Multi-resolution assessment of the foraging and reproductive ecology of red-cockaded woodpeckers in a Mississippi loblolly-shortleaf pine forest.

摘要

Red-cockaded woodpeckers (Picoides borealis) (RCWs) are a federally-endangered species endemic to mature pine forests of the southeastern United States. Since 1970, virtually all aspects of RCW ecology have been subject to intensive management, including habitat. Habitat management focuses on the provision of cavities and foraging habitat using long even-aged timber rotations, preventing of hardwood midstory encroachment, and prescribed burning. Habitat management guidelines were based primarily on research performed in longleaf pine (Pinus palustris) forests, however, approximately 50% of all RCW populations are located in loblolly pine (P. taeda ), shortleaf pine (P. echinata), or mixed pine forests. The U.S. Forest Service has requested that site-specific foraging habitat studies be performed in loblolly-shortleaf pine forests.; From 1997 to 1999, I studied 41 RCW groups at the Bienville National Forest in central Mississippi. I performed 5-hr visual focal runs on RCW groups annually and recorded a mean annual home range of 43.1 ha. Non-breeding season home ranges were greater than breeding season home ranges. RCWs selected large pine stems for >90% of their foraging activities, although hardwood stems were selected more frequently during the non-breeding season. At the stand level, RCWs preferred, in ranked order, pine sawtimber, hardwood sawtimber, pine poletimber, and pine regeneration. I developed regression models to predict RCW group status using number of pine poletimber patches, pine sawtimber mean patch size, and pine sawtimber edge density. I developed a regression model that predicted RCW reproductive success at 3 temporal scales using pine poletimber mean patch size, number of pine sawtimber patches, and pine poletimber core area.; Analysis of microhabitat variables in plots centered on foraging and randomly-selected stems indicated that RCWs selected microhabitat within habitat patches. However, logistic regression models using microhabitat variables were poor predictors of RCW foraging and randomly-selected stems. I used microhabitat variables to develop predictive models of RCW nest stage parameters. Various microhabitat variables were useful in developing models that predicted effectively RCW reproductive success parameters, however, a model to predict RCW fledging success was only moderately successful.