Allometric equations considering the influence of hollow trees: A case study for tropical peat swamp forest in Sarawak

Biomass estimations in tropical peat swamp forests are quite complex when hollow trees are frequently found due to the unavailability of data on hollow size and the limited data on accurate measures of biomass. Destructive samplings were done for both above(AGB) and belowground biomass (BGB) and hollow sizes of remained trees at a logged-over peat swamp forest in Sarawak, Malaysia. Subsequently, allometric equations taking hollows into account for both the aboveand belowground biomass of tropical peat swamp forests were also being developed. It was observed that these were hollows in Shorea albida and Combretocarpus rotundatus trees with diameters at breast height (DBH) exceeding 40 cm; S. albida is a dominant or co-dominant species, and C. rotundatus grows in peat swamp forests throughout Sarawak. The hollow volumes ranged from 0.23 to 1.08 m, and occupied 42.3% of stem volume on average. The larger biomass produced by previous allometric models were partially due to the presence of hollows. Thus, new models for estimating both AGB and BGB were developed that included one (only DBH), two (DBH and height [H] or wood density [WD]), or three (DBH, H, and WD) predictor variables, and [ln(DBH)] was added as predictor variable indicating the biomass loss by cavity formation. AGB model with three predictor variables and BGB model with one predictor variable performed the best where; they had the highest adjusted coefficients of determination and lowest Furnival index and Akaike information criterion (AIC).