Investigation of sampling methods for modeling of diameter distribution of Quercus persica trees in the Zagros forests

The aim of this research was to determine an optimum sampling method based on best-fit probability distribution functions for modeling diameter distribution of Quercus persica in Dalab forests in Ilam province (western Iran). A total area of 37.2 hectares was full callipered in this study. Also, 37 rectangular 1000 m² plots were sampled using both transect method, with a fixed length of 50 m, and systematic random sampling. Beta, Gamma, Exponential, Normal, Lognormal and Weibull probability distribution functions were fitted to diameter at breast height (DBH) distribution of the oak trees. The expected probability and probability derived from the above functions were compared using Kolmogorov-Smirnov and Anderson-Darling tests. The Kolmogorov-Smirnov analysis in census method showed that the applied probability distribution functions are incapable of fitting DBH distribution of the oak trees. Anderson-Darling test in our study showed that the Beta probability distribution function most appropriately fit with DBH distribution. In the systematic random sampling, the results of Kolmogorov-Smirnov showed that Weibull distribution is the most suitable function compared to other probability distribution functions. Though, Anderson-Darling test indicated that the Gamma, Weibull, Beta and Lognormal distributions are appropriate for fitting DBH distribution in a descending order. In transect sampling method, the results of Kolmogorov-Smirnov showed that the investigated probability distribution functions are unsuitable for fitting DBH distribution. According to Anderson-Darling test, the Gamma, Lognormal, Normal and Weibull are, respectively, the most applicable distribution functions in explaining the oak trees DBH distribution.