Time series analysis of forest carbon dynamics: recovery of Pinus palustris physiology following a prescribed fire

Abstract

Frequency and intensity of fire determines the structure and regulates the function of savanna ecosystems worldwide, yet our understanding of prescribed fire impacts on carbon in these systems is rudimentary. We combined eddy covariance (EC) techniques and fuel consumption plots to examine the short-term response of longleaf pine forest carbon dynamics to one prescribed fire at the ends of an edaphic gradient (mesic and xeric sites). We also introduce novel (to the EC research community) statistical time-series approaches to quantify the drivers of carbon dynamics in these systems. We determined that our mesic site was a moderate sink of carbon (−157.7 ± 25.1 g C m−2 year−1), while the xeric site was carbon neutral (5.9 ± 32.8 g C m−2 year−1) during the study. The fire released 408 and 153 g C m−2 year−1 for the mesic and xeric sites, respectively. When loss associated with fire was combined with net ecosystem exchange rates, both sites became moderate carbon sources for the year. Analyses of assimilation and respiration parameters (e.g., maximum photosynthesis, quantum efficiency, and daytime ecosystem respiration) showed a positive trend over time pre-fire and a negative trend over time post-fire for maximum ecosystem CO2 uptake rates, and the opposite relationship for daytime ecosystem respiration rates. Within 30 days following fire, ecosystem physiological activity was statistically similar to pre-fire and appeared to be driven by the pine canopy. Our results suggest that prescribed fire (low intensity, high frequency) maintains the existing structure and function (in this case, carbon flux rates) because longleaf pine ecosystems have evolved with fire. This study, 1 year in length, provides a foundational understanding of the complex interaction between fire and carbon dynamics for longleaf pine ecosystems. Moreover, it provides a case study for applying time series analysis methods to EC data where there are complex relationships between ecosystem physiological activity and environmental drivers. However, to elicit a broader understanding of the complex interaction occurring between fire and carbon dynamics long- term studies are needed.