PS 87-68
Quantifying uncertainties in tree-ring estimates of biomass at the Valles Caldera, NM

Friday, August 14, 2015
Exhibit Hall, Baltimore Convention Center
M. Ross Alexander, Laboratory of Tree-Ring Research, University of Arizona, School of Natural Resources and the Environment, University of Arizona, Tucson, AZ
Christine R Rollinson, Earth and Environment, Boston University, Boston, MA
Flurin Babst, Laboratory of Tree Ring Research, University of Arizona, Tucson, AZ
Marcy Litvak, Department of Biology, University of New Mexico, Albuquerque, NM
David J.P. Moore, School of Natural Resources and Environment, University of Arizona, Tucson, AZ
Valerie Trouet, Laboratory of Tree Ring Research, University of Arizona, Tucson, AZ
Background/Question/Methods

The recent pairing of tree-ring techniques with eddy-covariance systems has allowed for variability in forest carbon uptake to be analyzed at interannual time-scales.  Biomass estimates calculated from tree rings have provided insights into carbon allocation within the forest, but the uncertainties around these estimates are large and poorly quantified.  We have identified five major areas of uncertainty associated with tree-ring based assessments of aboveground biomass: 1) sample dating and dropout through time, 2) sample representativeness, 3) allometric uncertainty, 4) site history (disturbance and mortality), and 5) uncertainty around the mean tree response through time. We illustrate how these areas of uncertainty contribute to biomass reconstructions from two forests at different elevations in the Valles Caldera in northern New Mexico.  The lower-elevation site was composed of ponderosa pine and had a large spatial variability in density, whereas the upper-elevation site was dominated by Engelmann spruce and was relatively homogenous in density. Biomass for both sites was reconstructed from 1903 to 2011, and the ranges of the uncertainties around the mean biomass reconstruction were calculated over the entirety of the time series. 

Results/Conclusions

The overall uncertainty ranged from 280 ± 187% (mean ± SD) of the mean biomass at the upper-elevation site to 390 ± 245% at the lower-elevation site.  However, the relative contribution of various uncertainty components was not conserved across sites.  Tree density differed greatly between the upper and lower sites (0.34 vs. 0.15 trees/m2, respectively) and represented 40% of the overall uncertainty at the lower-elevation site (188 ± 40% of mean biomass).  At the upper-elevation site, tree density was more homogeneous, and thus was a minor contributor (6%) to the overall uncertainty.  At this site, mortality was the dominant contributor (40%) to biomass uncertainty (118 ± 72% of mean biomass).  Uncertainty around the mean tree response through time was 18% of the total uncertainty at both the upper and lower sites.  At the Valles Caldera, the areas of uncertainty relating to present and past forest density contributed the most to uncertainty surrounding aboveground biomass estimates.