مجله جنگل ایران (Jan 2013)
Fitting probability distribution functions for modeling diameter distribution of oak species in pollarded northern Zagros forests (Case study: Armardeh-Baneh)
Abstract
Forest pollarding is one of the major uses of NorthernZagrosForest. This kind of intervention cause major changes in the structure of these forests. Trees diameter distribution is one the most important attributes of forest stands structure. In this research, probability distribution of oak species including Q. branti, Q. infectoria and Q. libani in Armardeh Forests Located in Northern Zagros was investigated for studying the role of uncontrolled pollarding on forest structure. The models were fitted to diameter observations for 1021 Q. branti, 854 Q. infectoria and 1963 Q. libani trees. Diameter distribution of trees was derived from 178 0.1 hectare sample plots. Beta, Gamma, Normal, Log-normal and Weibull probability distribution function were fitted to diameter distribution of oak trees. Actual probability and probability which derived from pdfs for diameter distribution was compared using Kolmogrov-Smironov and Chi-square tests. Fitting showed different results for different species. Since, fitted properly to all species based on both tests, Beta function was chosen as the best probability distribution function. Based on results, beta probability distribution can be used for those who want to simulate changes of pollarded forests. Beside, diameter distribution showed a critical situation for forest regeneration.
