Comparison of estimation methods for diameter distribution models of Tectona grandis Linn.f. in Omo Forest Reserve, Nigeria

Accurate tree diameter characterisation is fundamental to growth and yield modelling and critical for sound decision-making in sustainable forest management. Despite its importance, there is a paucity of comparative studies on optimal probability density functions and parameter estimation approaches for Tectona grandis stands within the Omo Forest Reserves, Nigeria. This study evaluated the comparative performance of Maximum Likelihood Estimation (MLE) and Methods of Moments Estimation (MME) for diameter characterization using Diameter at breast height 10cm. Data from 658 tree samples aged from 9 to 15 years were used to fit six diameter distribution models, evaluated by log-likelihood (loglik), Akaike information criterion (AIC), Bayesian information criterion (BIC) and Kolmogorov-Smirnov (KS) statistics and graphical appraisal of the resultant probability density function. Results showed that the 3-parameter Weibull distribution consistently provided the best fit across both methods, with Loglik, AIC, BIC, and KS values of -1866.03, 3738.07, 3751.54, and 0.0399, respectively, under MLE. Similarly, under MME, it achieved AIC, BIC and KS statistics of 3747.52, 3760.98 and 0.0506, respectively, outperforming all competing models. Johnson’s SB ranked second under both methods, followed by generalized extreme value and Gumbel distributions, while the logistics and 2-parameter Weibull performed poorly. The stand structure for the pooled age revealed an inverse J-shaped mimicking uneven-aged structure signifying intense competition, with many trees in the smallest diameter classes. Both estimation methods are statistically viable, with MLE demonstrating superior flexibility and robustness. Consequently, the 3-parameter Weibull distribution estimated by MLE is recommended for Tectona grandis diameter characterisation in the study area.