Ecology of Iranian Forest

      In this research, modeling tree height distributions of beech in natural forests of Masal  located in Guilan province, was investigated. Inventory was carried out using systematic random sampling with network dimensions of 150×200 m and area sample plot of 0.1 ha. DBH and heights of 630 beech trees in 30 sample plots were measured. Beta, Gamma, Normal, Log-normal and Weibull probability distribution function were surveyed. Characteristics of distribution functions were estimated by use of maximum likelihood method. Actual probability and probability  derived from functions was compared using Kolmogorov-Smirnov and Anderson-Darling tests. Results of Kolmogorov-Smirnov test showed that only Beta probability distribution function was suitable for modeling height distribution of beech; while results of Anderson-Darling tests demonstrated probability distribution functions of Beta, Normal, and Weibull were fitted to the height distribution of trees. Therefore probability distribution function of Beta was selected for modeling height distribution of beech.