Crown ratio modeling of Populus deltoids trees in plantation of Guilan province

The tree crown component is one of the decisions making tools in units of forest management and commonly used in growth and yield models. One of the indices in related to tree crown in crown ratio which is the ratio of live crown length to the total height of the tree. The purpose of this study was to crown ratio modeling of poplar trees (Populus deltoides) using linear and nonlinear regressions (exponential, logistic and Chapman-Richard) in plantations of Guilan province.

For this purpose, the data related to 660 poplar trees including DBH, collar diameter, total height, crown length, crown width, crown area, crown diameter, basal area, tree slenderness coefficient, stem height and crown ratio were measured and recorded. To evaluate and compare different models were used coefficient of determination (R2), paired-samples t-test and four accuracy statistics including ME, RMSE, MEr and RMSEr.

The coefficient of determination of linear and nonlinear models showed that all models have acceptable coefficient of determination. Also, the results of accuracy statistics showed that the exponential and linear models were more accurate than the other models. But the results of mean comparison test indicated there was no significant difference between of measured mean with estimated mean of crown ratio derived from exponential model. While there was significant difference between of measured mean of crown ratio with estimated mean of linear, logistic and Chapman-Richard models. Regarding to the results of precision and accuracy of studied models, exponential model was selected as the more suitable model for estimation of crown ratio of poplar trees in plantations of Guilan province.

Based on the obtained results, it can be suggested to use the crown ratio index in order to select a suitable species for wood cultivation.