Using california gnatcatcher to test underlying models in habitat conservation plans

Habitat Conservation Plans are a widely used strategy to balance development and preservation of species of concern and have been used in southern California, USA, to protect the coastal California gnatcatcher (Polioptila californica). Few data exist on gnatcatcher abundance and distribution, and existing data have problems with issues of closure (i.e., sampling occurs in a short enough time period such that abundance or distribution are not changing), detectability, and proper attention to probability-based sampling schemes. Thus, a habitat model has been relied upon in reserve design. California gnatcatchers are the flagship and umbrella species of many plans and we provide the first estimates that incorporate probabilistic sampling and test predictions from the habitat model. Probability of occurrence was 26% (S (E) over cap = 0.06); however, occupancy varied by modeled habitat quality with slopes <40%, warm, and wet sagebrush habitat having higher occupancy probabilities. Interpreting abundance and occupancy probabilities by vegetation type was complicated by error detected in Geographic Information System vegetation metadata files. The slope (1.08, S<(E)over cap> = 0.66), temperature (0.79, S (E) over cap = 0.70), and precipitation (-2.62, S (E) over cap 1.21) variables associated with habitat models were stronger influences on occupancy than was patch size (0.48, S (E) over cap 0.66). Previous models weight patch size equal to slope and climate. Our work demonstrates that probabilistic sampling can be carried out on a large scale and the results provide better information for managers to make decisions about their reserves.