Passive acoustic monitoring (PAM) allows the study of marine species and their habitats through sound. PAM is one of the primary techniques used to monitor the presence and behavior of marine mammals. Among PAM technologies, autonomous acoustic recorders are widely employed, and the diversity of underwater soundscapes has driven the development of a variety of devices.
However, differences between devices can make results difficult to compare, risking misinterpretation when integrating datasets.
This study assessed the performance of five PAM devices (SoundTrap, RTsys, HydroMoth, F-POD, and C-POD) in detecting bottlenose dolphin (Tursiops truncatus) vocalisations, with a focus on improving comparability of their findings. Results confirmed that detection capabilities vary significantly between devices, which can lead to biased estimates of species presence. However, applying customised settings in post-processing software (e.g., PAMGuard) reduced false positives and improved accuracy. SoundTrap device consistently outperformed others, offering broader bandwidth and richer acoustic detail, while low-cost tools like HydroMoth showed potential for certain applications. Click loggers, though energy-efficient, may underestimate presence by missing certain vocalisations.
Results show that, with proper post-processing adjustments, vocalization detections become comparable, but a trade-off between specificity and sensitivity must always be considered. This study highlights the need for standardized processing to support reliable data integration for marine mammal monitoring and effective conservation.
promotor/supervisor feedback
nothing yet
