Use of LiDAR to define habitat thresholds for forest bird conservation
Quantifying species-habitat relationships provides guidance for establishment of recovery standards for endangered species, but research on forest bird habitat has been limited by availability of fine-grained forest structure data across broad extents. New tools for collection of data on forest bird response to fine-grained forest structure provide opportunities to evaluate habitat thresholds for forest birds. We used LiDAR-derived estimates of habitat attributes and resource selection to evaluate foraging habitat thresholds for recovery of the federally endangered red-cockaded woodpecker (Leuconotopicus borealis; RCW) on the Savannah River Site, South Carolina. First, we generated utilization distributions to define habitat use and availability for 30 RCW groups surveyed over a >4-h period twice per month between April 2013 and March 2015. Next, we used piecewise regression to characterize RCW threshold responses to LiDAR-derived habitat attributes described in the United States Fish and Wildlife Service recovery plan for RCW. Finally, we used resource utilization functions to estimate selection of specific habitat thresholds and used the magnitude of selection to prioritize thresholds for conservation. We identified lower and upper thresholds for densities of pines ≥35.6 cm dbh (22, 65 trees/ha), basal area (BA) of pines ≥25.4 cm dbh (1.4, 2.2 m2/ha), hardwood canopy cover (6, 31%), and BA of hardwoods 7.6–22.9 cm dbh (0.4, 6.07 m2/ha); we identified three thresholds for density of pines 7.6–25.4 cm dbh (56, 341, and 401 trees/ha). Selection rankings prioritized foraging habitat with <6% hardwood canopy cover (β = 0.254, 95% CI = 0.172–0.336), < 1.2 m2/ha BA of hardwoods 7.6–22.9 cm dbh (β = 0.162, 95% CI = 0.050–0.275), ≥1.4 m2/ha BA of pines ≥25.4 cm dbh (β = 0.055, 95% CI = 0.022–0.087), and ≥22 pines ≥35.6 cm dbh/ha (β = 0.015, 95% CI = 0.013–0.042). We identified habitat thresholds corresponding to open canopy structure, moderate densities of large and medium pines, and sparse hardwood midstory trees. Selection ranks prioritized multiple thresholds below USFWS range-wide recovery thresholds, indicating site-specific management goals may be beneficial for RCW conservation. Fine-grained LiDAR-derived habitat data coupled with GPS-derived habitat use can guide forest bird conservation by identifying the full range of structural conditions associated with threshold responses.