Soil CO2 efflux measurements in a mixed forest: Impact of chamber disturbances, spatial variability and seasonal evolution

Abstract

A closed-dynamic-chamber system (CDCS) was used to measure the spatial and seasonal variability of the soil CO2 efflux (Fs) in beech and in Douglas fir patches of the Vielsalm forest (Belgium). First the difference between natural and measured soil CO2 efflux induced by the presence of the CDCS was studied. The impact on the measurements of the pressure difference between the outside (natural condition) and the inside of the chamber was found to be small (0.4%). The influence of wind disturbance in the closed chamber was tested by comparison with an open-chamber system characterized by a different wind distribution. A very good correlation between the two systems was found (r2=0.99) but the open system yielded slightly lower fluxes than the closed one (slope=0.88±0.05). A measurement procedure has been developed to minimize the effect of the other sources of perturbation. The spatial and seasonal evolution of the soil CO2 efflux was obtained by performing regular measurements on 29 spots in the beech patch over a period of 12 months and on 24 spots in the Douglas fir patch over 8 months. For each spot, the experimental relationship between Fs and soil temperature was compared with the fitted line for an Arrhenius relationship with a soil temperature-dependent activation energy. Soil temperature explains 73% of the seasonal variation for all the data. The spatial average of the soil CO2 efflux at 10°C (Fs10) in the beech patch is 2.57 ± 0.41 μmol m-2 s-1, approximately twice the average in the Douglas fir patch recorded at 1.42 ± 0.22 μmol m-2 s-1. The litter fall analysis seems to indicate that soil organic matter quality and quantity may be one the reasons for this difference. Finally the annual soil CO2 efflux was calculated for the beech and Douglas fir patches (870 ± 140 and 438 ± 68 gCm-2y-1, respectively). The beech value would represent 92 ± 15% of the annual ecosystem respiration estimated from the eddy covariance measurements.