Sensitivity of soil CO₂ efflux to environmental conditions regulates how the largest terrestrial source of CO₂ is responding to global climate and land use changes.  Yet, little is known about the sensitivity of in situ soil CO₂ efflux to temperature and moisture in the tropics, and how it is affected by land use change.  This represents a large gap in our understanding of global C cycling, since tropical regions have the highest rates of net primary productivity, soil CO₂ efflux, deforestation and land use- land cover change.  Our study aims to provide valuable information about the controls on soil CO₂ efflux in a tropical forest of Xishuangbanna, southwestern China.  We compared relationships among aboveground litterfall and soil temperature, moisture and CO₂ production in a mature secondary forest and rubber plantation; the dominant land use change in the region.  Litter production, soil temperature and moisture were measured using litter traps, temperature and time domain reflectometry (TDR) probes, respectively.  Soil CO₂ efflux was measured using static field chambers and an infrared gas analyzer.  All measurements were made in 2m x 2m treatment plots within four experimental blocks in each site.  Treatments included aboveground litter exchange and litter exclusion to assess the influence of aboveground litter input on soil CO₂ efflux. 

Results/Conclusions

Our data indicate that replacement of tropical evergreen rainforests with rubber plantations changes patterns of aboveground C input via litterfall.  The forest site exhibited a rather steady litter input rate, with peak production occurring during the wet season.  Litterfall in rubber plantations occurs as a pulse during the cold, dry winter months.  Despite large differences in litterfall patterns, soil respiration was similar in both sites: increasing soil CO₂ efflux was observed during the warm, wet summer months and decreased during the cool, drier months.  Site differences in soil CO₂ efflux were only detected during the cool dry season.  Treatment differences were not detected, likely due to the high spatial variability of soil CO₂ efflux in both sites, and relatively low contribution of heterotrophic litter respiration to overall soil respiration.  Exponential regression models of soil CO₂ efflux and temperature were different between sites, but not strong at either site (R²<0.20). Soil moisture only marginally improved the model for the rubber plantation, but the importance of moisture as a predictor of soil CO₂ efflux is likely due to seasonal dormancy of the rubber trees rather than physiological limitations on soil microbes.