Harvard Forest >

Harvard Forest Symposium Abstract 2018

  • Title: Hide and go seek with temperature signals of Northeastern US Tree Species
  • Primary Author: Ross Alexander (Illinois Mathematics and Science Academy)
  • Additional Authors: Kevin Anchukaitis (University of Arizona); Daniel Bishop (Harvard Forest); Edward Cook (Columbia University in the City of New York); Jessie Pearl (University of Arizona); Neil Pederson (Harvard Forest)
  • Abstract:

    Improving reconstructions of past climate is vital in providing long-term context for regional climate change. There have been only two published reconstructions of past temperatures in the northeastern U.S. (NEUS) since the 1980s, one based on Picea rubens, and one out in 2017 based upon Chamaecyparis thyoides (Atlantic white cedar; AWC). Because increased species diversity generally improves dendrohydroclimatic reconstructions and both Picea rubens and AWC have limitations as paleoproxies due to land-use and air pollution, we conducted a series of tests to ask, "Does species diversity improve reconstructions of temperature history in the northeastern United States?" The first two tests were performed on AWC and then a network of AWC and Picea rubens. Subsequent tests added groups of species or genera from a network of 230 tree-ring chronologies beginning with those having the strongest relation between warming temperatures and increased growth. PC1 of the AWC test represented 40% of the variance and showed a significant positive relation with winter temperature (r = 0.38). As additional species were included, the "winter temperature PC" accounted for less of the overall variance, ranging from 26% of the variance in test 2 to 5% by test 5. We constructed a common period minimum winter temperature reconstruction from 1900-1970 to assess species contributions to the regression model. Cumulatively, AWC contributes the most to the model's explained variance. Individually species such as Nyssa sylvatica, Quercus stellata, Platanus occidentalis, and Liriodendron tulipifera contribute greatly to model performance. Based on these analyses we constructed a nested reconstruction of winter minimum temperature from 1393 - 1970 using the full suite of chronologies. This work highlights the need to resample established collections, and seek out underrepresented, but climate-sensitive, species to push the understanding of Northeast US climate further into the past.

  • Research Category: Historical and Retrospective Studies; Physiological Ecology, Population Dynamics, and Species Interactions; Regional Studies