Response of Japanese beech (Fagus japonica Maxim.) sprouts to canopy gaps

Vegetatio 124: 1-8, 1996. 1 © 1996 KluwerAcademicPublishers. Printedin Belgium. Response of Japanese beech (Fagusjaponica Maxim.) sprouts to canopy gaps T a t s u h i r o O h k u b o 1,*, T a k e o T a n i m o t o 1 & R o b P e t e r s 2 1 Department of Forest Science, Faculty of Agriculture, Utsunomiya Univ., 350 Mine-machi, Utsunomiya 321, Japan; 2 Department of Forestry, Wageningen Agricultural University, P.O,Box 342, Wageningen 6700 AH, The Netherlands (*author for correspondence: e-mail: ohkubo@cc.utsunomiya-u.ac.jp) Received 3 June 1994; accepted in revised form 8 December1995 Abstract The response of Japanese beech (Fagusjaponica Maxim.) sprouts to canopy gaps in natural beech forest in central Japan was studied using two contrasted gaps in which tree-ring chronologies of regenerating stems were analyzed. The gaps were created by uprooting of a single Quercus mongolica var. grosseserrata stem (diameter: 50 cm; gap size: 40 m2; 23 years old) and by concurrent uprootings of four E japonica stools (gap size: 180 m2; 30 years old). Japanese beech sprouts emerged before and after the gap formation and dominated stem populations in both gaps. In gaps, growth of F. japonica sprouts was equal or lower than growth of stems of seed origin, but most sprouts (E japonica, Acer mono var. marmoratum) appeared a few years before emergence of seedlings. The small gap created by single stem fall was dominated by some beech sprouts from stools adjacent to the gap. The multiple gap was not closed by beech sprouts from stools surrounding the gap, but some dominant beech stems were resprouts from the uprooted beech stools. The existence of a 'sprout bank' under the canopy may play an important role in the closing process of gaps in natural Japanese beech forest. Introduction The formation of canopy gaps is important in the dynamics of old growth beech forests (Watt 1947; Hara 1983; Nakashizuka & Numata 1982; Koop & Hilgen 1987; Maeda 1988; Ohkubo et al. 1988; Yamamoto 1989; Peters 1992). After formation of small - scale canopy gaps, initiated by uprooting or death of one or a few standing canopy stems, canopy closure occurs through crown expansion of adjacent canopy trees (Trymble & Tyron 1966) and/or vertical growth of saplings (Bormann & Likens 1979). Sprouting response of the main tree species in the canopy is also an important mode of canopy-gap closure (Putz & Brokaw 1989; Yamamoto 1992a 1992b; Ito 1993). For example, sprouting can result in persistence of early successional species, such as Populus alba and Populus tremula, even in smaller canopy gaps of shadetolerant species (Koop 1987). Yet, sprouting responses of canopy species to canopy openings have not been analyzed in many forests. Japanese beech (Fagus japonica Maxim.) is a canopy species in old growth forests of the cool temperate zone of the Pacific side of Japan, where it cooccurs with Ecrenata, Quercus mongolica var. grosseserrata and some Acer species. Throughout its geographic distribution range, Ejaponica stems naturally form vigorous sprouts from the root-collar (Tohyama 1965; Ohkubo et al. 1988; Peters & Ohkubo 1990; Ohkubo 1992). The sprouts are connected with each other at their epigeal parts and form a permanent base of a individual tree (here we call 'stool'). Thus a stool becomes a cluster of stems that reach into the canopy. These stems may become separated stools (Ohkubo 1992). Within a F. japonica stool, one or more of these sprouts may replace a canopy stem that has died. Thus, a E japonica stool can reach old age. Because new sprouts are formed on the outside of each stool, a stool increases in size. The genetic homogeneity of such clusters of stems has been shown through isozyme analysis (Kitamura et al. 1992). In this study we focused on the response of E japonica spouts to new canopy gaps. We compared 2 the initiation and initial height growth of F. japonica stems/saplings of sprout origin and other tree species. We compared these responses in a small gap created by a single tree fall with the response in a large gap created by a multiple-tree fall. Study sites and methods The two study sites were located on Mount Myojingatake (1594.5 m a.s.1.; 36 ° 53' N, 139 ° 48' E; Okukinu research site) and on Mount Shakagatake (1794.9 m a.s.l., 36 ° 55' N, 139 ° 38' E; Takahara research site) in the Kanto mountains, northern part of Kanto plain, the Pacific side of Honshu island, Japan. The Okukinu plot included a small gap and the Takahara plot a large gap. The distance between the two study sites is about 16 kin. The climate of the study sites is characterized by lower precipitation (snowfall) in winter. Annual mean precipitation in the areas is ca. 1500 m m -1700 m m and mean annual maximum depth of snow of the areas is 20 cm - 50 cm. The warmth indices (W.I.) are 54.9 in Okukinu and 62.4 in Takahara indicating that the areas are located in the temperate deciduous forest zone (Kira 1945). The study sites are in a Fagusjaponica - E crenata forest reserve of national forests, and E japonica dominates the areas whereas E crenata, Quercus mongolica var. grosseserrata and Acer spp. are also found in the forest canopy (Ohkubo 1992). In August 1984 a quadrat of 40 m x 20 m was set up in Okukinu, at an altitude of ca. 1220 m, on a northwest facing steep slope of 41 degrees. The quadrat was planned to be clear cut by the forest agency. In August 1986 another quadrat of 40 m x 10 m (belt-transect No. 3 in Peters & Ohkubo 1990) was set up in Takahara, at an altitude of 890 m, on a southeast facing slope of 28 degrees. In each quadrat, we identified woody species and measured diameter at breast height (DBH) and tree height of the stems taller than 1.3 m. We mapped stem position and crown projection of the stems taller than 5 m. Fagusjaponica is the first dominant species [78% of total Basal Area (BA)=Relative Dominance of Basal Area (RDBA)] in Okukinu and is the third one (9% in RDBA) in Takahara. But in a quadrat (50 m x 50 m) near the quadrat of this study in Takahara E japonica dominated (68% in RDBA). On the other hand E crenata is the first dominant species (59% in RDBA) in Takahara and is the second one (17.2% in RDBA) in Okukinu. Main medium sized tree and shrub species were Acer Shirasawanum, A. distylum and Fraxinus lanuginosa in Okukinu and Prunus grayana, Meliosma myriantha and Carpinus cordata in Takahara, respectively. The undergrowth was dominated by 0.6 m tall dwarf bamboo, Sasa nipponica, which covered 43% of the area in Okukinu and 88% in Takahara (Ohkubo 1992). In each quadrat we analyzed age and height growth patterns of the stems using stem disks and increment cores. In May 1985 in Okukinu, before clear cutting all standing stems (including four E japonica sprouts) in the canopy gap (n=27) and six of small sized-stems (including five E japonica sprouts) under the canopy area were cut and stem disks were taken at 0 m and every 1 m up from 0.3 m in height. After clear cutting, stem disks were taken from 214 out of the 227 stems at height of the stumps (50 cm from ground level). In September 1986 in Takahara all stems were cut in the gap area (two sub-quadrats B & C in Figure 2 of 10 m x 20 m) of the quadrat. Stem disks were taken at 0 m and every 1 m up from 0.3 m in height from 118 out of 138 stems. And in the canopy area (two sub-quadrats A & D in Figure 2 of 10 m x 10 m) of the quadrat increment cores were taken from 11 canopy stems out of 29 stems by increment borer. In the laboratory disks and cores were sanded or shaved to reveal the growth rings. Every five ring widths from the center were measured to the nearest 0.01 m m under a binocular microscope. The disks were cross-dated from four directions. Stem age was estimated using stem-disks or increment cores taken at 0.5 m in height. We assumed that the stems were 10 years old when they reached 0.5 m in height. In disks with missing the central regions, the missing number of tree rings was estimated by extrapolation using the average width of the 20 oldest rings visible. We defined canopy stems as stems that cannot be suppressed by neighboring stems. Canopy gaps are open areas in the overstory canopy not containing any leaves. Each of the two quadrats included a tree fall gap, the size of which was different. In Okukinu, the tree fall gap measured 40 m 2 and was created by uprooting of a Q. mongolica var. grosseserrata tree (50 cm in diameter), a small gap by single-tree fall (Figure 1). In Takahara, the tree fall gap measured 180 m 2 and was created by uprooting of four E japonica stools (ST1-ST4 in Figure 2), a large gap by multipletree fall. To estimate the age of the two gaps in each quadrat we used the sprouts that had emerged after the gap was created. These sprouts grew on uprooted, but still living stems and stools. From their direction of growth these sprouts were judged to have appeared after the uprooting of the mother stem (Suzuki 1981). Figure 1. A crown projection map of the crowns of stems taller than 5 m in a small gap created by single-tree fall in the Okukinuquadrat (20 m x 40 m). Solid and broken lines represent the margin of canopy and subeanopy crowns, respectively. Small solid and empty circles represent the positions of stem base of Fagusjaponica and other species, respectively. The largest shaded area represents the canopy gap created by the uprooting of Quercus mongolica var. grosseserrata. In the small gap in Okukinu we used an Acer mono var. marmoratum sprout that had emerged from the mother stem (146 years old), which was pushed over by the gap maker. One of two new emerged vertical sprouts on an uprootedA, mono var. marmoratum stem was 22 yr. old. The new sprout grew rapidly in height (1.97 m in the first 5 years) and occupied the upper part of the gap (8.8 m) (open square in Figure 4). In the large gap in Takahara there were twenty E japonica sprouts produced by four uprooted beech stools (ST1-ST4 in Figure 2). The sprouts in the stools were also used for gap dating. A m o n g them, emergence of seven sprouts occurred about 29 yr. ago. The oldest sprout was one of six F. japonica sprouts from one leaning living stem (110 yr. old) in the uprooted stool ST3. Including a year lag (Payette et al. 1990) the gap ages were around 23 yr. old in the small gap in Okukinu and around 30 yr. old in the large gap in Takahara. Nomenclature follows Satake (1989). Results Species composition and distribution of E japonica sprouts in the gaps and in the canopy area In each gap (Figures I & 2) stem density and basal area (BA) of stems involved were different: 27 stems (0.67 stems m -2, 13.9 cm 2 in B A m -2) in the small gap in Okukinu and 138 stems (0.77 stems m -2, 11.3 cm 2 in B A m -2) in the large gap in Takahara. In the small gap F. japonica was the second dominant species (RDBA: 35.6%) and Carpinusjaponica was the main co-dominant species. In the large gap E japonica was also the second dominant species (RDBA: 18.2%), and the other dominant species were Prunus grayana, Cornus controversa, Magnolia obovata and Betula grossa . In both sites all F. japonica stems were of sprout origin, with the exception of two F. japonica saplings (100 cm and 55 cm in height) of seed origin on the mound of the uprooted Q. mongolica var. grosseserrata stem in the small gap in Okukinu. In the small gap in Okukinu all E japonica sprouts originated from the beech stools adjacent to the gap (Fj3-Fj5 in Figure 1). On the other hand all sprouts in the large gap in Fe 0. Fe lore 0 I xx I x x . XI l A B C D Figure 2. A crown projection map of the stems taller than 5 m in a large gap created by multiple-treefall of four E japonica stools in the Takahara quadrat (10 m x 40 m). Upper: crown projectionmap of crowns of stems (height>5 m) in the canopy area A and D, and of crowns of species (0.5<beight_<5m) in the gap B and C. Solid circles representbases of stems. A shaded area represents the canopy gap created by the uprootingof four Fagusjaponica stools. Fc:Faguscrenata. Lower: Positionsof uprooted Fagusjaponica stools and stems (ST1-ST4),and living stems sprouted from the uprooted stools. Solid and open bars represent live and dead fallen stems, respectively. Shaded areas represent mounds. Open circles indicatehorizontalprojectionof crowns of Fagusjaponica stems. Dashed lines indicatethe boundarybetween areas of different topography. Takahara had emerged from the uprooted beech stools (ST1-ST4 in Figure 2) in the gap. Fagusjaponica crowns entirely covered the canopy area of the quadrat at the border of the small gap in Okukinu, but not the border of the large gap in Takahara (Figures 1 & 2). The canopy area adjacent to the quadrat of the large gap in Takahara was entirely covered by E japonica crowns (Ohkubo 1992). Stem bases of E japonica showed clump structure and formed stools, e.g., from stool F j l to stool F j l 2 in Okukinu and a stool Fj in Takahara (Ohkubo 1992). In each stool there were a few large canopy crown and some small sub-canopy crowns. Most of the subcanopy crowns were located below the large canopy crowns. Age structure of stems in the gaps and in the canopy area Comparison of size structure of stems before and after gap opening was made in the two gaps of different size. Size distributions of E japonica and other species stems were significantly different between two gaps (Kolmogorov-Smirnov test, p<0.001). Regardless of 35- 2520ti Sma,, gap in Okukinu 0 Canopy area in Okukinu ~10I 7- 0 E I I I I I I I I 50 I I I I I I I I I I I I 100 I I I 150 Age in 1986 300 I o ..................... .,',,": ,,=,,,,,i,,...i,..,..,.,,,,,;. o 5o lOO 15o 200 250 2520 E O iD. 10- 03 E Q.i ~6 6 z Canopy area in Takahara 'r-- 15- E Age in 1986 c'J Large gap in Takahara 0- 0 ~ ~ ,'if, . . . . . . . . . ,'if,,, ,, . . . . . . . . . . . 50 100 150 Age in 1 9 8 4 F1, '5 O Z +1 .....! .............. 0 50 ............... I] .... rl, 100 150 200 250 300 Age in 1984 Figure 3. Age frequency distributions of Fagusjaponica stems (closed bars)and the stems of the other species (open bars) in the gap (left) and the canopy area (right) in a small gap (upper) in the Okukinu quadrat and in a large gap (lower) in the Takahara quadrat, respectively. Arrow denotes the estimated age of the gap. gap size most of all analyzed stems (81% in the small gap in Okukinu and 96% in the large gap in Takahara) emerged after the canopy opening, this emergence was concentrated within 25 years (peak at 10 years after gap creation in the small gap in Okukinu) and within 30 years (peaks at 5 and 15 years ago after gap creation in the large gap in Takahara). In each gap emergence of E japonica sprouts was also within 5 years in Okukinu and 20 years in Takahara after canopy opening, respectively. Especially in the large gap in Takahara pioneer species like Clethra barbinervis, Zanthoxylum piperitum and Clerodendrum trichotomum emerged within 3 years after the opening of the canopy. Before gap creation the emergence pattern of stems was sporadic in each gap. Sixty percent in the small gap in Okukinu and 71% in the large gap in Takahara of all analyzed stems were E japonica sprouts. In the gaps F.japonica stems of 110 years old were leaning stems, which was caused by uprooting of the beech stool (ST3 in Figtire 2). In the gaps, the oldest stems were Acer mono var. marmoratum of 148 years old and Rhododendron quinquefolium of 190 years in Takahara. In the canopy area of the quadrat in Okukinu, 54% of all analyzed stems were younger than the canopy gap (30 years) because of presence of many the beech sprouts, but some sprouts were older than 200 yr. The oldest stem was a E japonica stem of 252 years old. The E japonica sprouts became established almost continuously since 165 years ago. The number of E japonica sprouts gradually decreased with increasing of age. The differences of the age distributions between the canopy areas at the two quadrats were derived from different species composition and different site histories of the canopy area. But in Takahara there were many E japonica stools around the quadrat comparable to those in Okukinu (Peters & Ohkubo 1990). Growth pattern of stems in the gaps In each gap, the mean annual height growth rates for first five years of E japonica sprouts that appeared after gap creation were 0.36 m y - I in the small gap in Okukinu and 0.32 m y - I in the large gap in Takahara, which is about three times (3.43 in Okukinu and 3.28 in Takahara) as high as that of the beech sprouts before canopy opening. In the small gap in Okukinu, the mean annual height growth rate for first five years of beech sprouts was almost the same as that of the sprouts that emerged Small gap in Okukinu (D ~>, E v v I (D o> 0.8 0.6(D (- 4, 0.2- i • , O,0~0, o No. of years from 1984 O o I ~ o~, ~ , o ~ e- -6o- o- 0-30-zo-lo ._o~ 10 Z~ t- .< i 8- .$-a ..£E "1- ICD ~ O 0.6- 6420 • io I a3D 1 030 10- E v 6- cO3 4- .m -r" I I -6o -go -;,o -do -~o -1'o 12 0 N o . of y e a r s f r o m 1 9 8 4 ,,333 I(~) IOD I~DO i ~ ~ • 8- I o No. of years from 1986 c- {- E v I O 0.8- 0.2- 0, ~D c- < I 0.4- u -6o- o- o-3o-zo- o c- a IO0 0 ,130 _ n 0 3 our 0.4- o t- Large gap in Takahara E • ICl300 • 20 -6o- • • I C ~ I OO¢:113:D OI C D ~ I 133£IDCD I O OO(3 I, o-,io-3o- o-1'o o No. of years from 1986 Figure 4. Annual height growth rate (upper) for first five years and height (lower) of E japonica sprouts (solid circles) and other trees (open circles) of different ages in a small gap in Okukinu (left) and in a large gap in Takahara (right). Vertical broken lines show the estimated age of the gap (23 years ago in Okukinu, 30 years ago in Takahara). Open square shows the value of a sprout from a leaned maple (Acer mono vat. marmoratum) stem pushed down to the ground by a fallen oak (Quercus mongolica var. grosseserrata) stem (gap maker) in the small gap in Okukinu. from a leaning A. re•no var. marmoratum stem, which was pushed over by the gap maker (0.39 m y - l ) . In Okukinu the mean annual height growth rate for first five years of stems other than E japonica (0.25 m y - 1) was lower than that of the beech sprouts. In the large gap in Takahara the mean annual height growth rate for first five years of E japonica sprouts that appeared after gap creation (0.32 m y-1)was not significantly different from that of stems of species other than E japonica (0.34 m y-1 ) (Mann-Whitney U-test,/9=0.87). In the small gap in Okukinu, the earlier initiation group of E japonica sprouts and A. mono var. marmoratum sprouts after canopy opening occupied upper part around 8 m in height, and the later initiation group of other saplings was lower than the former, ranging from 2 m to 6 m in height. Also in the large gap in Takahara the earlier initiation group of stems occupied upper part. The most upper part was dominated by the stems other than E japonica sprouts. The mean height of E japonica sprouts that appeared after gap creation was not significantly different from that of stems of species other than E japonica (Mann-Whitney U-test, /9=0.25). Discussion In both sites species composition and size structure of the studied forests were similar, so performance of F. japonica sprouts and the stems of other tree species responding to canopy opening of different size can be compared. Difference of gap size affects mode of gap closure by the F.. japonica sprouts and the stems of the other species. In case of the small gap in Okukinu, although gap maker was a single tree other than E japonica, crown extension growth towards the center of the gap by sprouts of E japonica from neighboring the beech stools was possible. The same was found by Ohkubo et al. (1988) for E japonica sprouts in Chichibu. All the beech sprouts in the small gap were composed of the ingrown sprouts from stools at the gap border. When a small canopy gap created by a single tree fall of Fagus japonica in a beech stool and the canopy gap is filled by beech sprouts in the same stool, canopy recruitment is more quick in the small gap in Okukinu. Uprooting of beech stools creates larger gaps because the crowns of the several stems in a single stool occupy a larger area than a single stem (Ohkubo et al. 1988; Ohkubo 1992). Our observation in Okukinu and in Takahara is that uprootings of E japonica stools are rare events, and that they occur mainly on steep slopes with shallow soils. The large gap in Takahara was closed by upgrowth of E japonica stems which were resprouted from the uprooted beech stools and saplings of other tree species in the gap. Gap creation initiates resprouting and growth of stems of both F. japonica and other species. The results suggest that resprouts of E japonica respond more quickly to canopy openings than stems of other tree species. F. japonica sprouts contribute more to canopy closure in small gaps than in large gaps. A similar case was reported in moist tropical rain forest in Panama (Brokaw 1985; Putz & Brokaw 1989). On the contrary in warm temperate forests of the southeastern U.S. Platt & Hermann (1986) indicated that regeneration of species that are positively associated with small gaps is important mode of overstory gap closure. Fagus japonica stool forms basal sprouts prior to disturbances beneath closed canopy regardless of stool size and the E japonica stools expand, and parts of the stools become disconnected (Ohkubo 1992). Similar cases were found in two beech species in Korea (E multinervis in Kim et al. 1986) and in China (E engleriana in Peters 1992), American basswood (Tilia americana) in U. S. (Fowells 1965), Tillia cordata in Poland (Pawlaczyk 1991)and Cercidiphyllumjaponicum in J a p a n (Miyabe & Kudo 1986). In the recovery of the subtropical forests of the Caribbean from hurricanes, resprouting of damaged trees plays an important role and the sprouting ability depends on species, size and intensity of damage (Basnet 1993; Bellingham et al. 1994; Zimmerman et al. 1994). In Jamaican montane forests stems more than 10 cm in DBH sprouted more frequently and produced more sprouts per stem, than those less than 10 cm in DBH (Bellingham et al. 1994). But in a moist tropical forest in Panama, after stem breakage, large trees often fall to sprout because of lack of resting buds or inability of the buds present to emerge through thick bark (Putz & Brokaw 1989). Stem breakage in large sized E crenata (above 20 cm) in Japan also results in limited the sprouting abilities (Maeda 1988; Kamitani 1993). After the formation of E japonica sprouts in a stool, the survivorship of sprouts might be affected by light levels determining the extent to which the beech produces sprouts beneath a closed canopy. Magnolia grandifolia in southeastern U. S. is capable of resprouting like E japonica, but does not exhibit this pattern in a closed canopy forest. Disturbances in the beech forests result in frequent local increases in light levels (Platt & Hermann 1986). The survivorship of E japonica sprouts might be affected by local increases in light levels, especially canopy openings due to death of stems inside a stool can change light level directly. Thus, presence of the populations of E japonica sprouts in closed canopy forests can play an important role as a 'sprout bank' analogous to seedling banks (Bellingham et al. 1994). The location of E japonica stools in relation to canopy gaps, the population structure of E japonica sprouts, and the orientation of the beech sprouts in the stool relative to the position of the canopy opening all influence the mode and process of gap closure by E japonica sprouts in canopy openings of different sizes. Acknowledgments We thank T. Hamaya, T. Maeda, M. Kaji, K.-E Cat and two anonymous referees for their advises and comments on the manuscript. We also thank M. Nakata, M. Watanabe, M. van Zalingen, A. Niwa, A. Oikawa and the members of the laboratory of Forest Botany, The University of Tokyo for their help in the field. The staffs of Utsunomiya and Yaita branch office of the Forestry Agency provided us with research facilities and help. References BaSher, K. 1993. Recovery of a tropical rain forest after hurricane damage. Vegetatio 119: 1-4. Bellingham, P. J., Tanner, E. V. L & Healey, J. R. 1994. Sprouting of trees in Jamaican montane forests, after a hurricane. J. of Ecology 82: 747-758. Bormann, F. H. & Likens, G. E. 1979. 'Pattern and Process in a Forested Ecosystem', Springer-Verlag, Berlin and New York, 253 pp. Brokaw, N. V. L. 1985. Treefalls, regrowth and community structure in tropical forests. Pages 53-69 in S. T. A. Pickett and P. S. White, editors. The ecology of natural disturbance and patch dynamics. Academic Press, New York, USA. Fowells, H. A. (Ed.). 1965. Silvics of forest trees of the United States. USDA Agric. Handb. No. 271: 692-698. Hara, M. 1983. A study of regeneration process of a Japanese beech forest. Ecol. Rev. 20: 115-119. Ito, S. 1993. Timing of sprouting in Hamabiwa (Litsea japonica Juss.) growing in a coastal dwarf forest. Bull. Kyushu Univ. For. 67: 1-8. Kamitani, T. 1993. Ecological studies on regeneration of beech (Fagm~crenata Blume) coppice forests in heavy snowfall region. Memoirs of Fat. of Agric. Niigata Univ. 30: 1-108. (in Japanese with English summary) Kira, T. 1945. A new classification of climate in eastern Asia as the basis for agricultural geography. 23 pp. Horticultural Institute, Kyoto Univ., Kyoto (in Japanese) Kitamura, K., Okuizumi, H., Seki, T., Niiyama, K. & Shiraishi, S. 1992. Isozyme analysis of the mating system in natural populations ofFagus crenata and E japonica.. Jpn. J. Ecol. 42: 61-69. (in Japanese with English summary) Koop, H. 1987. Vegetative reproduction of trees in some European natural forests. Vegetatio 72:103-110. Koop, H. & Hilgen, P., 1987. Forest dynamics and regeneration mosaic shifts in unexploited beech (Fagus sylvatica) stands at Fontainebleau (France). For. Ecol. Management 20:135-150. Maeda, T. Studies on natural regeneration of beech (Fagus crenata Blume). Spec. Bull. Coll. Agric., Utsunomiya Univ. no.46, 79 pp. (in Japanese with English summary) Miyabe, K. & Kudo, Y. 1986. Icones of the essential forest trees of Hokkaido. Hokkaido University Press, Sapporo. Nakashizuka, T. & Numata, M. 1982. Regeneration process of climax beech forests. I. Structure of a beech forest with the undergrowth of Sasa. Jpo. J. Ecol. 32: 57-67. Ohkubo, T. 1992. Structure and dynamics of Japanese beech (Fagus japonica Maxim.) stools and sprouts in the regeneration of the natural forests. Vegetatio 101: 65-80. Ohkubo. T., Kaji, M.& Hamaya, T. 1988. Structure of primary Japanese beech (Fagusjaponica Maxim.) forests in the Chichibu mountains, central Japan, with special reference to regeneration processes. Ecol. Res. 3: 101-116. Payette, S., Filion, L. & Delwaide, A. 1990. Disturbance regime of a cold temperate forest as deduced from tree-ring patterns: The Tantare Ecological Researve, Quebec. Can. J. For. Res. 20: 1228--1241. Pawlaczyk, P. 1991. Linden (1ilia cordata Mill.) vegetative regeneration and its ecological role in the Tilio-Carpinetum community of the Bialowieza national park. Phytocoenosis 3 (N.S.) (Seminarium Geobotanicum 1): 161-172. (in Polish with English summary) Peters, R. 1992. Ecology of beech forest in the northern hemisphere. PhD thesis, Wageningen Agric. Univ., 125 pp. Peters, R. & Ohkubo, T. 1990. Architecture and development in Fagus japonica-Fagus crenata forest near Mount Takahara, Japan. J. Veget. Sci. 1: 499-506. Platt, W. J. & Hermann, S. M. 1986. Relationships between dispersal syndrome and characteristics of populations of trees in a mixedspecies forest, pp. 309-321. In: Estrada, A. and Fleming, T. H. (eds), Frugivores and seed dispersal. Dr W. Junk Publishers, Dordrecht. Putz, F. & Brokaw, N. V. L. 1989. Sprouting of broken trees on Barro Colorado Island, Panama. Ecology 70: 508-512. Satake, Y. 1989. Wild flowers of Japan. Woody plants I, II, Herbaceous plants I-III. Heibonsha Publisher, Tokyo Suzuki, E. 1981. Regeneration of Tsuga sieboldii forest III. Regeneration under a canopy gap with low density level of conifer seedlings and a method for estimating the time of gap formation. Jap. J. Ecol. 31: 307-316. (in Japanese with English summary) Trimble, G. R.,Jr. & Tyron, E. H. 1966. Crown encroachment into openings cut in Appalachian hardwood stands. J. For. 64: 104108. Tohyama, M. 1965. Synecological study on the Fagus japonica forest on Mt. Ohmuro, Pref.Yamanashi -The forest vegetation of Mt. Fuji. II. Jap. J. Ecol. 15: 139-142. (in Japanese with English summary) Watt, A. S. 1947. Pattern and process in the plant community. J. Ecol. 35: 1-22. Yamamoto, S. 1989. Gap dynamics in climax Fagus crenata forests. Bot. Mag. Tokyo 102: 93-114. Yamamoto, S. 1992a. Gap characteristics and gap regeneration in primary evergreen broad-leaved forests of western Japan. Bot. Mag. Tokyo 102: 93-114. Yamamoto, S. 1992b. The gap theory in forest dynamics. Bot. Mag. Tokyo 105: 375-383. Zirnmerman, J. K., Everham III, E. M., Waide, R. B., Jeanlodge, D., Taylor, C. M. & Brokaw, N. V. L. 1994. Responses of tree species to hurricane winds in subtropical wet forest in Puerto Rico: implications for tropical tree life histories. J. of Ecology 82:911-922.