Development of a spatial terrestrial nitrogen model for application to Douglas-fir forest ecosystems

Though field data indicate that biogeochemical properties are heterogeneous, biogeochemical models are often applied without regard to local variation. In this paper, we examine how the spatial scale used to represent heterogeneity in ecosystem modeling affects aggregated forest stand nitrogen (N) dynamics. We conduct this spatial analysis using measured local variation in forest carbon and nitrogen distributions. By coupling field and modeling approaches, we address spatial scaling in the context of natural ecosystem dynamics. We model spatially explicit N dynamics in Douglas-fir (Pseudotsuga menziesii) forests using reaction-diffusion equations, in which Michaelis-Menten functions represent nitrogen transfer between plant, microbial, and soil pools, and nitrogen transport is approximated using diffusion. A stability analysis of the reaction terms in the absence of diffusion shows the model is very resilient to perturbation; consequently limit-cycle or oscillatory dynamics are not observed. The model is sensitive to variation in the maximum rate of plant nitrogen uptake as well as the Michaelis-Menten half-saturation constant for plant N uptake. The spatial scale of modeled heterogeneity does not significantly affect aggregated N dynamics in the Douglas-fir system; spatial resolution mainly affects the variance of stand nitrogen values. Analysis of post-perturbation dynamics shows that stand productivity is significantly reduced under frequent perturbations, especially when microbial biomass pools are seriously degraded. (C) 2005 Elsevier B.V. All rights reserved.