jomon culture, which is notable for being a pre-agricultural society which may have innovated to produce the world’s earliest pottery. Because everything before ~500 A.D. in Japan is prehistory (this date being generous) the relationship of the Jomon and Yayoi, and the ethnogenesis of the Japanese people as a homogeneous group, is somewhat speculative. But, it does seem one group of Jomon descended people persisted in northern Honshu, the Emishi. Finally, it is usually understood that the indigenous people of Hokkaido, the Ainu, have some relationship to the Jomon.
The Ainu are of interest because they do not exhibit many of the traditional characteristics of Northeast Asian people in their phenotype. Ergo, they are termed the “hairy Ainu.” Early anthropologists noting the physical uniqueness of the Ainu speculated that they were a lost branch of the white race. But even early ABO blood group analysis suggested that this was not the case. Rather, the closest relatives of the Ainu were the peoples of Northeast Eurasia. Nevertheless, it does seem striking that the Ainu do not exhibit the distinctive features of many other East Asians. I think this makes more sense though when you consider that it is likely that the East Asian physical type is not quite as primal as we might think. The latest work from genetics indicates very rapid population expansion out of the loci of agricultural activity on the North China plain over the past 10,000 years. The relatively uniformity of physical type and genetic relatedness across East Asia today may have more to due with demographic expansion that genetic connectedness through gene flow over long periods of time.
Much of this can be gleaned by inference, implication, and intuition. A new paper in Molecular Biology and Evolution uses explicit model testing to infer whether the modern Japanese the products of population replacement, admixture, or cultural diffusion. In other words, are the Japanese total cultural and biological descendants of the Yayoi. Are the Japanese total cultural and partial biological descendants of the Yayoi. Or, are the Japanese total cultural descendants of the Yayoi, but total biological descendants of the Jomon. The paper, Model-based verification of hypotheses on the origin of modern Japanese revisited by Bayesian inference based on genome-wide SNP data, reports that the middle hypothesis, genetic admixture, is highly supported when compared to the other two.
There are few interesting points. First, reference population matters. I’m rather sure that the Ainu samples they used as proxies for Jomon are imperfect because the Ainu today have recent Japanese admixture, and, as they note the Jomon themselves exhibit population structure (i.e., the Ainu are one lineage of post-Jomon people). This structure goes back to the late Pleistocene, over 10,000 years ago. The divergence between the ancestors of the Jomon and the Ainu goes back ~20,000 years ago, about when East Asian peoples began to diversify in the wake of the Last Glacial Maximum. The estimates for Jomon ancestry are probably inflated by the admixture in the Ainu, but dampened by the fact that the Ainu are not prefect proxies for the Jomon. Additionally, the reference population for the Yayoi, whether Koreans or North Chinese, has an effect.
But these are minor details. The major conclusion in a qualitative sense is that the modern Japanese are predominantly descended on the whole from the Yayoi farmers, but have a substantial minority component of indigenous Jomon ancestry. This almost certainly varies as a function of geography, the modern Japanese people likely can be used to obtain a “ghost phylogeography” of the Jomon whom they absorbed, as presumably admixture occurred locally a fair amount.