The abandonment of plantation sites after clearcutting is increasing in Japan; there is concern that if vegetation fails to recover or develops too slowly after clearcutting, there will be a decline in forest ecosystem services. An appropriate management strategy for rectifying the situation might include predicting vegetation recovery for each abandoned site and prioritizing sites for restoration. Here, we present a regional-scale study that investigated a large number of abandoned plantation clearcut sites distributed across Kyushu Island. The objective was to develop a better general understanding of early-stage vegetation recovery patterns and their determinants, which will be useful in prioritizing sites for restoration. Four vegetation classes -grassland (GL), few pioneer trees (FPT), non-pioneer trees (NPT), and pioneer trees (PT)- were detected by a vegetation survey. Sites that were (1) affected by deer browsing, (2) had relatively young tree stands at the time of clearcutting, and (3) that had been clearcut recently most often recovered to GL or FPT. Under these site conditions, efforts should be directed toward reducing deer grazing to enhance vegetation recovery during the initial stages of regrowth. Restoration effort should be allocated to sites classified as GL even 5 years after clearcutting, which includes a relatively large proportion of the study area. FPT, PT, and NPT sites should be monitored over the long term. This monitoring effort should be used to determine (1) whether tree species recovery is successful in FPT sites and (2) whether PT or NPT sites will convert successfully to evergreen broad-leaved forests (lucidophyllous forests).

View full abstract

Variation in the form-factors for stem surface area was studied in two even-aged pure stands of Japanese larch (Larix kaempferi). In the stands, all living trees were felled for stem analysis, and then the form-factors for stem surface area and stem volume at breast height and 0.7 and 0.5 in relative height, i.e., κ_b, κ_<0.7>, κ_<0.5>, λ_b, λ_<0.7> and λ_<0.5>, were calculated. The coefficients of variation (CV) of κ_<0.7> and κ_<0.5> ranged from 3.2 to 7.4% and from 3.9 to 8.8%, respectively. On the other hand, CV of κ_b ranged from 9.1 to 15.4%, indicating that the variation in κ_b was comparatively larger than those in κ_<0.7> and κ_<0.5>. The ratios of CV of form-factor for stem surface area to that for stem volume ranged from 0.3 to 0.6, indicating that the variation in the form-factors for stem surface area was smaller than that for stem volume. In conclusion, κ_<0.7> and κ_<0.5> would be effective form-factors in estimating the total stem surface area of a stand with the angle count sampling precisely.

View full abstract