Forest types and the ecological processes to form them have been discussed mainly by simple eco-physiological properties of the dominant species. Recent studies on cool-temperate forest in Japan, however, have demonstrated the importance of other processes like plant-animal interaction, natural and human disturbance, and geological history. Reviewing these papers, I propose three main types of cool-temperate deciduous forests, commonly distributing in East Asia; Mixed Deciduous Forest, Beech Forest and Mixed Conifer and Deciduous Broadleaf Forest. Beech forest distributes in most humid areas. The intermediate temperate forest, which was considered as inter zonal between Beech forest and Warm Temperate (Evergreen) forest, should be included into the Mixed deciduous forest, which commonly distributes in areas with relatively dry, and continental climate. Having drier climate, the Mixed deciduous forest is affected by fire disturbance, and human land use in recent several thousand years, while Beech forest has disturbance regime only with tree fall gap. Species composition and disturbance regime of them are analogous to those in eastern North America.
We reviewed our recent studies on the cause of geographic difference of Siebold's beech (Fagus crenata) forest vegetation in Japan with special reference to the effect of snow-pack on tree regeneration.
(1) A direct gradient analysis of forest vegetation along snowfall gradient showed that the species diversity index declined clearly along the snow depth gradient. One third of the tree species was not distributed in the snow-rich region ( >2.0-m in snow depth). Other species were distributed from snow-poor region (< 2.0-m in snow depth) to snow-rich region. However, they lost their relative importance in snow-rich conditions exceptF. crenata, which was the only species that gained very high dominance in snow-rich region. (2) Regarding the snow pressure tolerance hypothesis, traits of beech forest species; i.e. trunk bend, trunk height decrease, trunk damage and basal sprouting rate were compared along the snow pressure gradient. General trends recognized are 1) trunk bend and trunk height decrease; 2) trunk injury increase; and 3) maximum size decrease. However,F. crenata was the only species whose trunk form and maximum size were hardly influenced by snow pressure.(3) The causes and timing of seed death in early regeneration of F. crenata was studied at 15 sites along a snowfall regime in countrywide scale. Seed production did not significantly differ between snow-rich regions and snow-poor regions. Pre-dispersal seed mortalities by insects were higher in snow-poor regions than those in snow-rich regions, but this only seemed to be a minor factor influencing the population. A large proportion of the viable seeds were killed in winter. Mortality in winter was much higher in snow-poor regions than that in snow-rich regions, and this factor was strongly correlated with the geographic variation of seedling regeneration probability. Germination experiments made clear the reason of high mortality of beech nuts in winter, and suggested that effects of desiccation, rodents, fungi and bacteria were very strong. (4) These results showed the close relationships between the snowfall regime and F. crenata regeneration, and are consistent with its distributional and environmental development since Last Glacial time.
In human inhabited areas, forest vegetation develops under various anthropogenic influences. In the Kitakami Mountain Range in Northern Honshu, Japan, present secondary stands clearly showed differing species compositions according to the type of past land-use at the site and could be classified into several types. The species constituting each stand type had some specific trends in their characteristics which appeared to be advantageous to their subsistence under the disturbance regimes imposed by a certain land-use. Two congeneric trees (Betula maximowhicziana and Betula platiphylla var. japonica) were typical examples: Segregation in distribution at landscape scale was closely related to interaction of their specific characteristics with past land-use at the site. Meanwhile, spatial distribution of each land-use type was determined by the site environment. Consequently, present secondary stands in the Kitakami Mountain Range should be understood as the combined result of direct influences by the site environment and indirect influences by the site environment through land-use or disturbance regimes. Diversity of anthropogenic disturbance regimes due to various site environments might to a certain extent be responsible for the disparity found between vegetation growing on the Japan Sea side and on the Pacific Ocean side of Northern Honshu.
Cryptomeria japonica is considered to be the primary component of Japanese vegetation history, evidenced by pollen analytical data for the period since the Last Interglacial Epoch. The other main components include some Pinaceae and deciduous oak, although beech forest and evergreen-leaved forest were not distributed in a wide area, when compared with the Post-Glacial Period. This information differs from the supposed vegetation during wide climatic movement based on a temperature change. Therefore it is necessary to recognize the regional different forest under the same climate. From pollen analyses on the Pacific Ocean side of the Tohoku district, the deciduous broad-leaved forest, including beech, was distributed across a wide area until 4000 yr. B.P., at which time fir trees as well as evergreen oaks increased in the southern part of the district up to 3000 yr. B.P. Although the distribution of beech forest shrank during that period, it expanded again until 1000 yr. B.P. After that time, the recent forest vegetation was consisted of evergreen and deciduous oak in the southern part, while mixed deciduous beech forest accompanied by Fagus japonica around the northern part. These various changes might be caused by a climatic change and human impact.
Early Last Glacial fossil wood assemblages were found in and below the Toya Pyroclastic Flow Deposits (Tpfl) around Lake Toya (Toya Caldera), Hokkaido. Among the assemblages, three fossil zones were recognized (in an ascending order): Zone I consisting of Picea, Larix and Abies, accompanied with Salix and Lonicera; Zone II of deciduous trees and shrubs such as Alnus, Salix, Rosa, and Lonicera; and Zone III of Betula and Picea. A boreal coniferous forest mixed with Salix and Lonicera was reconstructed from Zone I assemblage. Zone II reflected wetland vegetation. A preliminary pollen analysis of Zone II indicated that Picea, Abies, and Betula dominated boreal coniferous forests on adjacent upland of the fossil wood sites. Zone III also indicated boreal coniferous forest. Zone I - III were correlated to Oxygen Isotope Stages 5d - 5c based on the age of Toya Tephra and proposed pollen zones in Hokkaido. A good correlation of wood assemblages with known pollen assemblages indicates that the forest was composed mainly of taxa detected by pollen analyses. Drastic vegetation change from the Last Interglacial to Last Glacial reconstructed in Hokkaido may be attributed to its location near the cool-temperate forest/boreal coniferous forest boundary.