MODELLING THE GROWTH OF THE LOGGED OVER INLAND PRODUCTION FORESTS OF PENINSULAR MALAYSIA WITH FOREST MANAGEMENT CONSIDERATION

The study objectives are (1) to develop growth models for inland tropical rain forests and (2) to propose changes to the current management systems and cutting regime options. In the Malaysia context inland forests encompass lowland, coastal hill, hill and upper hill forests, 'flic current management systems reviewed are those practised in Peninsular Malaysia. Three fundamental aspects of forest management are inventory, growth prediction and an understanding of the forest silviculture or ecology. Tropical rain forest contains the largest percentage of the world is terrestrial biodiversity and its ecosystem is one of the most complex, but studies on the growth of this forest is very much less compared to the temperate and plantation forests, To ensure sustainable forest management the forest silviculture has to be better understood and the growth prediction models used arc of a dynamic nature to cope with the stand content variations. High variability in stand content, even within site, means that as many growth models as possible should be developed. Hence a range of alternative models should be available to forest managers for adequate choice suited to different sites and constraints. The study contained three aspects, (1) species grouping, (2) model development and testing, and (.3) growth projection and review of management systems. Three sites in Peninsular Malaysia were used in this study, and the data used was 6 8 year growth period. Each study site was stratified into four stocking classes; superior, good, moderate and poor. Each stocking class was treated with a range of different cutting intensities. After felling, every tree in the study plot is recorded, measured and evaluated al a regular interval, free sizes measured were from seedling to big trees, hi this study only pole size tree and above were considered. Two sites were used for model development and model verification, and one site was used for model validation, 'The initial stands of the study sites were untouched/virgin forest. Species grouping were based on (be tree silvicultural and wood characteristics; species canopy status (emergent and main-emergent or A stratum, main or B stratum and understorey or C, D and E strata), mature tree diameter size (very large, large. medium and small) and wooti characteristics (light heavy hardwood, medium heavy hardwood anti heavy hardwood). These are stantlartl characteristics used in Malaysia to describe species physical characteristics, and the silvicultural characteristics are in lact standard practice in forestry worldwide. More than 400 species identifictl in this study were grouped by all the characteristics or one of the characteristics, flic groupings categorised trees with similar silvicultural characteristics and generally of the same potential growth rate. The species groups were then used in the growth modelling. Rotation period projection was made using the general diameter increment anti mortality rates that were derived in this study. Two types of rotation periods were considered; targeted diameter size based and basal area content based. Another type of rotation period considered was canopy species composition anti stand vertical structure based. The latest method is most ecologically sound, in this study it is termed stand maturity benchmark. Based on the projected growth it is concluded that under the current management practice (Selective Management System anti Malayan Uniform System) the forest is not managed to its ecological maturity. To manage to the ecological maturity cycle for the Selective Management System a 70 years rotation period is proposed anti for the Malayan Uniform System not less than 100 years rotation period is recommended. Type of growth models developed were diameter growth, ingrowth and models. Models were formulated by stem sizes (small pole, big pole, small tree and big tree) and species groups. Models were developed for different forest management levels: general planning and management at the compartment level (smallest management unit). Basically the developed models can be categorised into (1) general growth rate (fixed growth rates), (2) general growth model (prediction models irrespective of stand variations) and (3) growth model by stand condition (prediction models taking into account the site anti stand variations). The model type used to model growth was a stand development, transition based model, which is, in principle, similar to the state-transition model and succession model. Canopy species ratio was used to express the stand development stages. Regression method was used to formulate the diameter growth models and logistic regression was used to formulate the mortality and ingrowth. All growth models produced are empirical in nature. The study has successfully produced diameter growth, mortality and ingrowth models by stem sizes and species groups. It is concluded that the models developed in this study arc only applicable at the compartment level - the smallest area. Only the study arc only applicable at the compartment level - the smallest area. Only the diameter growth model, where its predictor is the previous growth rate, has the potential to predict growth to the quadrat level, even to the individual tree level. The study has also formulated a method of calculating the mean annual growth using short-term growth data. Another achievement was the formulation of stand maturity benchmark, a standard measurement of measuring stand ecological maturity. 'I he species grouping procedure formulated in this study offers a practical approach grouping species into similar silvicultural characteristics and potential growth rate. Lack of long term growth data is common in tropical countries, and consideration of data more representation of sites is more critical, lienee methodology using short-term growth data (which was used in this study) for growth study and modelling will be a great contribution to practical growth prediction for tropical forest countries.

Name dc:type.qualificationname

PhD

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

De Montfort University

Year dc:date.issued

2000

Author dc:creator

• ZAKARIA, MOHD YUNUS

Licence

• https://dora.dmu.ac.uk/bitstreams/ff815b65-e7f9-4aea-869f-0ab63b83b93b/download