Tuesday, August 3, 2010

PS 26-17: The role of prescribed fire on longleaf pine carbon dynamics along an edaphic moisture gradient

Andrew Whelan1, Gregory Starr1, Robert Mitchell2, Jason D. McGee2, Christina L. Staudhammer3, Jennifer Wright4, and Hank Loescher5. (1) University of Alabama, (2) Joseph W. Jones Ecological Research Center, (3) University of Florida, (4) University of Edinburgh, (5) NEON Inc.

Background/Question/Methods

Current general circulation models (GCMs) predict increased periods of prolonged drought in the Southeastern United States, and the average area burned annually in some forests may double by the year 2050.  These predictions are largely associated with catastrophic wildfires in forests with long fire return intervals.  These disturbances cause major changes in ecosystem carbon dynamics and forest structure, and recovery can take hundreds of years.  Longleaf pine forests in the southeastern United States experience frequent, low intensity fires that sustain woodland forest structure and the highest levels of biodiversity recorded in North America.  To date, only a limited number of studies have focused on prescribed fire's role in ecosystem carbon dynamics.  We used the eddy covariance method along an edaphic moisture gradient in longleaf pine forests at the Joseph W. Jones Ecological Research Center in Newton, Georgia to investigate how interactions between soil water availability and fire affect net ecosystem exchange (NEE).  We hypothesized that soil water availability would play a larger role than fire in regulating ecosystem carbon dynamics.  Observations began in the summer of 2008, and in the winter of 2009, prescribed fires were conducted on each site.
Results/Conclusions Results from the first year show that the mesic site was a carbon sink (NEE = -142.16 g C m-2yr-1), and the xeric site was a slight carbon source (NEE = 13.82 g C m-2 yr-1).  We attribute this to larger total leaf area index (LAI) and a faster rate of recovery from disturbance at the mesic site.  NEE was reduced at both sites during January 2009 (the month of the fires), but recovered quickly by February 2009 due to a change in the diurnal source and sink relationships.  These decreases in NEE appear to have been driven more by water availability associated with drought when the fires were conducted, than with the direct loss of photosynthetic capacity associated with the burn.  The quick recoveries of typical NEE rates at both sites are associated with the high evolutionary adaptation of these plant communities to low intensity fire.  These findings also suggest the need to characterize the fire-adapted ecosystem process rates as a function of a broad range of abiotic controls across fine temporal scales and larger regional scales.  Future analyses will focus on the seasonal and inter-annual variability of climate and soil water availability on ecosystem carbon dynamics.