13

Climate Change Impacts on Vegetation in Humid Asian Mountains

Masahiko OHSAWA

Institute of Environmental Studies

Graduate School of Frontier Sciences The University of Tokyo

5-1-5 Kashiwanoha, Kashiwa-shi, Chiba 277-8653, Japan e-mail: mohsawa@k.u-tokyo.ac.jp

Abstract Diverse landscapes or ecosystems of mountainous terrain pose a challenge to predicting the re-

sponse of mountain ecosystems to potential global change impacts. Isolated high mountain areas are intrinsically fragile habitat-islands surrounded by lowland vegetation that is impassable to highland biota. In humid East Asia, the more or less continuous extension of high mountain chains exhibits a latitudinal transition of vertical arrangement of forest zones on mountains. Since Humboldt, differ-ences between vegetation zonation in humid tropical and temperate mountains have been described, but the causal temperature factors responsible for the patterns have not been fully explored. This has made it difficult to evaluate the impact of climatic change on mountain vegetation. In temperate mountains, latitudinal transition of vegetation is controlled by increasing temperature seasonality or minimum winter temperature. In tropical mountains, a year-round decrease in temperature with no seasonality causes a shortage of energy for forest growth as altitude increases. By distinguishing between these two temperature factors, most of the structural diversification in terms of life-forms, biomass, stratifica-tion, and diversity, among tropical and temperate mountain-forests can be explained. Furthermore, the critical temperature factors controlling the patterns can be used for elucidating potential global- warming-induced changes in forest zones and the causal effects, particularly at the interface between tropical and temperate vegetation in the mid-latitudes.

Key words: evergreen forest, forest limit, temperature sum, tropical mountain, winter temperature

1. Introduction

Since the time of A. von Humboldt, many

researchers have conducted comparative studies on mountain vegetation zonation. Humboldt (1817) devised isotherms and tried to compare the altitudinal zones of mountain systems from a global perspective (Troll, 1962). Around the middle latitudes, where tropical zones transition into temperate zones, humid forest series exist only in South to East Asia where the monsoon climate provides enough rainfall throughout the year. On the Asian continent, many botanists/ geographers have explored mountainous areas like the Himalayas, upper Burma, the Hengtuan Mountains, southwestern China, etc. Among them, several re-searchers, e.g., Troup (1926), von Wissman (1939), Troll (1959), Schweinfurth (1957), Wolfe (1979), etc., proposed mountain-vegetation profiles as the basis for comparative study of diverse mountain worlds.

During the last several decades, we have been con-ducting comparative studies on mountain-vegetation zonation in South (the Himalayas), Southeast, and

East Asian humid mountains to compare the general pattern of forest zonation in a latitude-altitude matrix (Ohsawa et al., 1985; Ohsawa, 1990, 1995a; Ohsawa & Nitta, 1997a, b; 2002). Recently our comparisons were extended toward the Canaries where a relic type of laurel forest developed as the cloud forest sustained by fog drip provided by trade-wind inversion clouds (Ohsawa et al., 1999), and also toward Georgia, where the evergreen trees of the Colchic Tertiary relic vege-tation, are also sustained by fog/rain from the Black sea (Kikvidze & Ohsawa, 1999, 2001). These subtropical/warm-temperate evergreen broad-leaved forests are interesting from the standpoint of under-standing the processes of adaptation, diversification, and evolution of the original tropical montane rain forest formations of Southeast Asia, which are facing increasing stresses in the mid-latitudes, i.e. droughts in the west toward the Mediterranean and low winter temperatures due to seasonality to the east in monsoon Asia (Ohsawa et al., 1986; Kikividze & Ohsawa, 2001; Ohsawa & Nitta, 2002).

Global change issues are particularly serious from

Global Environmental Research ©2006 AIRIES 10(1)/2006: 13-20 printed in Japan

14 Masahiko OHSAWA

the standpoint of conservation of mountain ecosys-tems because of their environmental peculiarities. Mountains are often island-like isolated habitats among lowland ecosystems or human-dominated land-scapes, and it makes it difficult for the mountain biota to migrate freely. Mountain slopes have a steep tem-perature gradient, nearly 600 to 1,000 times steeper than that of horizontal zonation. In general, the tem-perature changes at a rate of 6°C/km on mountain slopes but 0.01-0.006°C/km of horizontal distance in the mid-latitudes (Ohsawa & Nitta, 1997a). This suggests tight species interactions in mountain envi-ronments compared to those of horizontal, latitudinal changes. At the same time, however, this facilitates migration of certain mountain biota towards neighbor-ing ecosystems or life-zones; e.g., in the lowlands of the humid eastern Himalaya at ca. 500 m a.s.l., we can observe tropical tree species, e.g., Terminalia myriocarpa (Combretaceae) coexisting with such a temperate tree species as Betula alnoides (Betulaceae) in the same community (Ohsawa, 1987a). These as-pects of mountain environment act as the migratory corridor for a certain biota, particularly in temperate mountains with the seasonal overlapping of tempera-ture conditions between different altitudes (cf. Janzen, 1967). Along with this, the complex habitat mosaics of mountain topography provides the ‘hiding places (refugia)’ for a certain biota from competitively supe-rior species (Ohsawa, 1987b).

The object of the present study is to explore the forest zonation patterns in humid Asian mountains and analyze the factors controlling the zonation in order to prospect possible impacts of global warming on mountain ecosystems.

2. Mountain Vegetation Zonation in Humid

Asia and Its Controlling Factors We have compared the latitudinal-altitudinal diver-

sification of vegetation zones on the basis of several descriptive studies on mountain vegetation zonation at different latitudes. On tropical mountains such as Mt. Kerinci (2°S, Ohsawa et al., 1985) and Mt. Kinabalu (5.6°N, Kitayama, 1991; Ohsawa, unpublished data), physiognomically all the forest zones rather monoto-nously consist of broad-leaved evergreen forests up to the forest limit. Even though tropical mountain for-ests have such physiognomic uniformity, there are drastic floristic and structural changes including spe-cies diversity that can generally be delineated into lowland and lower and upper montane zones (Van Steenis, 1972; Whitmore, 1984; Ohsawa, 1991; Ashton, 2003). Typical examples of humid tropical to temperate forest zonation are obtained from slopes surrounding the mountains of High Asia; the southern slopes in the Himalayas (Numata, 1983; Ohsawa, 1987a; Wangda & Ohsawa, 2006), the eastern slopes in Sichuan (e.g. Mt. Emei, Tang & Ohsawa, 1997), and some mountainous islands off the Asian continent,

i.e. Taiwan (Su, 1984) and Japan (Ohsawa, 1984a, b). In contrast to tropical mountains, temperate mountains have clear zonal transitions of physiognomic forest zones comprising highly diverse tree life-forms (ever-green, deciduous, and coniferous forests), though structural (tree height, biomass, etc.) and species- diversity traits showed no remarkable changes with altitude except around the forest limit (Ohsawa, 1995a). All of these data provide us invaluable infor-mation on forest zonation in tropical to subtropical and warm-temperate mountains in humid monsoon Asia (Fig. 1).

A more or less sharp transition of vegetation zona-tion from humid tropical to temperate mountains can be seen at ca. 20°N-30°N in the mountain-vegetation template for South/East Asia (Ohsawa, 1990, 1995a). The template depicts the pattern of altitudinal zonation on humid mountains along latitudinal gradients (Fig. 2). According to a comparative study of zonation patterns and climatic conditions of these mountains, the main reason for the gradual transition from tropical, to subtropical, to temperate types of zona-tions is the increasing seasonality or steep decline in winter temperatures with no appreciable depression of summer temperatures toward high latitudes (Ohsawa, 1990, 1995b; Ohsawa & Nitta, 2002). The transition zone between tropical and temperate mountain zonation also coincides with the planetary frontal zone

F

ig. 1 Altitudinal formation zones in typical high areas ofsouth/southeast/east Asian mountains overlaid on amountain vegetation template for humid Asia (Ohsawa,1990, 1993). Data sources: Mt. Kerinci, Sumatra (Ohsawa et al.,1985), Mt. Kinabalu, Borneo (Kitayama, 1991; Ohsawa,unpublished data), Mt. Yushan (Su, 1984), Mt. Namshila(Ohsawa, 1987a), Mt. Emei (Tang & Ohsawa, 1997, Mt.Miyanoura (Ohsawa, 1984a), Mt. Kirishima, Mt. Fuji(Ohsawa, 1984b), and Mt. Poroshiri (Ohsawa et al.,1973).

Climate Change Impacts on Vegetation in Humid Asian Mountains 15

starting around 20°N-30°N (Weischet, 1988). The northward extension of evergreen broad- leaved trees of tropical origin is blocked by a minimum monthly mean temperature of –1°C (Iversen, 1944; Ohsawa, 1990) around 42°N in East Asia. Beyond this limit, deciduous and coniferous forests replace the evergreen. In turn deciduous broad-leaved trees of cool- temperate zone are blocked by an extreme minimum temperature of – 40°C further north. Both deciduous and coniferous trees in the high latitudes of the northern hemisphere have adapted to increasing seasonality of temperatures, particularly to cold winters by shedding plant parts or evolving special means of tolerating freezing temperatures (Sakai, 1982). Over geological time scales, deciduous broad-leaved trees are thought to have evolved to suit an increasing annual temperature range of more than 20°C (cf. Wolfe, 1979), and these temperature condi-tions start to appear around 30°N in maritime East Asia at present (Ohsawa, 1990).

3. Thermospheres and Corresponding Forest Zones in Mountain Environments On the latitude-altitude matrix of temperature con-

ditions for all of humid South/East Asia, there are sev-eral thermospheres that are lacking in the other latitu-dinal ranges. The southernmost or northernmost thermospheres have reached their distributional limit in either direction. For example, the thermosphere for tropical upper montane forests (TUMF) is found only in high mountains at latitudes lower than ca. 30°N (Fig. 2, cf. Ohsawa, 1991). The temperature conditions delineating the TUMF are a tem-perature-sum (Warmth Index, by Kira, 1948) of 85°C·months for the lower altitudinal limit (at ca. 2,500 m a.s.l. in the tropics), and 15°C·months for the upper limit, which coincides with the forest limit on tropical mountains (ca. 3,600-4,000 m a.s.l.). On tropical mountains, a monthly mean temperature of the coldest month (CMT) of –1°C, which limits ever-green broad-leaved forests on temperate mountains, is

Fig. 2 Mountain vegetation template with effective boundary temperature factors overlaid on a

mountain profile of humid Asia. Tropical zonations are controlled by the temperature sum (WI) and temperate zonations, by low temperature, except at the forest limit.

16 Masahiko OHSAWA

located far above the forest limit at ca. 4,500 m (Fig. 2), and the coldest monthly mean temperature at the forest limit is ca. 5-6°C throughout the year. Evergreen broad-leaved trees with microphylls in the TUMF, therefore, reach their energy limit for making tall forests at the forest limit before reaching their physiological tolerance limit for low temperatures. The major evergreen trees of the TUMF zone are the ones mainly constituting the understorey layer of the zones below that (lower montane and lowland forests). The characteristic tree groups are mainly Symploca-ceae, Myrsinaceae, Theaceae, Aquifoliaceae, and other ‘symploco-myrsinaceous’ trees, with charac-teristic features such as microphylls, hypsophyllary

buds, and sylleptic branchings (Ohsawa, 1991; Ohsawa & Nitta, 1997b, 2002).

Outside the tropics, the isotherm of CMT = –1°C (the lower temperature limit for evergreen forests) is located at gradually lower altitudes with increase in latitude. The line of CMT = –1°C crosses the line of WI=15°C·months (forest limit) at ca. 20°N, where the temperature conditions stop evergreen forests at certain altitudes far below the forest limit (Fig. 2). The isotherm of CMT = –1°C finally reaches sea level at around 40°N.

Figure 3 illustrates the diversification of the ther-mosphere in South/East Asia, mainly focusing on tropical broad-leaved evergreen forests with some

Fig. 3 Effective temperature factors for the differentiation of forest zones in tropical mountains and the northern

limit areas of evergreen broad-leaved forests around central Japan (based on the data and original diagram of Ohsawa, 1990). Three thermal conditions are proposed as the limiting conditions for evergreen broad-leaved trees, as follows:

▽: altitude at which CMT = –1°C, ○: altitude at which CI = –10°C·months, ●: altitude at which WI = 85°C·months. Temperature sum conditions, i.e., WI = 15°C·months for the forest limit and WI = 240°C·months for

tropical lowland forests, are indicated only by lines without dots. CASE 1 and CASE 2 indicate two different scenarios predicting rising temperatures only in winter and

summer, respectively. The range of the rises is not proportional, and only the differences in response pattern for each quadrant are illustrated.

Climate Change Impacts on Vegetation in Humid Asian Mountains 17 adjoining thermospheres in the vicinity of their north-ern limits.

Near the northern limit of broad-leaved evergreen forests at around 30°N-40°N (around central Japan), the altitudes of CMT = –1°C nearly overlap with the altitudes of WI = 85°C·months (Fig. 3, cf. Ohsawa, 1990). From a wider geographical perspective, the regression curves of the two factors (altitude for CMT = –1°C and WI=85°C·months) crossing at around 30°N-40°N, diverge clearly into two toward the equatorial tropics as shown in Fig. 3.

Although two boundary-temperature conditions for evergreen broad-leaved trees are located at around the same altitudes at 30°N-40°N, the biological significance of the two temperature factors differ. A CMT of –1°C is effective at limiting evergreen trees physiologically in temperate latitudes (cf. Iversen, 1944; Ohsawa, 1990), even though the temperature sum factor in summer is still sufficient for establishing tall forests. On the other hand a temperature sum of 85°C·months may not be a simple factor, but if the condition is extended toward tropical mountains, it coincides with the boundary between lower and upper montane zones at ca. 2,500 m a.s.l. (Ohsawa, 1990). The boundary also approximately coincides with the upper limit of notophyllous lauro-fagaceous forests, and above it, canopy dominants are replaced by sym-ploco-myrsinaceous trees which are evergreen micro-phylls.

Near the northern limit of evergreen forests, the two temperature conditions of tropical and temperate zonation can affect the zonal differentiation. Thus winter temperatures of CMT = –1°C and a summer temperature sum of WI = 85°C·months can overlap and divide the region into four quadrants (Q-A, B, C and D) as shown in Fig. 3. Around central Japan the crossings of the two factors and the resultant thermos-pheric quadrants (within the square in the schematic thermosphere on the latitude-altitude matrix in Fig. 3) can be correlated precisely with the extant forest types around the northern limit of evergreen broad-leaved forests in temperate mountains as follows (Fig. 3):

Quadrant A (thermosphere > WI 85°C·months and >CMT –1°C), a warm-temperate forest zone dominated by evergreen notophyllous lauro-fagaceous trees

Quadrant B (CMT –1 > thermosphere > WI 85), a mid-temperate forest zone dominated by mid- temperate deciduous trees of the genera Quercus, Carpinus, etc., and temperate conifers of Abies, Tsuga, Cryptomeria, etc.

Quadrant C (CMT –1 < thermosphere < WI 85), evergreen microphyllous forests dominated by symploco-myrsinaceous trees (tropical upper montane forest types)

Quadrant D (thermosphere < CMT –1 and <WI 85), cool-temperate deciduous forests dominated by Fagus.

A detailed analysis of the two boundary factors

clearly indicates a diversification of forest types at the contact zone between evergreen broad-leaved forests of tropical origin and deciduous forests of temperate origin. The pattern also indicates a very limited ther-mosphere for temperate deciduous and/or relict conifers (Q-B) in between evergreen and deciduous forest zones in East Asia (cf. Fig. 3).

It also clearly indicates that the major determinant of the formation type (tree life-form) at the tropi-cal/temperate interface is winter coldness (CMT –1°C), and evergreen broad-leaved trees are limited their northward extension by CMT –1°C. On the other hand the energy level indicated by the WI-value is effective for differentiating between notophyll ever-greens (Q-A) and microphyll evergreens (Q-C), mainly in tropical mountains (Fig. 3).

4. Global Warming Impacts on Mountain

Vegetation According to the present analyses, temperature fac-

tors that control zonal differentiation of forests differ between tropical and temperate mountains. While in tropical mountains, forest zones are differentiated through temperature sum conditions, winter low tem-peratures below freezing control the northern limit of evergreen broad-leaved forests in temperate moun-tains (Figs. 2 & 3). This brings disparity to their re-sponses to global warming.

Around the mid-latitudes, there are two different predictions for temperature changes due to global warming. One is that winter temperatures will rise with no appreciable change in summer temperatures, or at least that the rise in winter temperatures will sur-pass that of summer (Yoshino, 2006). The other is that temperatures will rise with no appreciable change in the seasonal pattern of temperatures (Nakagawa et al., 1998).

If global warming makes winter temperatures rise to a certain threshold, e.g., CMT above –1°C, the broad-leaved evergreen trees will be affected directly and start to move northward beyond their present northern limit, perhaps replacing the deciduous/ coniferous forests. The tropical/temperate interface affected by the two temperature conditions CMT and WI could produce the four quadrants shown in Fig. 3.

The two wedge-shaped thermospheres (Q-B and C) both represent fragile vegetation types vulnerable to temperature changes. Thermosphere Q-B (85°C· months < thermosphere < CMT –1) will be lost if winter temperatures rise (CMT > –1) due to climate change, as in Case 1 of Fig. 3. The forest type which has developed in thermosphere Q-B is the so-called mid-temperate zone forest mainly dominated by de-ciduous broad-leaved trees, e.g., Quercus, Carpinus, etc., and more importantly Tertiary relic temperate conifers such as Cryptomeria, Tsuga, Abies, Chamaecyparis and Pseudolarix, which are all valu-able timber trees in Japan. Ginkgo, Cryptomeria,

18 Masahiko OHSAWA Taiwania and Chamaecyparis are also vicariant genera in China and Taiwan.

Replacing these lost trees will be those of the newly expanding thermosphere resulting from rising winter temperatures, Q-C (CMT –1 < thermosphere < WI 85°C·months) (Fig. 3, Case 1). Quadrant-C is equivalent to the thermosphere of tropical upper montane forests, but it does not exist extensively in temperate mountains under the present climatic condi-tions. As exemplified in tropical mountain zonation, the evergreen forests which develop under lower en-ergy (temperature sum < 85°C·months) levels are symploco-myrsinaceous, evergreen, microphyllous forests, as in the tropical upper montane zone.

This vegetation change scenario is interesting be-cause there are several cases of fossil evidence (huge amounts of the remains of seeds and fruits) of the oc-currence of true Symplocos forests in Central and Western Europe which was at the northern limit of tropical broad-leaved evergreen forests during the Tertiary. Those forests became extinct later during the Quaternary glaciations (Nooteboom, 1975). Thus in the geological past there may have been forest types similar to the symploco-myrsinaceous forests of the present tropical upper montane forests, near the past northern limits of the broad-leaved evergreen forests at ca. 50°N. Even in the present situation at the northern limit of broad-leaved evergreen forests in temperate Japan, broad-leaved evergreen tree species with microphylls and hypsophyllary buds (Ilex crenata, Illicium religiosum, Daphniphyllum macropodum, etc.) are invading into the understorey of cool-temperate deciduous forests dominated by Fagus, Quercus, Acer, Carpinus, etc. Those broad- leaved evergreen trees of the understory are mostly vicarious species with the trees of the upper montane zone on tropical mountains (Ohsawa et al., 1985; Ohsawa, 1991; Ohsawa & Nitta, 1997b).

On the other hand in the case summer temperatures rise with no change in winter temperatures, thermo-sphere Q-C will be wiped out and Q-B will expand drastically (Fig. 3, Case 2). Moreover in this case the forest limit may also shift upwards as shown in Fig. 3, Case 2.

5. Conservation of Mountain Ecosystems

The overall temperature rise may shift all the zones

in a parallel way, and this would have a particularly serious effects on small peaks having a narrow cap of distinctive ecosystems at the top. Those isolated patches may surely be driven off of the peak. Moun-tains having complex topographies, however, may somehow harbor the relict vegetation. Several iso-lated, scattered Pinus parviflora communities (nor-mally distributed above ca. 1,500 m a.s.l.), thrive on rocky ridges of less than 500 m a.s.l. on the Boso Peninsula, Chiba Pref., in central Japan. They are a good example of such relic communities from the gla-

cial periods (Numata, 1979). These pine stands are now severely threatened in the area because of their isolated small populations being subjected to overall temperature increased, infectious wood nematodes, and other threats.

In a different case, in which only winter tempera-tures rise due to global warming with no change in summer temperatures, the situation would be different at the northern limits of broad-leaved evergreen for-ests. Only the broad-leaved evergreen forests and/or deciduous forests would shift upwards with no associ-ated movement of the forest limit (controlled by the temperature sum), This could suppress the subalpine forest zone below the forest limit, and the cool- temperate deciduous broad-leaved forests (dominated by Fagus crenata, Quercus mongolica var. grosseserrata, etc.) in between the subalpine conifer-ous and broad-leaved evergreen forest zones.

As already mentioned above, East Asia is one of the major refugia for Tertiary relic plants, particularly around the interface between evergreen and deciduous forests, ranging from warm-temperate to cool-tem-perate zones, and those relics can hardly compete with modern trees appearing in most cases as the dominant species in both cool and warm-temperate forests (Bond, 1989; Ohsawa, 1987). Most of those relic plants are inferior in competition and grow mainly in disturbed sites as fugitive pioneers, or as isolated patchy communities in harsh environments, as in the case of the relict pine stands on the Boso Peninsula. There are many similar examples in China, e.g., mono-typic genera such as Ginkgo, Tetracentron, Euptelea, Davidia, Pseudolarix, Cathaya, etc., growing in heav-ily disturbed sites, on rocky ridges, or at any other sites where modern trees are excluded by stressful habitat conditions (cf. Bond, 1989). They can some-times invade into even mesic, fertile sites as pioneers, and can survive by becoming tall emergent trees (Ohsawa, 1987b), but their regeneration is dependent on the disturbance or topographical complexities on a regional scale.

The only means of conserving those relict or other diverse plant communities in mountainous areas is to protect areas large enough to encompass their diverse habitats including specific regeneration sites (rocky ridges, cliffs, flood plains, etc., as well as mesic sites) as a whole. If the areas are insufficient in scale to include all the possible habitat types or the necessary population sizes to survive, those protected areas should at least be connected to each other by an eco-logical corridor that facilitates migration of biota (Ohsawa et al., 1998). In Bhutan, for example, there are nine protected areas which are inter-connected with each other by such ecological corridors to facilitate conservation of rare wildlife such as tigers, takin, and snow leopard (Dorjee, 2000). In total, the nine protected areas cover ca. 26% of the land area, and the conservation corridors contain in total ca. 9.2% of the land area in Bhutan.

Climate Change Impacts on Vegetation in Humid Asian Mountains 19 6. Conclusion

Humid mountains in South, Southeast and East

Asia provide the best sites for investigating zonal dif-ferentiation of forests along temperature gradients on a continuous latitude-altitude matrix. Such a series of vegetational gradation mainly governed by tem-perature cannot be found which are not subject to dry spells in other parts of the world. Analyzing the se-ries of mountain forests rigorously, we could identify two contrasting mountain zonation types, i.e. a tropi-cal zonation type at low latitudes below 20°N-30°N, and a temperate zonation type at higher latitudes. The two zonation types are controlled by two contrast-ing temperature factors, i.e. temperature sum for tropi-cal zonation and minimum winter temperature for temperate zonation. Only the forest limits of both tropical and temperate mountains are commonly con-trolled by the temperature sum. These contrasting temperature factors cause different responses of zonal vegetation to global warming. The shifting of vegeta-tion on mountainous terrain will have serious impacts and the best possible way to conserve mountain vege-tation diversity is to connect the isolated mountain environments with biological corridors that makes it possible for the biota to migrate to survive serious global change impacts.

References Ashton, P.S. (2003) Floristic zonation of tree communities on

wet tropical mountains revisited. Perspectives in Plant Ecology, Evolution and Systematics 6: 87-104.

Bond, W.J. (1989) The tortoise and the hare: ecology of angiosperm dominance and gymnosperm persistence. Biol. J. Linnean Soc. 36: 227-249.

Dorjee, K. (2000) Biodiversity Assessment and Conservation Planning. Bhutan. WWF Nepal Program.

Humboldt, A. von (1817) Des libnes isothermes et de la distri-bution de la chaleur sur le globe. Mem phys. et Chem. de la Societe d’ Arcueil, III, 462-602. (cited by Troll, 1962)

Iversen, I. (1944) Viscum, Hedera and Ilex as climate indicator. A contribution to the study of the Post-Glacial Temperature climate. Geol. Foren．Fuehandl., 66: 463-483.

Janzen, D.H. (1967) Why mountain passes are higher in the tropics. The American Naturalist 101:233-249.

Kikvidze, Z. and M.Ohsawa (1999) Adjara: east Mediterranean refuge of Tertiary vegetation. In: M.Ohsawa, W. Wildpret and M. del Arco, eds., Anaga Cloud Forest, Chiba University, pp. 297-315.

Kikvidze, Z. and M.Ohsawa (2001) Richness of Colchic vege-tation: comparison between refugia of south-western and East Asia. BMC Ecology, 1.

Kira, T. (1948) On the altitudinal arrangement of climatic zones in Japan. Kanti-Nougaku, 2:143-173.

Kitayama, K. (1991) Vegetation of Mount Kinabalu Park, Sabah, Malaysia. East-West Center, Honolulu.

Kitayama, K. (1992) An altitudinal transect study of the vegeta-tion on Mount Kinabalu, Borneo. Vegetatio, 102:149-171.

Nakagawa, S., M. Sugi, T. Motoi and S. Yukimoto (1998) Climate change projections. In: S. Nishioka and H. Harasawa, eds., Global Warming. The Potential Impact on Japan. Springer, pp. 1-33.

Nooteboom, H.P. (1975) Revision of the Symplocaceae of the Old World. New Caledonia excepted. Leiden University Press.

Numata, M. (1979) Eco-geographical consideration of floristic and vegetational changes in the Quaternary-Particularly re-lated to the Boso Peninsula. Quaternary Research, 17: 223-233. (in Japanese)

Numata, M. (1983) Biota and Ecology of Eastern Nepal. Chiba University.

Ohsawa, M. (1984a) Vegetation structure and dynamics in the Yaku-shima Wilderness Area, south Japan. In: Conservation Reports of the Yaku-shima Wilderness Area, Kyushu, Japan. Environment Agency of Japan, pp. 317-351.

Ohsawa, M. (1984b) Differentiation of vegetation zones and species strategies in the subalpine region of Mt. Fuji. Vegetatio, 57: 15-52.

Ohsawa, M. (1987a) Vegetation zones in the Bhutan Himalaya. In: M. Ohsawa, ed., Life Zone Ecology of the Bhutan Himalaya. Chiba University, pp. 1-71.

Ohsawa, M. (1987b) Habitat differentiation and the ecological niche in vegetation. Physiol. Ecol. Japan 24 (Special No.): s15-s27.

Ohsawa, M. (1990) An interpretation of latitudinal pattern of forest limits in south and east Asian mountains. Journal of Ecology, 78: 326-339.

Ohsawa, M. (1991) Structural comparison of tropical montane rain forests along latitudinal and altitudinal gradients in south and east Asia. Vegetatio, 97: 1-10.

Ohsawa, M. (1993) Latitudinal pattern of mountain vegetation zonation in southern and eastern Asia. Journal of Vegetation Science, 4: 13-18.

Ohsawa, M. (1995a) Latitudinal comparison of altitudinal changes in forest structure, leaf-type, and species richness in humid monsoon Asia. Vegetatio, 121: 3-10.

Ohsawa, M. (1995b) The montane cloud forest and its grada-tional changes in Southeast Asia. In: L.S. Hamilton et al., eds., Tropical Montane Cloud Forests, Springer-Verlag, pp. 254-265.

Ohsawa, M., T. Eguchi, N. Tanaka, H. Ikeda, M. Nemoto and Y. Hada (1998) Impacts on natural ecosystems. In: S. Nishioka and H. Harasawa, eds., Global Warming. The Potential Impact on Japan. Springer, pp. 35-99.

Ohsawa, M., P.H.J. Nainggolan, N. Tanaka and C. Anwar (1985) Altitudinal zonation of forest vegetation on Mount Kerinci, Sumatra: with comparisons to zonation in the tem-perate region of East Asia. Journal of Tropical Ecology, 1:193-216.

Ohsawa, M. and I. Nitta (1997a) Northern distributional limit of evergreen trees and its conservation implications in a changing mountain environment in East Asia. Proceedings of the Second Conference on National Parks and Protected Areas of East Asia, Japanese Organizing Committee for the Second Conference on National Parks and Protected Areas of East Asia. National Parks Association of Japan, pp. 273-285.

Ohsawa, M. and I. Nitta (1997b) Patterning of subtropical/ warm-temperate evergreen broad-leaved forests in east Asian mountains with special reference to shoot phenology. Tropics, 6: 317-334.

Ohsawa, M. and I. Nitta. (2002) Forest zonation and morphological tree-traits along latitudinal and altitudinal environmental gradients in humid monsoon Asia. Global Environmental Research, 6: 41-52.

Ohsawa, M., P.R. Shakya and M. Numata (1986) Distribution and succession of west Himalayan forest types in the eastern part of Nepal Himalaya. Mountain Research and Develoment, 6: 143-157.

Ohsawa, M., T. Tezuka and M. Numata (1973) Altitudinal distribution of forest vegetation on Mt. Poroshiri, the Hidaka range, Hokkaido, Japan. Memoirs of the National Science

20 Masahiko OHSAWA

Museum, 6: 183-200. Ohsawa, M., W. Wildpret and M. del Arco, (1999) Anaga

Cloud Forest. A Comaparative Study on Evergreen Broad- leaved Forests and Trees of the Canary Islands and Japan. Chiba University.

Sakai, A. (1982) Extraorgan freezing of primordial shoots of winter buds of conifer. In: P.H. Li and A. Sakai, eds., Plant Cold Hardiness and Freezing Stress, 2, Academic Press, NY.

Schweinfurth, U. (1957) Die horizontale und vertikale Verbrei-tung der Vegetation im Himalaya. Bonner Geographische Abhandlungen, 20: 373.

Su, H-J (1984) Studies on the climatic and vegetation types of the natural forests in Taiwan. Quarterly Journal of Chinese Forestry, 17: 1-14, 57-73.

Tang, C.Q. and M. Ohsawa (1997) Zonal transition of ever-green, deciduous, and coniferous forests along the altitudinal gradient on a humid subtropical mountains, Mt. Emei, Sichuan, China. Plant Ecology, 133:63-78.

Troll, C. (1959) Die tropische Gebirge. Ihre dreidimensionale klimatische und pflanzengeographische Zonierung. Bonner Geographische Abhandlungen, 25:1-93.

Troll, C. (1962) Die dreidimensionale Landschaftgliederung der Erde. Herman von Wissmann-Festschrift. Tübingen (Selbstverlag d. Geogr. Inst. der Universität), 54-80.

Troup, R.S. (1926) Problems of forest ecology in India. In: A.G. Tansley and T.F. Chipp, eds., Aims and Methods in the Study of Vegetation, The British Empire Vegetation Committee, London, pp. 283-313.

Van Steenis, C.G.G.J. (1972) The mountain flora of Java. E.J. Brill. Leiden.

von Wissman, H. (1939) Die Klima- und Vegetationsgebiete Eurasiens. Zeitschrift der Gesellschaft für Erdkunde zur Berlin, 1939. 1-14.

Wangda, P. and M. Ohsawa (2006) Gradational forest change along the climatically dry valley slopes of Bhutan in the midst of humid eastern Himalaya. Plant Ecology, 186: 109-128.

Weischet, W. (1988) Einführung in die Allgemeine Klimatologie. Teubner Studien bücher, Stuttgart.

Whitmore, T.C. (1984) Tropical Rain Forests of the Far East. 2nd ed., Oxford Univ. Press, Oxford.

Wolfe, J.A. (1979) Temperature parameters of humid to mesic forests of eastern Asia and relation to forests of other regions of the northern hemisphere and Australasia. Geological Survey Professional Paper 1106, USGPO, Washington.

Yoshino, M. (2006) Global warning and moutain environment. Global Environmental Research, 10: 3-12.

(Received 24 July 2006, Accepted 28 July 2006)