By: Ige P. O.; ADESOYE P. O.;
Abstract:
Non-spatial explicit inter-tree competition indices were evaluated to determine their efficiency in stem growth prediction of Gmelina arborea trees in Omo Forest Reserve. The aim was to quantify and describe the neighbouring effects on the tree diameter and basal area growth. Data were collected from four age series of G. arborea stands (32, 26, 20 and 16 years old). Stratified random sampling method was used in this study to select ten sample plots which were allocated in each stand age. Each sample plot was 20m x 20m (i.e 0.04ha) in size. Diameters at breast height of all trees were measured. Seven non-spatial competition indices were separately devised into growth models as a predictor variable to assess the ability of the diameter and basal area growth models before and after adding competition measures. The fitness of each model to the data was assessed on the basis of adjusted coefficient of determination (AdjR2), akaike information criterion (AIC) and standard error of estimate values (SE). All the models were then compared for predictive ability using an independent validation data set. The result revealed that inter tree competition proved to be a major factor contributing to individual tree growth. For diameter growth model, model with competition index computed as the sum of ratio of dbh of neighbors to the subject tree dbh in a plot (CI3) had the highest AdjR2 (98.97%), smallest AIC (15.1272) and least SE (0.0021) and the greatest prediction precision (0.0431%). For basal area growth, growth model with competition index which was computed with the ratio of sum of the basal area of trees larger than the subject tree to the relative spacing index (CI7) seems more appropriate as a predictor variable based on its highest AdjR2 (98.95%), smallest AIC (14.0328) and least SE (0.0001) and the greatest prediction precision (0.0210). In addition, these competition indices can be used due to the fact that they can be determined easily and inexpensively during field data collection.
Keywords: Non-spatial, Inter tree competition, Growth models, Gmelina arborea
Click here to read and download paper