Understanding anuran calling activity patterns is important for maximizing efficiency and value of call survey data collection and analyses. Previous studies have primarily focused on identifying and quantifying abiotic variables that influence anuran calling activity, and investigating relationships between calling activity and population estimates. In this study we investigated the use of a predictor pond approach to guide call survey effort. In this approach, calling activity at a subset of breeding sites (e.g., ponds) is used as a predictor of calling activity at additional breeding sites, with the goal being to minimize sampling effort while simultaneously maximizing sampling efficiency. We explored the efficiency of this approach using call survey data collected on the endangered Houston Toad (Bufo [Anaxyrus] houstonensis) at 15 known breeding ponds over 9 survey years. We found that if calling activity at 3 predictor ponds was used to decide if additional call surveys would occur at the remaining 12 ponds, we would have hypothetically correctly assumed calling activity was not occurring at non-predictor ponds on 92.1% of survey nights, and we would have hypothetically detected 93.9% of the total number of detected individuals over the 9 survey years. We found the predictor pond approach performed well in our case study, and believe it could be a valuable tool for many anuran monitoring programs.
Frog calling activity detection using lightweight CNN with multi-view spectrogram: A case study on Kroombit tinker frog
