Comparison of eddy covariance, chamber, and gradient methods of measuring soil CO2 efflux in an annual semi-arid grass, Bromus tectorum

Eddy covariance measures net ecosystem exchange of CO2 (NEE) at a scale between chamber-based measurements of CO2 exchange processes and large-scale models of CO2 flux dynamics. As the intermediate, it represents a link between small and large-scale estimates of NEE. Accuracy is therefore critical. However, estimates of nighttime ecosystem respiration based on scaled-up measurements of soil and leaf CO2 exchange are most often larger than from eddy covariance. Identifying the source of the discrepancy is difficult due to large measurement uncertainties associated with high variability of fluxes in complex ecosystems. This study compared measurements in a simple system that allowed for minimal uncertainty. We compared measurements of soil efflux using (1) soil chambers, (2) the soil CO2 gradient technique and ecosystem respiration using the (3) the eddy covariance method from a surface that was covered with living vegetation, straw, and snow in turn through a year. Results showed general agreement among measurements in a range of conditions during canopy absence indicating that each measurement technique is theoretically sound. However, we found disagreement among measurements in specific conditions that indicated certain limitations with each method. Nighttime eddy covariance measurements of ecosystem respiration were below the uncertainty limits of soil respiration measurements during the period of active canopy growth (leaf area index from approx. 0.3–1.0 m2 m−2. This raises questions about the accuracy of nocturnal eddy covariance measurements over more complex surfaces. There was indication that the chamber method estimates were unrepresentative of the footprint in certain conditions due to within collar surface treatment and due to collar interaction with the environment. Lastly, the gradient method failed to represent surface fluxes during summer rain. To measure soil efflux in all conditions typical of this site, a combination of all three methods is recommended. A combined NEE estimate for 2005 for soil efflux was 406 ± 73 g C m−2.