Cross-site analysis of tree carbon reserves in Eastern US forests

Background/Question/Methods

Responses and feedbacks of forests to global change are a major source of uncertainty in predicting future climate forcings and feedbacks. Among the processes least understood in forests is the role that tree carbon reserves play in buffering trees against environmental stressors. It has been suggested that tree carbon reserves (non-structural carbohydrates, NSC) play a large role in reducing tree mortality rates by providing a carbon reserve buffer during stressor events. However, data on NSC reserves are limited to small spatial, temporal and taxonomic scales. In order to better understand NSC reserves and the drivers of NSC variability, we established forest inventory plots, in the summer of 2011, at 10 sites across the Eastern US and measured sapwood NSC levels in all adult trees. At the stand level, we hypothesized that early successional species will have the lowest NSC and that canopy dominant trees will have higher NSC than understory trees. At the site level, we hypothesized that NSC will decrease as slope and/or elevation increase. At the regional level, we hypothesized that NSC will decrease as precipitation decreases and temperature increases. Observed NSC levels were also compared to predicted responses from the Ecosystem Demography model v2.2 (ED2).

Results/Conclusions

Simulated responses of NSC pools by ED2 predicted that at the stand level, NSC will be highest in mid- and lowest in early-successional tree species. Stand-level preliminary results were similar, with early-successional species tending to have lower NSC levels than mid- or late-successional. At the site level, NSC is predicted to decrease gradually with local surface slope and decrease at both low and high elevations. Preliminary results show that NSC tends to decrease with slope, although this relationship is expected to become clearer as sites with steeper topographic gradients are analyzed. At the regional level, NSC is predicted to decline rapidly at very low levels of precipitation and decrease with above- and below-average temperatures. Clear patterns at the regional level have not yet been determined from preliminary field data analysis. However, the similarity in model predicted responses and preliminary observations indicate that NSC levels are responding to environmental variables in expected, and physiologically meaningful, ways. Ongoing analyses are being used to determine the role of NSC in forest community dynamics (e.g. growth, mortality, reproduction) and also to improve and validate (with data) the representation of carbon reserves used in the current generation of ecosystem models.