Harvard Forest >

Harvard Forest Symposium Abstract 2012

• Title: Effects of warming on tree species’ recruitment in deciduous forests of the eastern U.S.A
• Primary Author: James Clark (Duke University)
• Additional Authors: Chelsea Baldino (Marine Biological Laboratory); Frank Bowles (Research Designs); Jerry Melillo (Marine Biological Laboratory); Jacqueline Mohan (University of Georgia); Carl Salk (Duke University); Lindsay Scott (Marine Biological Laboratory); William Werner (Marine Biological Laboratory)
• Abstract:

  Near-term increases in productivity will occur if phenology responds to climate warming with prolonged growing seasons. This possibility is supported by some long-term observational studies showing advancing budbreak that coincides with a warming climate. Other studies suggest that local adaption cued by photoperiod can limit phenological response. Limited response to climate change has adaptive value, guarding against damage from late frost in spring and early frost in fall. The price of this muted response is the missed opportunity to fully exploit the potential growing season. Due to the challenges for experiments and models it has been difficult to determine how responsive phenology will be to climate change.
  
  We are currently conducting two large field experiments to study the effects of climate warming on tree phenology and tree recruitment in transplant gardens in deciduous forests at the Harvard Forest in central Massachusetts and at Duke Forest in the Piedmont Region of North Carolina. Seeds and seedlings from Georgia, North Carolina and Massachusetts are being planted in open-top chambers at the two research sites. The experimental treatments at each site included manipulations of temperature (ambient, +3°C, +5°C) and light (shade, gap). The study has been operating for two full growing seasons. For nearly all species in both years, warming has hastened budburst and germination. Individual species showed substantial variability of all events’ dates within treatments. The data are being analyzed in the context of a new modeling framework.
  
  The modeling portion of this study began with the current state-of-the-art thermal (degree day) models. We found that, due to unrealistic assumptions, current thermal models for budbreak cannot be used to identify the onset of development (the threshold temperature), they do not actually incorporate the assumptions of adaptation to freezing risk, and they cannot be used predict phenology at a location other than where they are fitted. We then constructed a new modeling approach that accommodated continuous phenological development, combined with observations that were discrete and interval censored in time and discrete in phenological state. Results from this study challenge the basic assumptions of thermal models for phenology. We find substantial phenological response and large differences between individuals of different species. The actual effect of increasing temperature on phenology is underestimated with current models. Our results indicate that phonological response to temperature increase will be greater than previously thought.
  
  

• Research Category: