Contribution of soil moisture to seasonal and annual variations of soil CO2 efflux in a humid cool-temperate oak-birch forest in central Japan

To quantify the contribution of soil moisture to seasonal and annual variations in soil CO2 efflux in a cool-humid deciduous broadleaf forest, we measured soil CO2 efflux during the snow-free seasons of 2005-2008 using an automated chamber technique. This worked much better than manual chambers employing the same steady-state through-flow method. Soil CO2 efflux (g C m(-2) period(-1)) during the snow-free season ranged from 979.8 +/- A 49.0 in 2005 to 1131.2 +/- A 56.6 in 2008 with a coefficient variation of 6.4 % among the 4 years. We established two-parameter (soil temperature and moisture) empirical models, finding that while soil temperature and moisture explained 69-86 % and 10-13 % of the temporal variability, respectively. Soil moisture had the effect of modifying the temporal variability of soil CO2 efflux, particularly during summer and early fall after episodic rainfall events; greater soil moisture enhanced soil CO2 efflux in the surface soil layers. High soil moisture conditions did not suppress soil CO2 efflux, leading to a positive correlation between normalized soil CO2 efflux (ratio of the measured to predicted efflux using a temperature-dependent Q (10) function) and soil moisture. Therefore, enhanced daily soil CO2 efflux following heavy rainfall events could significantly reduce net ecosystem exchange (i.e. daily net ecosystem production) by 32 % on some days. Our results highlight the importance of precisely estimating the response of soil CO2 efflux to changes in soil moisture following rainfall events when modeling seasonal carbon dynamics in response to climate change, even in humid monsoon regions.