Abstract

A comparison was made of annual net ecosystem productivity (NEP) of a closed canopy Sitka spruce forest over 2 years, using either eddy covariance or inventory techniques. Estimates for annual net uptake of carbon (C) by the forest varied between 7.30 and 11.44 t C ha(-1) year(-1) using ecological inventory (NEPeco) measures and 7.69-9.44 t C ha(-1) year(-1) using eddy covariance-based NEP (-NEE) assessments. These differences were not significant due to uncertainties and errors associated with estimates of biomass increment (15-21%) and heterotrophic respiration (12-19%). Carbon-stock change inventory (NEP Delta C ) values were significantly higher (27-32%), when compared to both NEPeco- and -NEE-based estimates. Additional analyses of the data obtained from this study, together with published data, suggest that there was a systematic overestimation of NEP Delta C -based assessments due to unaccounted decomposition processes and uncertainties in the estimation of soil-C stock changes. In contrast, there was no systematic difference between NEPeco and eddy covariance assessments across a wide range of forest types and geographical locations.