| Summary: | 碩士 === 國立臺灣大學 === 森林學系 === 86 === Abstract
In the forest management stressing ecological principles, whole-
stand growth models offering information of stands can''t meet our
requirements in some aspects. In order to get more quintessential
information of trees like the variations of the growth of the different
trees in the same stand, developing individual-tree growth models is
an important issue in forestry. The competition is very important in
many ecological relations affecting the growth of an individual tree,
and the intertree competition is just the most important factor for a
tree in a stand. We can derive the growth of the trees according to
the quantifying of competition. This thesis used 15 size-ratio
competition indexes representing intertree competition to derive
individual-tree growth models. Mathematical functions of these
models were established with multiple regression empirical models
and non-linear semi-empirical models. The models will predict
periodic increments of diameter at breast height(DBH) and those of
basal area(BA) from 23-year-old to 32-year-old Taiwania
cryptomerioides in the Experimental Forest of National Taiwan
University. The different competition indexes, formulas for
determining the competitors, and evaluations of the models will be
compared and analyzed.
The study result showed, compared with original empirical
models, the models established by stepwise regression could
determine the independent variables more effectively. Among growth
models deriving from competition indexes of 15 types, the non-linear
semi-empirical model deriving from Hegyi''s competition index was
the best; the search radius using DBH angle gauge method was better
than the fixed radius method.
|
|---|
