Major contribution of grass roots to soil carbon pools and CO2 fluxes in a mesic savanna.

Aims: There is a trend of increasing woody biomass in tropical savannas. Here we ask what effect this increase may have on soil carbon pools and fluxes. Methods: Using a field experiment we determine the amount of soil carbon directly under grasses, a juvenile tree among grasses and a juvenile tree with no grasses. We also measure CO2 efflux at the soil surface and use gas wells to extract CO2 from several soil depths. Results: Our results show that grasses contribute substantially more than trees to both soil carbon and soil respiration. Grasses also make a disproportionate contribution to the δ13C value of SOC in the tree only treatments. The surface gas efflux data show that soil respiration increased with an increase in volumetric soil moisture and temperature and plots with both grasses and trees had higher respiration rates than plots with trees only or with grasses only. Conclusions: The highest soil respiration is in the top 20 cm of the soil with grasses the primary contributors to both δ13CSOC and δ13CR. Any increase in woody biomass will result in a decline in SOM turnover and nitrogen mineralization rates resulting in higher SOC pools. The associated increases in SOC and above ground biomass will however be associated with negative economic and biodiversity impacts. [ABSTRACT FROM AUTHOR]