Old-growth Cryptomeria japonica forest on Yakushima Island has been affected by large-scale logging activities that began approximately 350 years ago and continued over 300 years. Forests on the island currently consist of 200-300 year-old, regenerated Cr. japonica, and 400- to over 1000 year-old Cr. japonica that survived logging activities. The objective of the present study was to understand the long-term growth patterns of Cr. japonica on Yakushima Island over the last several hundred years. Tree-ring analysis using samples obtained from 28 Cr. japonica individuals that regenerated subsequent to the inaugural year of logging was employed to develop an understanding of long-term growth patterns in regenerated Cr. japonica. Growth rate of basal area increment (BAI) was calculated from tree-ring series and results indicated diversity among individual diameter growth curves. BAI growth rate increased with age until approximately 110 years, a period of increase longer than that observed in Cr. japonica in an artificial forest. BAI growth rate across diameter classes showed an initial rapid increase until 30 cm, followed by a slow increase between 30-50 cm before plateau. Growth patterns beyond 70 cm could not be determined as sample size were inadequate. Comparison of these results with monitoring results suggested that growth rates were higher 100-150 years ago than they were during the last 30 years, which further indicated that gap formation resulting from large-scale logging activity may have had a positive impact on growth rate in Cr. japonica.

View full abstract

Old-growth Cryptomeria japonica forests on Yakushima Island, Japan have been affected by logging activities. The most ancient record related to logging of Cr. japonica on Yakushima Island dates back to 1563. Systematic large scale logging activities of Cr. japonica occurred over a 300 year period starting in 1642. Forests on the island currently consist of 200-300 year-old regenerated Cr. japonica, although 400 to over 1000 year-old trees have survived logging activities. The objective of the present study was to identify the points in time and the scale of past disturbances and to verify of ancient records of logging activities using dendroecological approaches. Tree-ring analysis using samples obtained from eight Cr. japonica individuals was employed to develop an understanding of and pinpoint the time of past disturbances. Percent growth change (%GC) was calculated to detect release events caused by gaps created by human or natural disturbances and basal area increments (BAI) were calculated to detect growth rates. One older sample tree showed evidence of release events from the middle of 1700s to about 1800 and at the end of the 1900s, and another old-aged sample tree showed similar evidence from 1600 to the middle of 1900s. The BAI value showed an increase for one old-aged sample tree from the middle of 1700s to the beginning of 1900s and the other old-aged sample tree from 1800 to the end of 1900s; thus both trees showed high BAI values for 150 years after releases. Germination year of six regenerated trees subsequent to the inaugural year of logging was estimated within the relatively narrow range between 1791 and 1835. This regeneration timing was consistent with release events followed by high BAI values of old-aged trees. Evidence showing all regenerated samples germinated on stumps and logs indicates the detected releases might have been caused by large scale of logging activities. This study clarified that large scale of logging activity encouraged the growth rate of approximately 500 to 600 years old trees, and also large scale of disturbance was important for regeneration of Cr. japonica.

View full abstract

The objectives of this study were to formulate a long-term harvest scheduling model, involving two types of rotation constrained by the available size of the work force, using 0-1 integer programming and then apply the model to plantation forests in the University of Tokyo Chiba Forest as a case study. The following three silvicultural systems were considered: an 80-year and a 160-year rotation clearcutting system and a non-clearcutting system. The minimum amount of labour required to harvest the minimum area was calculated and then that figure was increased to model its effect on harvesting. Subcompartments better suited to timber production tended to be assigned to clearcutting. There was a tendency for subcompartments with a better site class to be assigned to the shorter rotation and subcompartments with a shorter yarding distance to be assigned to the longer rotation. As the size of the available work force increased, subcompartments less well suited to timber production were also assigned to clearcutting and the harvest volume per person-day and the clear cut area per person-day decreased. The longer rotation was efficient with a smaller work force and the shorter rotation appeared more efficient as the size of the work force increased.

View full abstract

We propose a method for estimating the stand density in an even-aged pure stand on a flat slope by hemispherical photography. This method is an application of the method for estimating the stand density proposed by Suzuki, i.e., distance method, and the method for estimating canopy-gap size using two photographs taken at two different heights. First, we assume a homogeneous tree height within an even-aged pure stand on a flat slope, and the straight and vertical axes of all stems. At a sample point selected from the stand randomly, two hemispherical photographs are taken with the vertically mounted camera equipped with a fish-eye lens at two different heights. Next, the radial distances between the center of the hemispherical image and the tip position of tree having the third-largest elevation angle (third-smallest zenith angle) are measured on each photograph taken at the two different heights. By substituting the measured radial distances into the relationship between radial distance and zenith angle of the fish-eye lens used, the elevation angles between camera and tip of the third-nearest tree can be obtained. Using these values of elevation angle, the distance from the photographic sample point to the third-nearest tree can be geometrically computed, and then the stand density can be estimated by substituting the computed distance into the equation of Suzuki's distance method.

View full abstract