Geographic variation in fitness and foraging habitat quality in an endangered bird

Models are important tools for conservation, but the usefulness of any given model for decision-making depends on its accuracy and precision. Few models designed for conservation purposes are validated with real-world data, and such models are even less likely to be revisited and improved with post-implementation results. We test the performance of a model frequently used and heavily relied-upon for the management of the endangered red-cockaded woodpecker (Picoides borealis). The RCW Foraging Matrix Application incorporates spatially-explicit forest stand data and woodpecker territory locations to produce quantitative assessments of foraging habitat quality. Model parameters were based on expert opinion and research performed on several key populations at a time when range-wide habitat quality was relatively poor. Since the model's inception, many red-cockaded woodpecker populations have been monitored intensely in restored habitat, providing an opportunity to evaluate model performance range-wide. We assessed the relationship of habitat quality, as measured by the RCW Matrix Application, to group size and fledgling production from populations across the species range in the southeastern United States. We also evaluated foraging habitat quality directly by relating woodpecker fitness components to foraging habitat metrics through regression tree analyses. Results showed that some, but not all, of the habitat metrics included in the RCW Matrix Application were consistently related to fitness components range-wide, but threshold values for these habitat metrics identified by regression tree analyses were site-specific rather than universal. Our findings indicate opportunities for improving on “one-size-fits-all” range-wide models with analyses of additional locally-relevant foraging habitat metrics.