You are here

Harvard Forest >

Harvard Forest Symposium Abstract 2006

  • Title: Reconciling landscape-scale measurements of soil respiration with estimates of total ecosystem respiration
  • Primary Author: Eric Davidson (University of Maryland - Center for Environmental Science)
  • Additional Authors: Kathleen Savage (Woods Hole Research Center)
  • Abstract:

    Annual estimates of chamber-based soil respiration (SR) and tower-based total ecosystem respiration (TER) at the Harvard Forest are consistent with a growing body of literature indicating that CO2 efflux from soil represents about 60-80% of TER in most temperate forests. This consistency in the annual C budget, however, masks inconsistencies at higher temporal resolution. At the Harvard Forest, for example, our estimates of summertime SR are often equal to 100% of measured TER and occasionally higher (Figure 1), which obviously cannot be correct. Analytical errors or biases in one or both measurements systems might be indicated, but temporal and spatial mismatches between measurement systems are also plausible explanations. Eddy covariance estimates of TER are made at night, when the footprint of the measurement is often very large, whereas daytime manual measurements of SR are usually made over smaller areas, and nighttime autochamber measurements of SR are limited to a small cluster of chambers. In this study, we conducted a broader survey of daytime manual SR within a 500m radius of the Prospect Hill EMS tower, including the Montauk and Charlton soil series to the north and east of the tower, which we had not previously studied (Figure 2). Average SR rates were 17% and 13% lower, respectively, on these two soil series compared to the mean from the Canton series located near the tower and in the dominant northwest and southwest above-canopy wind directions. Using spatially area-weighted estimates of SR from each of the dominant soil series (Table 1) for the entire area (and ignoring wind direction) the mean annual SR flux is 91% of our previous estimates based primarily on SR measurements on the Canton soil series. Adjusting areal SR estimates for the area occupied by exposed rocks further lowers this value to 88%. Finally, we note that diel variation in autochamber SR measurements on the Canton series is larger than expected based on the application of seasonal Q10 functions to diel temperature variation (Figure 3). Hence, it may be possible that daytime estimates of TER based on nighttime measurements that are adjusted for temperature using seasonal Q10 functions might underestimate daytime estimates of TER. We expect that additional efforts to reconcile SR and TER data will yield new insight into temporal and spatial heterogeneity of respiration processes.

  • Research Category: Forest-Atmosphere Exchange