Wood decomposition model for boreal forests

Abstract

The decomposition of woody litter is a biochemical process, controlled by physical, chemical and biological environmental conditions. To develop a practicable model of this complex process, it is necessary to identify the major controlling factors and quantify their effects. We used four data sets (total N= 2102) on mass loss of decomposing woody litter in Northern Europe to extend an earlier decomposition model of non-woody litter and make it suitable for describing also decomposition of woody litter. We compared alternative ways to model the effects of size and chemical composition of woody litter on decomposition using the Bayesian model selection theory. The best model fitted to the diverse data sets (woody litter of four tree species, diameter 0.5-60 cm, time series up to 70 years) with little systematic error. Based on this result, we concluded that the extended model is suitable for describing decomposition of woody litter of the common tree species in the boreal forests studied. According to the model developed, frequently observed sigmoidal patterns in mass loss of woody litter are related to high concentrations of slowly decomposing lignin compounds and these patterns become stronger with an increasing size of decomposing woody litter.