Abstract
Forests play an important role in terrestrial carbon cycling processes, acting as sinks for the
methane (CH4) and carbon dioxide (CO2). Process-based ecosystem models are used to estimate
the exchange of these greenhouse gas fluxes. Soil microorganisms play a crucial role in regulating
these fluxes, but microbial parameters are often not estimated for the site being modeled. In this
study, we parameterized the specific oxidation rate (µmax) and half saturation constant (Km) related
to aerobic heterotroph, methanotroph, and nitrifier populations for a forested watershed in
Huntington Wildlife Forest, Newcomb, NY. We trained an ecosystem-level model, ecosys, for this
purpose and incorporated site-specific measurements of trace gas fluxes (CH4, CO2) and microbial
abundance (ammonia oxidizers) into Bayesian analysis using a Markov Chain Monte Carlo
(MCMC) Metropolis-Hastings algorithm. We found a reduction of model bias for CO2 and CH4
flux by 9.38% and 2.04% respectively. Our analysis also provided information about the
relationships between nutrient cycling processes and the uncertainties involved in model
projections. We used our parameterization to project heterotrophic respiration, CH4 flux exchange
and net biome productivity under future climate change scenarios with elevated temperature and
CO2 concentration. We found CH4 uptake to be nearly the same as present with high uncertainty
range, while heterotrophic respiration decreased by 0.373 gC m-2 year-1
and net biome productivity
increased by 0.59 gC m-2
year-1
.