Soil respiration (Rs), including autotrophic respiration from roots (Rr) and heterotrophic respiration (Rh) from microbes and soil fauna, is a key process that regulates the carbon balance in forest ecosystem. Ecosystem warming experiments, modeling analyses and fundamental biokinetics all suggest that Rs should change with elevated temperature following recent global climate change. The Meta analysis of worldwide Rs observations (1,434 points data) by Bond-Lambety and Thomson (2010) found the air temperature anomaly (the deviation from the 1961-1990 mean) was significantly and positively correlated with changes in Rs, and estimated an increase in global Rs by 0.1 Pg C yr^-1 between 1989 and 2008. This suggests the recent climate changes might have change the carbon balance of terrestrial ecosystems. The study site of this study, the Takayama flux site, is the oldest forest site of the AsiaFlux network. The long-term continuous monitoring of soil respiration began in 1994, enabling analysis of changes in soil respiration with temperature and other associated factors. The aim of our study was to present multiyear measurements of chamber-based soil respiration over 13 years from 1994 to 2010 (lack data of 1996-1998) and to determine the contribution of temperature anomaly to interannual variability in soil respiration in temperate deciduous forest ecosystem.

The study site was located at central Japan (36^o08’N, 137^o25’E, 1420 m a.s.l) with a cool temperate climate under the influence of the Asian monsoon. The secondary deciduous broad-leaved forest is established on a brown forest soil (Dustric Cambisol), and is primarily dominated by oak (Quercus crispula) and birch (Betula ermanii and Betula platyphylla), with a canopy height of 15 to 20 m. The forest floor is covered with a dense dwarf bamboo (Sasa senanesis) community. The ground surface is covered with snow from December to April. The soil respiration was measured every two weeks with manual chambers based on open-flow IRGA method from 1994 to 2005 (except 1996-1998). The CO₂ efflux from snow surface was also measured. We improved an automatic chamber technique based on the same open-flow IRGA method, and measured soil CO₂ efflux continuously over snow-free seasons since 2005.

Soil temperature exerted principle control on the seasonal and annual variability of soil CO₂ efflux (e.g. explained 86.4% of variability in 2005). Soil moisture modified the temporal variability of efflux especially during summer and early fall. Furthermore, soil moisture enhanced the efflux may be explained by the increase of CO₂ production in topsoil layer. The estimated annual soil respiration with and without considering the effects of soil moisture, showed less than 2% deviation to the measured cumulative efflux if using the empirical models derived from each target year. However, if using the model derived from the two investigated years, the deviation would increase to more than 8%. The temperature-associated changes in annual soil respiration were examined with the relationships of air temperature anomaly (the deviation from the 1961-1990 mean), annual litterfall and annual net ecosystem productivity by the eddy covariance method.