You are here

Harvard Forest >

Harvard Forest Symposium Abstract 2016

  • Title: Comparing tree-ring and permanent plot estimates of aboveground woody production in three Eastern U.S. forests
  • Primary Author: Alex Dye (West Virginia University)
  • Additional Authors: Audrey Barker Plotkin (Harvard Forest); Daniel Bishop (Harvard Forest); Amy Hessl (West Virginia University); Neil Pederson (Harvard Forest); Ben Poulter (Montana State University)
  • Abstract:

    Globally, forests account for a large portion of sequestered carbon, much of which is stored as wood in trees. Measurements of the rate of carbon accumulation in the aboveground woody materials of trees, or woody net primary productivity (wNPP), quantify annual to decadal variations in forest carbon sequestration. Permanent remeasurement plots are often used to estimate stand-level wNPP but are usually not annually resolved and take many years to establish a long dataset. Tree rings are a unique and relatively infrequently used source for measuring wNPP, and benefit from fine spatial (individual trees) and temporal (annual) resolution. Because of this precision, tree rings are a complementary addition to permanent plots and the suite of tools used to study forest productivity.



    For a lowland evergreen (Howland, Maine), mixed deciduous (Harvard Forest, Massachussets), and mixed mesophytic (Fernow, West Virginia) forest in the eastern United States, we demonstrate that annual estimates of wNPP developed from tree rings approximate estimates derived from permanent remeasurement plots, even when plots are not directly co-located. We simulate random draws of subsets of permanent plot data to describe the distribution of possible wNPP estimates given a sampling area size equivalent to the tree-ring plots. Though mean tree-ring wNPP overestimates permanent plot wNPP slightly at the Maine and central Massachusetts sites and underestimates at the West Virginia site, it is still within the distribution of random draws at all sites.

    We observe variability in plot-level estimates of wNPP at each site, and attribute the magnitude of this variability to species and size heterogeneity across the forest. Inter-plot wNPP variability is highest at the most diverse site (Fernow), and lowest at Howland where species and structural diversity is minimal. In developing stand-level estimates of wNPP using tree-ring plots, we show it is essential to sample a representative selection of species and size classes, especially in forests with high diversity. Furthermore, we develop the benefits and concerns associated with using tree rings to reconstruct annual wNPP and call for the application of tree rings in carbon cycle studies across a broader range of species diversity, productivity, and disturbance histories.

  • Research Category: Historical and Retrospective Studies
    Large Experiments and Permanent Plot Studies
    Regional Studies