Tukasa Hukusima’s research while affiliated with The Open University of Japan and other places

In cool temperate forest zones in eastern Japan, certain forb species like Cimicifuga simplex in light⁻snow regions grow on slopes of beech forests, while in heavy⁻snow regions its habitat is rather limited to valleys adjacent to beech forests. This research study reveals reasons for this difference in forb species habitats. As key factors, the study focuses on differences in the lighting environment during the leafing season as well as the state of tree leaf litter sedimentation on the forest floors in both light⁻snow and heavy⁻snow environments. Regardless of the amount of snow, the volume of tree leaf litter sedimentation on the forest floor was large on slopes of beech forest and small in valleys. In heavy⁻snow beech forests, the number of layers of litter sedimentation as well as the density of litter compaction within a certain thickness of sedimentation were larger compared to light⁻snow regions. In terms of the lighting environment during the summer season, no difference was observed on slopes of beech forest between light⁻snow and heavy⁻snow environments. However, over the spring season in light⁻snow regions, the lighting environment of the forest floors was good during the leafing seasons, because foliation begins with the herbaceous layer and leaf expansion of canopy trees comes later. On the other hand, in heavy⁻snow regions, the lighting environment of the herbaceous layer was poor during the leafing season, since foliation starts from canopy trees as snow covering the floor remains longer. In valleys, foliation of herbaceous layer preceded that of shrubs even in heavy⁻snow regions. This phenomenon was attributed to the fact that lingering snow in the valley floor covers the shrubs alongside the valley, which in turn delays foliation of the shrubs. Two factors are considered reasons restraining the growth of forb species in beech forests in heavy⁻snow regions: 1） the large number of layers and the high density of litter sedimentation at the forest floor constrict the establishment and germination of forb species, and 2） the poor lighting environment from spring to early summer （which are supposed to be productive seasons） is due to lingering snow that makes the crown canopy of beech forests foliate before forb species. However, the study found that forb species are able to grow in valleys in heavy⁻snow regions, because the volume of litter sedimentation is small and herbaceous layer begins and completes foliation before shrubs do.

Within the wide-ranging beech forests of East Asia, differences in flora can be seen to reflect differences not only in the capacity of compose plant communities, depending from the climatic and ecological conditions, as it was described in the syntaxonomical section of this study, but also in the evolutionary history of each plant group and the migrations following the last glacial maximum in each area. For these reasons there are very few species that are uniformly diffused across the whole range of the study area. However, on the other hand, in many cases, even though species were different, similar structural properties and functional adaptations can be observed at the genus or family levels.

Speciation of beech (genus Fagus) in East Asia has been intensive and beech forests are widely, though disjunctly, distributed within Japan, Korea, China and Taiwan. Beech forests lie between the evergreen broad-leaved forests in the south (lower limit) and the evergreen coniferous trees in the north (upper limit), making them a key vegetation type in East Asia. In this study, to better distinguish the features of beech forests in East Asia, we compared species composition of beech forests based on the known relevés and on new relevés we carried out in the field. As a result of that comparison, we were able to divide the beech forests of East Asia into three classes. Classes I and II are characterized by the abundance of deciduous flora. These 2 classes contain 4 orders, 8 alliances, 20 associations and 48 subassociations. Class III is composed of evergreen broad-leaved forest with the presence of Fagus. There are few common species among the classes, especially in China. Because of that, it is hard to describe the similarities and differences of plant communities only considering the species level. For that reason, we further compared also the genus and family composition, which were considered to have larger distribution between the different vegetation types. East Asian beech forests are characterized by genera endemic to East Asia, as well as by genera common to beech forests of the northern hemisphere. Classes I and II show a large overlapping in genus composition. Within each order, there are many characteristic genera, a fact that confirms the independence of each order. In contrast, few characteristic genera are found at the alliance level, although the frequency of genera within each alliance varies. That is not the case, however, in the association and the subassociation levels. We also compared life form composition for each vegetation unit. We found the following: (1) beech forests in East Asia always contain scrub bamboo species; (2) classes I and II contain evergreen conifers of PEN and NPEN life forms. Class I contains proportionally more deciduous flora, classified as life forms of PD, NPD, H scap and H rosul. Class III, in contrast, contains proportionally more evergreen flora, classified as life forms of PE, NPE and G rhiz. Class II contains an intermediate proportion of deciduous and evergreen floras. At the order level, the frequency of life forms for shrub and herb species is characteristic. At the alliance level there is much less difference in life form composition within each alliance. At both the alliance level and the order level, when the proportion of life forms indicative of broad-leaved forests is higher, the proportion of herbaceous life forms is lower. There are no significant differences in life form composition within each association and subassociation; a comparison of life form compositions within the lower vegetation units would not, therefore, be useful. We also compared climatic conditions between the three classes. Our comparison shows that the mean annual temperature of class I is less than 10 °C, which is lower than the mean annual temperatures of classes II and III when adjusted for the same elevation of 1,000 m asl. Further, at an elevation of 1,500 m, most areas of class I are 6 °C or below, while most areas of class II are above 10 °C, and class III, a warmer environment still, is above 16 °C. The pattern of temperature difference corresponds to the distribution patterns of evergreen and deciduous flora. In terms of annual precipitation, there are great differences between the three classes. In addition, regional differences in precipitation are reflected in differences at the order level. Beech forests in Japan and Taiwan in general occur in areas with rainfall over 2,000 and 2,600 mm respectively; those in Korea and China mostly have less than 1,500 mm rainfall. Our comparison of climatic differences suggests, therefore, that class level differences are strongly related to differences in thermal conditions while order level differences are strongly related to differences in the quantity of precipitation.

Table 3.1 is a synoptic table of the forests dominated by the genus Fagus in East Asia. According to Table 3.1, the phytosociological system of beech forests in East Asia was classified as in Table 3.9 and the distribution map of the classified classes, orders, alliances and associations was depicted as Fig. 2.1. In Table 3.1, 68 species categorized as the species group 49 are the common species occurring in the beech forests in East Asia, although there is a regional bias. Many of the 68 species are character species of beech forests in each region.

This book describes the mountain forests of East Asia (Korea, Japan, China and Taiwan), the tree layers of which contain different species of the genus Fagus. The vegetation is primarily deciduous in the northern regions, whereas in South China evergreen trees can also be found: a total of 21 plant communities are described, with data on species composition, dominance, geographical distribution and ecology. A general comparison is provided by synoptic Table 1, which details the frequencies of ca. 1500 species growing in the Fagus forests; biodiversity and evolution are discussed. The book, which is the fruit of a major international collaboration, presents a synthesis of extended original investigations by the authors and hardly accessible specialist literature.

From 1965 to 2007, we researched the dynamics of the pioneer tree Idesia polycarpa, which shows contrasting behavior in natural and urban forests, and discussed it in relation to the transitions of the preserved urban forest in the garden of the Institute for Nature Study, Tokyo. The number of I. polycarpa individuals increased rapidly between 1965 and 1983, leveled off between 1983 and 2002, and decreased from 2002 to 2007. Nevertheless, the basal area of I. polycarpa increased continuously from 1965 to 2007. The diameter at breast height (DBH) class distribution of I. polycarpa showed a single peak from 1965 to 1992, while at least two indistinct peaks appeared in and after 1997. I. polycarpa that reaches the canopy can survive by avoiding shade. In addition, I. polycarpa is thought to maintain growth by expanding the crown when a peripheral individual dies. It is suggested that urban forests repeatedly undergo the periods to be left undisturbed or protected, and those to be harshly disturbed, allowing I. polycarpa to regenerate and grow well.

We developed statistical distribution models for two evergreen Quercus species (Q. acuta and Q. salicina) by using classification tree analyses with phytosociological releve data, in Japan. Two different models were constructed for each of the two species: (1) species at least present (SP-model) and (2) species dominant in the canopy layer (DO-model). Four climate variables: warmth index (WI), temperature of the coldest month (TMC), summer precipitation (PRS) and winter precipitation (PRW), were used as predictor variables. WI and TMC were the most influential factors for each of the two species. The lower temperature range of each species was generally coincided with the averaged minimum temperature of the coldest month warmer than or equal to -1℃ The DO-models for the two species were mainly contributed by WI, and the predicted optimal habitats were narrower than those of the SP-models. The predicted distributions for the two species were similar to the geographical patterns of PRS. Specifically, Q. acuta was mainly predicted to occur under high precipitation conditions, such as on the western part of the Kanto region on the Pacific Ocean side. On the other hand, Q. salicina was predicted to occur both on the Pacific Ocean side and the Sea of Japan side. PRW was influential factors for the DO-models. Q. acuta was predicted to occur under low PRW conditions.

On the floodplains of gravel-bed rivers in Central Honshu, we can often observe deciduous broad-leaved forests dominated by Quercus serrata. In order to clarify ecological characteristics of these Quercus serrata forests, we surveyed species composition and substratum of the ground surface on the floodplains of the Sabi, Daiya, and Komu river. The substrata were sandy alluvial deposits in all stands, showing those Quercus serrata forests were established after river flooding and accumulation of sandy deposits. Riparian Quercus serrata forest stands had different species composition compared with coppice forest stands around each river. Riparian stands included elements of ravine forest dominated by Zelkova serrata, whereas they were lacking Ericaceae species and tall herb species that were common in coppice forest. WI are about 85℃・month and CI are less than -10℃・month respectively in the studied areas. This means that the areas are located in intermediate warm-temperate zone. Therefore, we concluded that the Quercus serrata forest established on flood-plain was regarded as a type of riparian forests of gravel-bed rivers in low elevation area of the Central Japan.

A new association of deciduous broad-leaved riparian forests, namely, Kerrio japonicae-Quercetum serratae is described on the Pacific Ocean side in central Honshu, Japan. This association is dominated by Quercus serrata and is accompanied with Zelkova serrata, Castanea crenata, Prunus spp., Acer spp., and Carpinus spp. in the tree layer. Comparing with the other associations of Quercus serrata, Kerrio japonicae-Quercetum serratae is characterized by the elements of Fraxino-Ulmetalia Suz.-Tok. 1967, such as Kerria japonica, Rhamnus japonica var. decipiens, Ostrya japonica, Acer mono var. ambiguum, and Lilium cordatum. The distribution of this association is inland area of central Honshu of the intermediate temperate zone between warm temperate zone and cool temperate zone. The habitats of this association are riverside stable terraces in the middle reaches with surface deposits consisting of cobbles, gravels, and coarse sands. This association can be classified into the following upper unit: Carpino-Quercion serratae Miyawaki et al. 1971, Quercetalia serrato-grosseserratae Miyawaki et al. 1971, Querco serrato-Pinetea densiflorae Nakanishi, Takeda et Hattori 1977. The Kerrio japonicae-Quercetum serratae can be regarded as one of the edaphic climax of alluvial plain.

Acorns (nuts and cupules) of 88 Quercus serrata Thunb. individuals in central Japan were collected in 2000 and 2001, and the relationships between twelve acorn traits and several environmental factors, such as altitude, warmth index (WI), annual mean temperature and precipitation were examined. All nut traits, three cupule traits and a peduncle trait were correlated negatively with altitude and precipitation during the acorn growing season, and positively with WI and annual mean temperature. These results indicated that the higher the precipitation was and the lower the WI, the smaller were the size and volume of nuts and cupules. This suggests that the differences in the length of the acorn growing season among the sampling sites may be responsible for the trends in acorn size and volume. The acorn size traits of Q. serrata were negatively correlated with precipitation during the acorn growing season, especially within the latter part of the growing season. Thus, we conclude that acorn size of Q. serrata had some strong relationship to less precipitation in the latter part of the acorn growing season. However, it was difficult to separate the effects of precipitation from those of temperature on acorn traits in our analysis. Moreover, Q. serrata nut length was more dependent on thermal conditions than was nut width. This difference may be caused by the different developmental patterns between these two dimensions, especially after August when the width stopped increasing but the length did not.