Statistically Rigorous Carbon Stock Predictions of Forest Restoration in the Southern United States
Dr. Duncan Wilson | Oklahoma State University
Forest restoration in the southern United States typically involves thinning of the overstory and re-introduction of surface fires, resulting in an open pine-oak savanna that was historically common across much of the region. These restoration treatments cause a shift in the ecosystem productivity from the overstory trees to a vigorous grass and forb understory that is beneficial to many wildlife species. This shift in the plant community also changes the carbon dynamics from largely aboveground to largely belowground sequestration. There is currently no consensus of how carbon sequestration is expected to change under such a scenario; however, preliminary evidence indicates that belowground carbon stocks can shift to being twice as high in the savanna ecosystem.
The objective of this project is to address how to: (1) accurately and precisely model carbon dynamics in forest and savanna ecosystems in the southern United States, (2) extend these results to other areas through field calibration, (3) formulate a statistical error budget for carbon predictions, and (4) integrate models of carbon dynamics with other ecosystem services to facilitate trade-off comparisons.
In this project researchers will examine how to quantify the short- and intermediate-term carbon stock of military bases undergoing widespread forest restoration to more open, savanna-like conditions. Department of Defense (DoD) installation managers also need tools to evaluate the trade-offs (or mutual gains) between carbon sequestration and other ecosystem services such as wildlife habitat or biodiversity. Researchers will develop and parameterize a mechanistic model describing how carbon sequestration changes in a newly restored savanna that is part of a long-term regional experiment and contrast these to closed canopy forest. A calibration approach will be evaluated to enable transfer of these results to wider areas, including other DoD installations. The calibration approach will be validated at Camp Gruber, Oklahoma, which should help extend the scope of inference of a key regional experiment and increase the robustness of predictions for military installations under different restoration scenarios. Researchers will also fully develop an error budget for carbon predictions to provide estimates of uncertainty (e.g., confidence limits). Characterizing the uncertainty around carbon predictions is critical to providing defendable estimates for use in planning and policy discussions.
This project will directly estimate the carbon stock for the 33,000 acre Camp Gruber under no management and different forest restoration scenarios. The carbon and ecosystem services models developed also will be directly applicable to Fort Chaffee, Arkansas. In addition, the field calibration and statistical approaches to estimating carbon stock (but utilizing different regional models of carbon dynamics) will be directly transferable to other forested DoD installations in the United States as well as other large ownerships such as national forests, state-owned forests, or forest industry lands. (Anticipated Project Completion - 2015)