Simulation model of natural tropical forest dynamics

Abstract

Assessments of the long-term natural regeneration of tropical forests following selective logging are today mostly based on extrapolation of limited empirical observations. Mounting evidence suggests that current logging policies overestimate forest regrowth by a wide margin, and are therefore not sustainable. The need for more reliable assessment of natural forest growth dynamics has led to the development of a vertically and spatially structured dynamic simulation model of natural forest development. The model accounts explicitly for biomass and tree numbers in five distinct canopy layers (seedlings, saplings, poles, main canopy, emergents). Energy accounting of assimilation and dissimilation rates leads to biomass, diameter, and height growth in each layer. Leaf mass in each layer determines photoproduction, and light and photosynthesis conditions in lower layers. Transitions to higher layers, mortalities in each layer, and seed events are accounted for. The basic model simulates the development dynamics of a forest gap. Parallel computation of spatially distributed single gap models with stochastic mortalities and seed dispersal provides a simulation of the spatial dynamics of the tropical forest mosaic. Simulation results are provided with parameters corresponding to unlogged and logged Malaysian lowland dipterocarp forest.