Thursday, August 9, 2007 - 2:50 PM

OOS 45-5: Developing a decision-support tool for longleaf pine conservation in the Southeast

Kevin Kleiner1, Allison Vogt2, Mark D. Mackenzie1, and James B. Grand1. (1) Auburn University, (2) East Gulf Coastal Plain Joint Venture

The longleaf pine woodland was the historically dominant ecosystem in the southeastern coastal plain.  Once stretching from North Carolina to Texas, it has now been reduced to a fraction of its previous areal extent.  The East Gulf Coastal Plain Joint Venture (EGCP JV), recognizing longleaf pine ecosystems as priority habitat for bird conservation, began development of GIS-based decision support tools created to focus the collective efforts of EGCP JV partners on areas within longleaf pine ecosystems predicted to be of highest conservation value.  This paper describes a method for prioritizing restoration sites using a spatially explicit decision support system.
An essential component of restoration site identification is identifying where longleaf currently exists.  We incorporated a longleaf pine ecosystem map developed through the Alabama Gap Analysis Program (AL-GAP).  A logistic regression model was created relating longleaf occurrence to spectral values in a satellite image.  This model was then applied to the satellite image and a map was generated specifying the probability of longleaf occurrence at each pixel.  Other important spatial data layers included location of historic longleaf pine woodland sites, existing land cover, biological landscape metrics (minimum patch size, fragmentation tolerance, total area requirement), projected future urban growth, and land value.  Data layers were incorporated into a GIS and spatial queries were performed to identify priority restoration sites.
Although this study’s focus was limited to the area within the conservation planning boundary of the EGCP JV, these methods can be applied throughout the historic range of longleaf pine.