Anuran Calling Survey Optimization: Developing and Testing Predictive Models of Anuran Calling Activity

Amphibian populations, particularly anurans, are declining worldwide, and programs that use calling surveys have been established to monitor anuran populations. Models that describe the environment's influence on calling may be useful to increase detection allowing optimization of surveys. Using an automated recording system, we evaluated the calling activity of Pseudacris crucifer, Pseudacris feriarum, and Rana sphenocephala at an ephemeral wetland in the Piedmont of North Carolina. We used stepwise logistic regression to model environmental variables that significantly affected calling activity. Models revealed that, for P. crucifer, water and air temperature positively influenced calling, whereas day of year, barometric pressure, and light intensity negatively influenced calling. For P. feriarum, air temperature positively influenced calling, and day of year, relative humidity, wind speed, barometric pressure, and light intensity negatively influenced calling. Finally, air temperature positively influenced calling for R. sphenocephala, whereas water temperature, relative humidity, wind speed, and light intensity negatively influenced its calling. Using these results, we developed comprehensive as well as simpler, user-friendly models, predicting the best conditions under which to conduct anuran calling surveys. The user-friendly models were tested using previously collected data from calling surveys performed in the same region of North Carolina and found to accurately predict calling activity approximately 70% of the time. We discuss how weather forecast data may be applied to these models to determine the best times to conduct calling surveys and how models such as those developed in this study can be used to interpret data previously collected during amphibian calling surveys.