Study on the Stand Dynamics of Natural Forest on Wutoushan Area

• Main Authors: Shu-Chen Shen, 沈淑貞
• Other Authors: Chaur-Tzuhn Chen
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/03592401473769375106

碩士 === 國立屏東科技大學 === 森林系 === 94 === This research defines the natural forest of the southern Wutoushan as an experimental area. By three times in fifteen years (1989~2004) continuous surveys of fifty-five permanent plots on the ground, and cooperating with Weibull probability density functions and the technology of geographic information system, this paper examines structures of the natural forest, the composite tree types, and conditions of spatial changes, in order to provide as the functions of long-term ecological researches and references of forest managements. Using Weibull probability density functions to estimate b and c values of vegetation types and major tree species. It shows that the b values of vegetation types between 10~20, c values between 1~2. Otherwise, the b values of major tree species are from 10~30, c values are from 1~3. So the results of this study indicate that middle- and small-size trees, whose shapes appear as the bell-shaped right-bent curve, are the major populations in the structure of the investigated natural forest. This condition shows that the renewal status of these trees is good. Through investigated data sampling from the t-test and nonparametric K-S test among different periods, all results show that different types of trees and main types of trees are situating in a continuous evolving status. Otherwise, using stepwise discriminant analysis to explore how different periods influence the environmental factors of b and c values from Weibull probability density functions. The result indicates that altitude, slope, aspect, whole light sky space, moisture gradient and soil influence the diameter distributions of trees, and altitude, slope, aspect, whole light sky space, moisture gradient, geology, and soil influence the distributions of forest types. Combining the results above, environmental databases, spacial analysis of geographic information systems, and simulating b and c values from Weibull probability density functions in the investigated area, it can be correctted that the natural forest in this area exists dynamic changes.