What the F‐POD? Comparing the F‐POD and C‐POD for monitoring of harbor porpoise (Phocoena phocoena)

Abstract

AbstractPassive acoustic monitoring (PAM) is a cost‐effective method for monitoring cetacean populations compared with techniques such as aerial and ship‐based surveys. The Cetacean POrpoise Detector (C‐POD) has become an integral tool in monitoring programs globally for over a decade, providing standardized metrics of occurrence that can be compared across time and space. However, the phasing out of C‐PODs following the development of the new Full waveform capture POD (F‐POD) with increased sensitivity, improved train detection, and reduced false‐positive rates represents an important methodological change in data collection, particularly when being introduced into existing monitoring programs. Here, we compare the performance of the C‐POD with that of its successor, the F‐POD, co‐deployed in a field setting for 15 months, to monitor harbor porpoise (Phocoena phocoena). While similar temporal trends in detections were found for both devices, the C‐POD detected only 58% of the detection‐positive minutes (DPM), recorded by the F‐POD. Differences in detection rates were not consistent through time making it difficult to apply a correction factor or directly compare results obtained from the two PODs. Generalized additive models (GAMs) were used to test whether these differences in detection rates would have an effect on analyses of temporal patterns and environmental drivers of occurrence. No differences were found in seasonal patterns or the environmental correlates of porpoise occurrence (month, diel period, temperature, environmental noise, and tide). However, the C‐POD failed to detect sufficient foraging rates to identify temporal patterns in foraging behavior, which were shown by the F‐POD. Our results suggest that the switch to F‐PODs will have little effect on determining broad‐scale seasonal patterns of occurrence but may improve our understanding of fine‐scale behaviors such as foraging. We highlight how care must be taken interpreting F‐POD results as indicative of increased occurrence when used in time‐series analysis.