To determine the means and variations in CH4 uptake and N2O emission in the dominant soil and vegetation types to enable estimation of annual gases fluxes in the forest land of Japan, we measured monthly fluxes of both gases using a closed‐chamber technique at 26 sites throughout Japan over 2 years. No clear seasonal changes in CH4 uptake rates were observed at most sites. N2O emission was mostly low throughout the year, but was higher in summer at most sites. The annual mean rates of CH4 uptake and N2O emission (all sites combined) were 66 (2.9–175) µg CH4‐C m−2 h−1 and 1.88 (0.17–12.5) µg N2O‐N m−2 h−1, respectively. Annual changes in these fluxes over the 2 years were small. Significant differences in CH4 uptake were found among soil types (P < 0.05). The mean CH4 uptake rates (µg CH4‐C m−2 h−1) were as follows: Black soil (95 ± 39, mean ± standard deviation [SD]) > Brown forest soil (60 ± 27) ≥ other soils (20 ± 24). N2O emission rates differed significantly among vegetation types (P < 0.05). The mean N2O emission rates (µg N2O‐N m−2 h−1) were as follows: Japanese cedar (4.0 ± 2.3) ≥ Japanese cypress (2.6 ± 3.4) > hardwoods (0.8 ± 2.2) = other conifers (0.7 ± 1.4). The CH4 uptake rates in Japanese temperate forests were relatively higher than those in Europe and the USA (11–43 µg CH4‐C m−2 h−1), and the N2O emission rates in Japan were lower than those reported for temperate forests (0.23–252 µg N2O‐N m−2 h−1). Using land area data of vegetation cover and soil distribution, the amount of annual CH4 uptake and N2O emission in the Japanese forest land was estimated to be 124 Gg CH4‐C year−1 with 39% uncertainty and 3.3 Gg N2O‐N year−1 with 76% uncertainty, respectively.

[1] CO 2 efflux in the period of snow cover can be a large carbon source in the yearly carbon budget of snowy ecosystems. However, the behavior of CO 2 in snowpacks and the mechanisms of the snow surface efflux are still unclear. We performed continuous (half-hourly) midwinter measurements of CO 2 concentrations in a conifer-broadleaf mixed forest snowpack, and found that concentrations in the snowpack fluctuated significantly as wind speeds varied. The snow surface efflux was evaluated as the sum of the CO 2 storage change in the snowpack and the CO 2 input from the soil to the snowpack, taking into account the mixing due to airflow. The median value over 52 days (49 mmol m À2 d À1 ) was almost the same as the daily net ecosystem exchange rate in this forest (50 mmol m À2 d À1 ) estimated by the eddy covariance technique and the storage-change flux in the air column. These values are clearly larger than the value we estimated using Fick's law of diffusion. These results show that airflow can be a dominant cause of mixing within snowpacks in midwinter. In addition, in the soil pores under the snowpack, the CO 2 concentration was primarily related to air temperature, implying that soil respiration responds directly to air temperature, not to soil temperature, even beneath a 1-m-thick snowpack. We infer that the air temperature affected the root activity of trees through their trunks and that the variation in root respiration strongly affected the CO 2 concentration fluctuation in soil under the snowpack.

We determined the interannual variation of annual litterfall rate in cool-temperate forests [three mixed-forests (Mx1-Mx3), an evergreen coniferous forest (Ec), and a deciduous conifer plantation (Dc)] of northern Hokkaido over a 16-year observation period (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011) and evaluated the effect of meteorological and phenological variables on the annual litterfall production. Total solar radiation during spring (from March to May) positively correlated with the annual litterfall rate in the current year at three mixed forests with statistical significance. A warm spring advanced the day of snow melt and the day of leaf expansion, however, the early leaf expansion did not enhance the annual litter production at any of the studied forests. In conclusion, spring solar radiation was the best explanatory factor among the studied factors that determines the interannual variation of the annual litterfall rate at cool-temperate mixed forests, although the mechanisms behind this relationship remain unknown. The early snowmelt and leaf expansion caused by a warm spring did not directly link to the enhancement of the litterfall rate. This implies that global warming or changing rainfall patterns do not necessarily affect the annual litterfall amount in these forests, at least within the range observed during the 16 years.