Modelling growth of a tropical rain forest in East Kalimantan, Indonesia
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
A study on modelling growth of individual trees in a tropical rain forest in East Kalimantan of Indonesia using PT. ITCI and PT. INHUTANI I data was conducted with the main aims being to provide a tool for predicting growth and yield, and to offer recommendations to improve the planning of timber harvests and management of the forests. Individual tree-based distance-independent modelling was the approach used to predict diameter growth of trees in the above forests. This approach was considered to be more applicable to selection cutting and planting system called TPTI (Tebang Pilih Tanam Indonesia), the silvicultural system applied for achieving management goals at harvest in tropical rain forests in Indonesia. Various ways for grouping species (using maximum attainable size and growth characteristics) were examined and different functional forms (linear models, probabilistic and modified beta functions, asymptotic nonlinear equations) were tested through several steps using PT. ITCI data. An empirical approach using the above characteristics was found to offer a useful way to aggregate species for PT. ITCI data. A modified form of the Gompertz projection equation which incorporated stand attributes and locality factors proved to be the best model among functional forms tested in this study. The modified Gompertz projection form was then used to model PT. INHUTANI I and the combined PT. ITCI/PT. INHUTANI I data, using the same criteria for species aggregation as for the PT. ITCI data (maximum attainable size and growth characteristics). The overall best performance shown by the Gompertz projection equation (an asymptotic nonlinear equation) among the functional forms tested for PT. ITCI data, was the major contribution from this study to individual tree-based growth modelling research in tropical rain forests. Individual tree-based modelling in this type of forest, for reasons unknown has traditionally relied heavily on the use of linear models. Outcomes from modelling PT. ITCI, PT. INHUTANI I, and the combined PT. ITCI/PT. INHUTANI I data, provide useful insights into what further research is needed in modelling tropical rain forests in these two localities. Although the growth models developed in this study still require further improvements, these models offer a useful guide for improving silvicultural prescriptions which are currently based on an assumed tree diameter growth rate of 1 cm/year for commercial trees.