Listening for a needle in a haystack: passive acoustic detection of dolphins at very low densities

Abstract

Passive acoustic surveys have potential for detecting trends in abundance and habitat use by rare cetaceans. We deployed commercially available acoustic data loggers (T-PODs) in 4 har- bours on the west coast of New Zealand's North Island between 2005 and 2008 to investigate the dis- tribution of Maui's dolphin Cephalorhynchus hectori maui and assess whether current protection measures are sufficient. A set of decision rules was developed to minimise the potential for false pos- itive detections. Over 3211 'T-POD days' of acoustic monitoring, 39 click trains which satisfied all of our decision rules were detected, indicating the presence of Maui's dolphins in Manukau and Kaipara Harbours. Data from the site with the most detections were fitted to 3 models, showing that the number of detections varied temporally (p < 0.001). The models were also used to show to what degree dolphins could have been present at monitored locations yet remain undetected. The study highlighted the challenges of passive acoustic monitoring of rare species, particularly of small del- phinids in an environment which is both physically and acoustically challenging. Nonetheless, we demonstrated that T-PODs are effective in studies of Maui's dolphin distribution, that Maui's dolphins are found in North Island harbours and remain at risk from gillnet bycatch. We make a number of rec- ommendations concerning acoustic monitoring studies of rare cetaceans, principally that a thorough understanding of the target signals and the acoustic environment being monitored is essential for maximising acoustic detection rates.