A model comparison of fire return interval impacts on carbon and species dynamics in a southeastern U.S. pineland

Abstract

Ecosystem process models can be used to predict forest response to disturbances at a range of scales. Selection of the spatial class of model should depend on the scale of the research or management question, and model type should depend on the ecosystem attributes of interest. In some cases, multiple classes of models could be used to address a single research question, with evaluations at each scale having potential benefits and drawbacks. This study examines two classes of models relative to how fire return intervals impact carbon and species dynamics in a southeastern U.S. pineland. A model that can be run as a global class model (ED) and a landscape class model (LANDIS-II) were parameterized with species inventory data from an experimental Pinus Palustris (longleaf pine) forest in southwest Georgia, and simulations were calibrated with literature values, then validated with eddy-covariance data from the study site. A variety of fire scenarios that included prescribed fire with a 2-yr return interval, fire exclusion, and three wildfire scenarios (20-, 50-, and 100-yr return intervals) were used for model runs. Results were compared and evaluated with regard to ecosystem carbon and species dynamics. Both models illustrated that prescribed fire provided the greatest carbon sequestration potential and most stable aboveground biomass through time when compared to the wildfire scenarios. The fire exclusion scenario for LANDIS-II was the only scenario where prescribed fire did not provide the greatest carbon sequestration potential. However, fire exclusion on the order of centuries was a condition of this outcome and the occurrence of such long fire-free periods is considered unrealistic in this fire-prone landscape. Differences between models were primarily the result of the underlying characteristics of each model class, namely the spatial resolution and number of species included. In the end, two vastly different scale models supported the conclusion that high frequency prescribed fire in southeastern U.S. pinelands stabilizes carbon and maintains species composition in an ecosystem that is a known ecological hotspot.

Opportunities for Research on Carbon Management in Longleaf Pine Ecosystems

“…While these stands contain more C in competing woody vegetation, they have a higher risk of catastrophic wildfire than naturally regenerated longleaf pine stands managed with prescribed fire. Researchers used forest landscape models to compare C outcomes among prescribed fire exclusion, prescribed fire, and wildfire scenarios [11,12]. These studies indicated that prescribed fire scenarios resulted in lower C emissions and greater aboveground biomass over time compared to wildfire scenarios.…”