Carbon dioxide, water vapor, and energy fluxes were measured using eddy covariance (EC) methodology over three adjacent evergreen forests in southern Washington State to identify stand-level age-effects on ecosystem exchange. The sites represent Douglas-fir forest ecosystems at two contrasting successional stages: old-growth (OG) and early seral (ES). Here we present eddy flux and meteorological data from two early seral stands and the Wind River AmeriFlux old-growth forest during the growing season (March–October) in 2006 and 2007. We show an alternative approach to the usual friction velocity (u*) method for determining periods of adequate atmospheric boundary layer mixing based on the ratio of mean horizontal (U) and vertical (w ) wind flow to a modified turbulent kinetic energy scale (uTKE). This new parameter in addition to footprint modeling showed that daytime CO2 fluxes (FNEE) in small clear-cuts (< 10 hectares) can be measured accurately with EC if micrometeorological conditions are carefully evaluated.
Peak midday CO2 fluxes (FNEE = – 14.0 to -12.3 µmol m-2 s-1) at OG were measured in April in both 2006 and 2007 before bud break when air and soil temperatures and vapor pressure deficit were relatively low, and soil moisture and light levels were favorable for photosynthesis. At the early seral stands, peak midday CO2 fluxes (FNEE = – 11.0 to -8.7 µmol m-2 s-1) were measured in June and July while spring-time CO2 fluxes were much smaller (FNEE ¬= – 3.8 to -3.6 µmol m-2 s-1). Overall, we measured lower evapotranspiration (OG = 230 mm; ES = 297 mm), higher midday FNEE (OG FNEE = -9.0 µmol m-2 s-1; ES FNEE = -7.3 µmol m-2 s-1) and higher Bowen ratios (OG β = 2.0; ES β = 1.2) at the old-growth forest than at the ES sites during the summer months (May–August). Eddy covariance studies such as ours add critical land-atmosphere exchange data for an abundant, but rarely studied Douglas-fir age class.
