Because of growing wood consumption, pulp and paper demands, plantations of the fast-growing tree species, managed under short rotations, have a mounting significance for the sustainability of industrial wood raw material. Consequently, the efficient utilization of fast growing plantations can have an enormous impact on productivity. The study is based on longitudinal data obtained from Sappi land holdings in coastal Zululand of eastern South Africa. During the first two years of growth, repeated measurements of stem radius were made using dendrometer attached to 18 trees. The objective of the study is to suggest a growth model for two clones of Eucalyptus tree, and compare their growth potential. Two nonlinear growth curves (Logistic and asymptotic regression with an offset) were fitted to the stem radius data. The asymptotic regression model with an offset having random effects for all its parameters appears well suited to represent the data. It was also observed that the asymptotic regression with offset fits better than the logistic regression model. All the three parameters of the logistic and asymptotic regression with an offset were considered as mixed effects. A significant difference was observed between the asymptotes of the two clones. In that GU clone grew faster than the GC clone, indicating better genetic potential for rapid growth.

Keywords: asymptotic regression with an offset, dendrometer trial, Loess, logistic regression, longitudinal study, nonlinear mixed effects.