Comparing chamber and eddy covariance estimates of ecosystem respiration during bark beetle mortality in a subalpine spruce-fir forest

Background/Question/Methods

Bark beetle epidemics have impacted millions of hectares of forests in western North America, with many stands experiencing tree mortality exceeding 90% of stand basal area.  Models predict that the loss of nearly all of the live tree biomass will decrease net ecosystem exchange (NEE) of carbon by decreasing photosynthesis and increasing heterotrophic respiration from the decomposition of newly produced fine litter (dead needles and fine roots) and dead wood (stems and branches and coarse roots).  Declines in photosynthesis should scale with loss of leaf area, but because ecosystem respiration is derived from many autotrophic and heterotrophic sources, the response of specific components and the net response  of ecosystem respiration is poorly understood.  We examined the response  of components of the ecosystem carbon balance to a spruce beetle outbreak that killed over 80% of the tree basal area from 2005 to 2011 at the Glacier Lakes Ecosystem Experiment Site (Ameriflux) near Centennial, Wyoming.  We used eddy covariance (EC) measurements to assess changes in annual NEE as leaf area was lost to mortality and compared eddy flux estimates of nightly respiration with the soil, foliage and woody  components of ecosystem respiration measured with chambers.

Results/Conclusions

Annual NEE estimated by eddy covariance declined from  ~ 300 g C m^-2 yr^-1 in 2005 (carbon sink) to ~-100 g C m^-2 yr^-1 in 2011 (carbon source).  Ecosystem respiration, estimated by EC plus changes in atmospheric carbon storage during turbulent nights (u* > 0.2 m s^-1) showed no effect from the loss of > 80% of the overstory foliage and was similar from 2005-2011.  Average soil respiration, measured with chambers, did not vary from 2005-2011 (3.8 µmol m^-2 s^-1).  Chamber measurements of foliage respiration at night declined with the loss of leaf area from an average of ~2.8 to 1.2 µmol m^-2 s^-1 from 2005 to 2011.  Chamber measurements of 24-hour wood CO₂ efflux declined with the loss of live trees from an average of ~0.7 to 0.4 µmol m^-2 s^-1 from 2005 to 2011. Summed estimates of chamber respiration as an estimate of ecosystem respiration declined with tree mortality from 2005-2011 and were substantially greater than night EC measurements of NEE on turbulent nights.  The lack of a response of EC ecosystem respiration and chamber  soil respiration to > 80% basal area mortality is striking and surprising, as is the difference between the EC and chamber estimates.