Harvard Forest >

Harvard Forest Symposium Abstract 2019

• Title: Tree Competition, Growth, and Mortality in the Environmental Measurement Site Eddy Flux Footprint
• Primary Author: Timothy Whitby (Harvard University)
• Additional Authors: J. William Munger (Harvard University); Laura Puckett (Virginia Polytechnic Institute and State University); Meghan-Grace Slocombe (Greenfield Community College)
• Abstract:

  The Environmental Measurement Site eddy flux tower has shown that Harvard Forest’s Prospect Hill tract has consistently accumulated carbon on an annual scale for the last 27 years. However, the long-term prospects of this carbon sink are uncertain, as rates of tree growth and mortality are expected to reach equilibrium based on accepted successional theory. Tree mortality is a rare, and relatively unpredictable event, but recent trends suggest that carbon loss through tree mortality may be on the rise. In the summer of 2017, with the assistance of two REU students, we investigated whether the spatial sampling of trees in our permanent plots was sufficient to capture forest-wide trends in mortality. To dive further into the mechanisms of tree mortality, we investigated whether competition (as measured by crowding) with surrounding trees correlated with recent mortality.
  To sample an approximately equal area outside of our existing plots, we ran 5 m wide transects between neighboring plots extending away from the tower. We found that the mean species-level mortality rates did not differ from those in transects (t = 0.88, df = 23.07, p-value = 0.38), although it varied widely among species. At the plot/transect level the number of dead trees / hectare (to account for varying sizes of individual transects) also did not differ between sampling methods (t = -1.50, df = 9.22, p-value = 0.17). Unfortunately, the sample size of dead trees of a single species were still too small to test for individual differences. Qualitatively, our plots may under-sample red pine mortality, as they are in distinct patches in the tower footprint, and we may over-sample white spruce, birch, and cherry, which are relatively rare species (Figure 1).
  By mapping the diameter and location of all the trees in our plots, we were able to calculate a competition index (CI) that was the distance weighted ratio of competitor tree to the focal tree. Overall, CI varied widely (figure 2), and dead trees did not necessarily experience high competition based on this metric. However, it appears that something is causing recently dead trees (black triangles in the plot) to experience near-zero or negative annual diameter growth. If we look at the three most common species – red oak, red maple, and hemlock – we see that competition index is negatively correlated with growth rate, especially in red oak (Figure 3). The negative relationship of growth with competition index is subdued in hemlock and maple. Interestingly, red maple mortality occurs across the whole range of CI, which may indicate that some other external factor is causing them to die. In oak, the dominant species in in terms of biomass, the only mortality events did occur at the high end of CI. Not shown in this data, is a large die-off event of red pine, related to some possible pest or fungal pathogen. We expect that nearly all the former plantation red pine will die off in the next year.
  In conclusion, the sampling regime does capture the local patterns of tree mortality accurately, although species-level trends are idiosyncratic. Competition may influence mortality risk in a given tree, but there are unaccounted for factors that may be the ultimate cause of mortality. In this set of trees, the best indicator we have of pending mortality is slow or negative stem diameter growth. Further analysis of this data will investigate intraspecific v. interspecific competition, species community composition change over time, and accounting for environmental variables that may confound the growth rate and competition relationship.
  

• Research Category: Forest-Atmosphere Exchange
  Biodiversity Studies


• Figures:
fig1.jpg
fig2.jpg
fig3.jpg