Spatial ecology and site occupancy of the Northern red-legged frog (Rana aurora) in a coastal dune environment

The southern distribution of the northern red-legged frog (Rana aurora) extends only into the three most northern coastal counties of California and seems to be restricted to the coastal belt. Because of recent declines in the population and destruction of its habitat, the California Department of Fish and Game has declared it a Species of Special Concern in California and it is also federally listed as a Sensitive Species by the US Forest Service. This study investigated the spatial distribution and habitat selection of the species on the North Spit of Humboldt Bay, CA in a locally isolated dune environment. The first part of this study concentrated on natural history facets of the species such as differences in relative abundance, movement, and breeding phenology between and among permanent and seasonal pond-breeding habitats in the study area. The second part of this study described the spatial distribution of the northern red-legged frog within seasonal wetlands of the dune swales using site occupancy models. It also gleaned knowledge on the species' habitat preferences during the breeding and post-breeding seasons and serves as a baseline experiment for future monitoring of this species distribution. Detection probability was also examined in the egg mass site occupancy models. Frog abundance was greater in the smaller, permanent pond habitat. Movement towards the ponds was greatest during the breeding season in both habitats. Juveniles were most active prior to and during the breeding season, whereas adults were more active during and after the breeding season. Prior rainfall increased movements in both juveniles and adults, although adults were not as affected except prior to the breeding season. No differences in breeding chronology and physical characteristics were found between the habitats. The egg mass occupancy models showed that gravid females preferred oviposition sites with deeper water and less than 50% emergent vegetation. Egg mass detection probabilities were affected by fluctuating water levels throughout the survey period. Egg mass occupancy rates were underestimated when detection was not incorporated into the models. Adult and juveniles in the post-breeding season selected sites with longer hydroperiods that were farther from sites which had contained egg masses. Occupancy results indicate that water depth and hydroperiod length are important factors in determining the spatial distribution of R. aurora in palustrine wetlands, detection probability is an important parameter to include when surveying the species, and the adult and juveniles use more than just the breeding ponds in the wetland habitat. This is the first study to research R. aurora egg mass site occupancy incorporating detection probability. The top model can be used as a foundation to predict egg mass occurrence in palustrine wetlands and as a baseline for development of a larger monitoring program to research spatial dynamics and population fluctuations of the species.