Abstract

It is crucial to advance the understanding of soil CO2 efflux and its components for a better comprehension of carbon dynamics in terrestrial ecosystems. The process-based PATCIS model was applied to a first rotation young Sitka spruce stand in order to simulate the seasonal contribution of soil respiration components to the overall soil CO2 efflux. We evaluated the performance of the model with observed measurements and compared it with empirically derived regressions. Once the model was parameterised, it explained 75% of the seasonal variation in total Soil CO2 efflux. Similar seasonal trends and annual estimates of soil CO2 efflux were obtained with either empirical or the process-based PATCIS models. Heterotrophic and autotrophic respiration contributed almost equally to total CO2 efflux during the early and late part of the year, while a larger contribution of autotrophic respiration to total CO2 efflux Occurred during the growing season. The overall annual contribution of autotrophic respiration to total Soil CO2 efflux was 54.7%. Most of root respiration took place in both the litter humus layer and the A, horizon as a result of their large concentrations of fine roots. We observed an accumulation of organic matter in the litter-humus layer, and a net loss from the mineral soil, which had much larger organic matter content compared to the litter-humus layer. The organic matter turnover rate calculated for the mineral soil was 45 years (mean residence time). The sensitivity analysis showed soil temperature as the most important factor controlling soil respiration. The influence of soil moisture was more variable and had an overall negative effect on soil respiratory rates, except for periods of low soil water content, such as summer drought. The episodic occurrence of very wet conditions at the deeper soil layers was responsible for their low contribution to total soil respiration. In general, gas transport within the soil was not an important constraint for soil CO2 efflux since most of soil respiration was produced in the highly porous litter-humus and top mineral layers. The autotrophic component was more affected than heterotrophic respiration by changes in soil water content. Other factors such as changes in litterfall inputs were shown to have a more limited impact on Soil CO2 efflux. This work shows that the use of a process-based model to simulate soil CO2 efflux may be a useful tool to separate soil respiration components. (c) 2007 Elsevier B.V. All rights reserved.