N2O, CH4 and CO2 emissions from seasonal tropical rainforests and a rubber plantation in Southwest China

The main focus of this study was to evaluate the effects of soil moisture and temperature on temporal variation of N2O, CO2 and CH4 soil-atmosphere exchange at a primary seasonal tropical rainforest (PF) site in Southwest China and to compare these fluxes with fluxes from a secondary forest ( SF) and a rubber plantation ( RP) site. Agroforestry systems, such as rubber plantations, are increasingly replacing primary and secondary forest systems in tropical Southwest China and thus effect the N2O emission in these regions on a landscape level. The mean N2O emission at site PF was 6.0 +/- 0.1 SE mu g N m(-2) h(-1). Fluxes of N2O increased from < 5 mu g N m(-2) h(-1) during dry season conditions to up to 24.5 mu g N m(-2) h(-1) with rewetting of the soil by the onset of first rainfall events. Comparable fluxes of N2O were measured in the SF and RP sites, where mean N2O emissions were 7.3 +/- 0.7 SE mu g N m(-2) h(-1) and 4.1 +/- 0.5 SE mu g N m(-2) h(-1), respectively. The dependency of N2O fluxes on soil moisture levels was demonstrated in a watering experiment, however, artificial rainfall only influenced the timing of N2O emission peaks, not the total amount of N2O emitted. For all sites, significant positive correlations existed between N2O emissions and both soil moisture and soil temperature. Mean CH4 uptake rates were highest at the PF site (-29.5 +/- 0.3 SE mu g C m(-2) h(-1)), slightly lower at the SF site (-25.6 +/- 1.3 SE mu g C m(-2) h(-1)) and lowest for the RP site (-5.7 +/- 0.5 SE mu g C m(-2) h(-1)). At all sites, CH4 uptake rates were negatively correlated with soil moisture, which was also reflected in the lower uptake rates measured in the watering experiment. In contrast to N2O emissions, CH4 uptake did not significantly correlate with soil temperature at the SF and RP sites, and only weakly correlated at the PF site. Over the 2 month measurement period, CO2 emissions at the PF site increased significantly from 50 mg C m(-2) h(-1) up to 100 mg C m(-2) h(-1) (mean value 68.8 +/- 0.8 SE mu g C m(-2) h(-1)), whereas CO2 emissions at the SF and RP site where quite stable and varied only slightly around mean values of 38.0 +/- 1.8 SE mg C m(-2) h(-1) ( SF) and 34.9 +/- 1.1 SE mg C m(-2) h(-1) (RP). A dependency of soil CO2 emissions on changes in soil water content could be demonstrated for all sites, thus, the watering experiment revealed significantly higher CO2 emissions as compared to control chambers. Correlation of CO2 emissions with soil temperature was significant at the PF site, but weak at the SF and not evident at the RP site. Even though we demonstrated that N and C trace gas fluxes significantly varied on subdaily and daily scales, weekly measurements would be sufficient if only the sink/source strength of non-managed tropical forest sites needs to be identified.