Modelling the management of forest ecosystems: importance of wood decomposition

Scarce and uncertain data on woody debris decomposition rates are available for calibrating forest ecosystem models, owing to the difficulty of their empirical estimations. Using field data from three experimental sites which are part of the North American Long-Term Soil Productivity (LTSP) Study in south-eastern British Columbia (Canada), we developed probability distributions of standard wood stake mass loss of Populus tremuloides and Pinus contorta. Using a Monte Carlo approach, 50 synthetic decomposition rate values per debris type were used to calibrate the ecosystem-level forest model FORECAST. Significant effects of uncertainty of pine stake mass loss rates on estimated tree growth were found, especially in moderately managed forests, as estimations of available nitrogen were affected. Consequently, our work has shown that projections of tree growth under management conditions depend on accurate estimations of woody debris decomposition rates, and special effort should be done in create reliable databases of decomposition rates for their use in tree growth and yield modelling