Summary 1We used aerial photographs to analyse canopy dynamics over 32 years in a 4-ha permanent plot of an old-growth evergreen broad-leaved forest in the Tatera Forest Reserve, south-western Japan. 2 We used aerial photographs taken in 1966, 1983, 1993 and 1998 to create digital elevation models (on a 2.5-m grid) of the canopy surface in a 10-ha area that included the 4-ha plot. A topographic map of the ground surface of the 4-ha plot was also created from ground measurements taken in 1990. The canopy height at each location was calculated and the 'vegetation height profile technique' was applied to quantify canopy dynamics. 3 The size distributions of gaps at each observation time showed peaks at both small and intermediate sizes. The total number of gaps and gap density increased over the study period, while total gap area, mean gap size and maximum gap size fluctuated. While the overall mean rates of gap formation and closure (0.58 vs. 0.50% year − 1 ) were nearly balanced, we found large temporal variation. 4 Of the 6400 of 2.5 × 2.5-m quadrats, 3.4% remained as gaps throughout the 32-year period. These tended to be in the centres of large, old gaps, whereas those that experienced re-disturbance (2.5%) were distributed along the edges of old gaps. A large proportion (74.5%) maintained a closed canopy state. 5 Meteorological data suggested that the variation in gap formation and gap closure rates may be related to year-to-year variations in the number and severity of typhoons. 6 This study suggests that long-term and large openings are essential for the establishment of deciduous broad-leaved and shade intolerant pioneer species in predominantly evergreen broad-leaved forests.

Summary1 Long-term canopy dynamics in a large area of temperate old-growth beech forest in the Daisen Forest Reserve, south-western Japan (11.56 ha studied over 43 years), were investigated using digital elevation models (DEMs) of the canopy surface, constructed from aerial photographs taken in the growing season (i.e. with foliage) in 1958, 1978, 1992 and 2001. A ground surface DEM at the same resolution (a 2.5 × 2.5 m grid) was constructed using aerial photographs taken when foliage was absent (winter 2002). Canopy height data were obtained by calculating differences in elevation between the canopy and the ground surface, and a canopy height profile was constructed. 2 Topographic data for a 4-ha plot, located within the 11.56-ha area, were obtained via a ground survey and used to validate the ground surface DEM derived from aerial photographs. 3 Canopy height class distributions changed significantly over the 43 years. The total number of gaps, defined as areas where canopy height was ≤ 15 m, decreased but total gap area increased over time. Total gap area in 2001 was twice that of 1958. The density of gaps decreased as gap size increased. 4 Gap formation rates increased from 0.47% year − 1 (1958-78) to 1.30% year − 1 (1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001), with a mean of 0.77% year − 1 , and substantially exceeded closure rates, which fluctuated from 0.28% year − 1 (1958-78) to 0.54% year − 1 , with a mean of 0.39% year − 1 . Gaps generally expanded and became connected to each other. 5 Temporal variation in gap formation and closure might be correlated with the frequency and severity of typhoon disturbances but, if the observed trends continue, this old-growth beech stand may become an open stand. The long-term dynamics of this forest type appear to be far from equilibrium.

No abstract

The long-term changes of the area of Sasa nipponica grassland, bark stripping damage to trees by sika deer, Cervus nippon, and sika deer densities on Mt. Ohdaigahara were assessed with aerial photographs, information from previous studies, and field experiments. The grasslands expanded year after year. The expansion rate was highest from 1982 to 1992, and lowest from 1992 to 1997. Bark-stripping damage on coniferous trees was heavier than that on broadleaf trees in both the early 1980s and 1998. The deer density in the eastern part of the area had been high (14.4-64.3 km -2 ) from 1984 until 1998. We presume that rich growth of S. nipponica and heavy feeding pressure by sika deer over the long term caused expansion of S. nipponica grasslands by killing canopy trees by bark stripping and preventing forest regeneration.

No abstract