Yasushi Mitsuda

Use of GIS-derived environmental factors in predicting site indices in Japanese larch plantations in Hokkaido

The objectives of this study were (1) to evaluate the effects of environmental factors derived from GIS on tree-height growth of Japanese larch (Larix kaempferi) and (2) to develop a best-fit regression model for its site index. Based on data from 40 sample plots situated in an even-aged (38 years), pure, and undamaged Japanese larch stand, multiple regression models for a site index of Japanese larch were constructed using environmental factors as independent variables. The average slope gradient, effective relief, distance from ridge, flow accumulation, degree of exposure, shading, solar radiation index, and gravitational water index were used as environmental factors and calculated on GIS using digital elevation model data. These factors were related to the Japanese larch site index through multiple-regression analysis. The result showed that the most effective factor for estimating site index was the degree of exposure. Through a backward stepwise procedure, the degree of exposure, shading, and average slope gradient were selected for a best-fit regression model. This model explained 72% of the variance in site index, with standard error estimates of 1.75 m. This strong relationship suggests that GIS-derived environmental factors can be used to predict site indices of Japanese larch.

Effects of Competitive and Cooperative Interaction among Neighboring Trees on Tree Growth in a Naturally Regenerated Even-aged Larix sibirica Stand in Considering Height Stratification

This study attempted to explain the variation in growth of individual trees in a naturally regenerated, even-agedLarix sibirica stand using indices that represented the competitive and cooperative interactions among neighboring trees. These interaction indices and DBH were used in stepwise multiple regression procedures to model the growth of individual trees. However, when the data from all trees were used, DBH was the only factor accepted in the growth model. Since DBH can be influenced by the cumulative effect of past interaction and other environmental factors, we stratified the stand into three height strata and repeated the stepwise procedure for each stratum to remove the cumulative effect represented by DBH. Several competition and/or cooperation indices were accepted in growth models of the lower, middle and upper strata. In each stratum, the residual mean square of the growth model was smaller than that of all strata. These facts suggested that height stratification was generally successful in reducing the cumulative effect of past interaction and other factors. The cooperation indices that suggested protection from wind stress by neighboring trees was a significant variable in the growth models of all three strata. This demonstrated that cooperative interaction should be considered in the explanation of variation in tree growth in dry and windy climates such as the present study region.