Vegetation history of Japan since the last glacial based on palynological data.
Nobuo Ooi（大井信夫）, p.1-101
To clarify the geographical history of Japanese vegetation and flora since the last glacial, the pollen assemblages were compiled into nine periodic groups 40 ka, 20 ka, 14 ka, 12 ka, 9 ka, 7 ka, 5 ka, 2.5 ka, and 0.8ka named by their median ages that approximately corresponded to Japanese pollen zones. Twenty-five major arboreal pollen types were selected for statistical analyses to map the distribution of pollen types and pollen assemblages. To clarify physiognomical changes in vegetation 25 pollen types were divided into eight physiognomical classes. Chronological succession of the physiognomical classes and pollen types apparently showed vegetation changes since the mid-glacial. During the mid-glacial (period 40 ka), temperate forests of Cryptomeria or deciduous broadleaved trees dominated in central to western Japan, and evergreen pinaceous forests did in northern Japan and the mountainous areas of central to western Japan. At the peak of the cold full-glacial (period20 ka), deciduous coniferous forests of Larix prevailed in Hokkaido, and evergreen pinaceous forests did in the other parts of Japan, with a considerable area of deciduous broadleaved forests in the lowland of central to western Japan. During the late-glacial (periods 14 ka, 12 ka), Betula forests expanded with evergreen pinaceous forests in the mountainous areas of central to western Japan and in northern Japan, and deciduous forests prevailed in the lower areas of central to western Japan with scattered Cryptomeria forests. During period 9 ka of the early post-glacial, deciduous broadleaved forests prevailed except in Hokkaido, where Betula and pinaceous forests dominated, and evergreen broadleaved forests appeared in southern Japan. At the hypsithermal (period7 ka), evergreen broadleaved forests expanded in southwestern Japan, and deciduous broadleaved forests did in northeastern Japan. These trends continued until period 2.5 ka. In the recent (period 0.8 ka), Pinus forestsexpanded around the lowland areas with an increase of human activities except in Hokkaido. The chronological succession of pollen types, however, showed little evidence of plant migration in spite of the above physiognomical changes in vegetation. Horizontal and vertical distribution of Fagus and Cryptomeria pollen types, for example, changed little chronologically, whereas their frequencies or relative abundance in pollen assemblages changed greatly. Similarly, the chronological succession of other pollen types showed little evidences of migration, but the falls and rises in the dominance of mother species. To clarify changes in plant communities, pollen assemblages were categorized into 20 groups based on the dominance of pollen types with empirical information. The pollen assemblage groups did not have any direct correlation with plant associations, but definitely had a close correlation with vegetation. The distribution of pollen assemblage groups through the nine periods did not simply shift in latitude or altitude, but expanded and shrunk. Thus, vegetation changes since the midglacial were not caused by plant migration, but by increases or decreases of existing species.