Simulation modeling of nitrogen trace gas emissions along an age gradient of tropical forest soils

Abstract

Applications of a process simulation model for ecosystem production and soil nutrient cycling were carried out to test hypotheses concerning the controls on nitrogen trace gas fluxes in tropical forests. Assuming that emissions of trace gases (N 2O and NO) can help characterize patterns in N cycling among forests of different age and nutrient status, we applied a new daily time step version of the CASA (Carnegie-Ames-Stanford approach) model in an attempt to reproduce soil biophysical conditions previously measured in three forest stands that comprise a soil age gradient (ranging from 200 to 185 000 years old) on the island of Hawaii. We compared model-predicted soil moisture, temperature, N mineralization, and N trace gas emission rates to measurements at forest sites made during 1990 and 1991. Results showed that predicted water filled pore space (WFPS) in the soil is consistently lower than measured WFPS, possibly due to incomplete understanding of moisture holding capacity, drainage properties, and small scale variability of these volcanic soils. Simulations correctly predicted the observed difference of an order of magnitude between N 2O emission fluxes in the youngest and the oldest forest soils. Nevertheless, specific day-by-day comparison of predicted and measured N 2O fluxes reveal only occasional agreement within an order of magnitude tolerance level. Although soil moisture conditions appear favorable for emission on NO at these sites, as predicted by the generalized model design, lack of observed NO emissions from the previous field and laboratory studies suggest that: (1) our hypothesized NO emission levels as a function of WFPS should be truncated at lower levels of soil moisture availability; and/or (2) nitrification potentials of natural forest soils should be considered in revised versions of the model.