Long term carbon dioxide exchange above a mixed forest in the Belgian Ardennes

Eddy co-variance measurements of water vapour and CO2 fluxes were performed for 15 months at the Vielsalm experimental site (Belgian Ardennes). The forest is patchy, made up of mixed coniferous (Douglas fir) and deciduous (beech) plots. The experiment was developed as a part of the EUROFLUX and ECOCRAFT projects. The experimental set-up has been shown to provide reliable flux estimates provided that an adequate frequency correction is applied. A simple footprint model was used in order to separate the contributions to the flux of the conifer and the beech sub-plots. Saturation net ecosystem exchange decreased with increasing air saturation deficit and with decreasing temperature below 15 degreesC. Dark respiration and quantum efficiency decreased with decreasing soil water content and, below 15 degreesC, with decreasing temperature. These responses explain most of the seasonal behaviour of the CO2 fluxes. Generally, the differences between the beech and the coniferous sub-plots were low during the full vegetation period. The differences were significant in early spring and late autumn when the leaves in the beech sub-plot were not fully developed or senescent. A relation between night CO2 flux (N-e) and temperature was established. During periods of little turbulent mixing at night, there was a considerable scatter in the plot of N-e versus temperature; however, after selection of the data using an appropriate turbulence criterion, a well-defined relationship between N-e and temperature was established. An optimal criterion was found to be u* > 0.4-0.5 ms(-1). No clear difference in N-e between the conifer and the beech sub-plots was detected, probably because of bad footprint conditions at night. A first estimate of the annual carbon sequestration was made by summing the half-hourly N-e and filling in the data gaps resulting from system failures or from removal of data that did not fit the quality criteria. It was found to be -0.60 kg m(-2) per year. (C) 2001 Elsevier Science B.V. All rights reserved.