Growing stock assessment and growth prediction system for managed hill dipterocarp forest of Peninsular Malaysia

Graduation date: 2002The silviculture of mixed dipterocarp forests of Peninsular Malaysia can potentially be improved by developing quantitative tools and guidelines for this forest type. This thesis addresses three key objectives; 1) to test the hypothesis that a maximum size-density relationship can be identified in hill dipterocarp forests of Peninsular Malaysia, 2) to develop growth equations for predicting annual diameter or basal area increment from initial stand structure and tree conditions, and 3) to develop guidelines for applying the maximum size-density concept to growing stock assessment and control in hill dipterocarp forest. Data from temporary and permanent sample plots were examined to test the applicability of the maximum size-density concept. The maximum size-density limit in mixed dipterocarp forests was related to stand structure as described by the skewness and variance of dbh[superscript m]. Relatively little of the variation in maximum size-density limits was attributable to species composition as represented by different community types. Temporary plots did not allow for the assessment of the maximum size-density limit, at least at the spatial scale investigated. In contrast, data from long-term permanent sample plots indicated that the maximum size-density limit was quite consistent. Several stand trajectories deviated from the limit for short periods, but returned to the same limit after a few years. Silvicultural guidelines were developed by applying the maximum size-density concept a stand density index approach took into account the residual stocking, allocation of growing stock by size class, and species mixture before and after treatment. Appropriate silvicultural strategies involving stand density manipulation require considerable insight into future stand development, especially the competitive effects at both the tree and stand level, and this effort requires quantitative assessment of tree growth through statistical modeling. A mixed-effects model with a random species effect was applied for modeling individual tree dbh increment. This model was compared with classical fixed effects models using data from second growth hill dipterocarp forest. The mixed-effects model offered the advantage that all species could be run simultaneously in one equation, and also allowed variation in slope for each species when an interaction between random species effects and other covariates were included in the model. Diameter growth was a function of tree size, inter-tree competition, and stand structure