Modeling carbon accumulation dynamics in tropical peat swamp forests

Abstract

Tropical peatlands provide a number of ecosystem services – including biodiversity, habitat, carbon and water cycling, and commodity products – and have among the highest carbon densities of any forest on earth. Almost half of tropical peatlands supporting the growth of peat swamp forests are in Indonesia. Rates of deforestation there and hence C emissions are very high, but few studies have examined C dynamics of these forests. We present the initial results of the development of a model of peat accumulation in tropical swamp forests over millennia. In this study, the Holocene Peat Model (HPM), which has been successfully applied to northern peatlands, was modified for tropical ecosystems. HPMTrop is a one-dimensional, non-linear, dynamic model with a monthly time step that simulates mass remaining in annual peat cohorts as a balance between vegetation inputs and decomposition. We utilized as model parameters published data on vegetation characteristics, including net primary production (NPP); NPP partitioning into leaves, wood, and roots; and litter decomposition rates. A stochastic mix of wet and dry years controlled the peat swamp water table depth. Over 8,000 years, HPMTrop simulated a peat accumulation of 3.9 m, which is equivalent to a net accumulation of about 2,200 Mg Cha-1. At the end of the simulation, 55% of the accumulated peat carbon was derived from wood inputs, 31% from roots, and 14% from leaves. In a modeled scenario of dominance by non-tree vegetation (sedge), peat accumulation was dramatically lower; only about 1 m (about 550 Mg Cha-1) over 8,000 years. These modeled C accumulation rates will be compared with peat accumulation estimated of peat cores collected from Berbak National Park, Jambi, Indonesia, and Tanjung Puting National Park, Kalimantan, Indonesia, using radiocarbon dating.

Department

Earth Sciences, Earth Systems Research Center

Publication Date

2013

Document Type

Conference Proceeding

Share

COinS