Odontocete Ecology off Northumberland, UK: Advancing the Application of Passive Acoustic Monitoring

Abstract

Effective conservation requires robust data on species occurrence, abundance, and area usage. Providing this information for odontocetes is difficult given that they are highly mobile, occupy areas at low density and are visually cryptic. The first chapter presents the result of a structural topic model used to evaluate abstracts representing odontocete research over the last 50 years. The results provide a comprehensive overview, acts as an introduction to the research field and identifies the need for further conservation related research. It further identified Bioacoustics and Communication as important areas of odontocete research. Passive Acoustic Monitoring (PAM) has proven a viable approach to studying odontocetes as all species produce sounds (e.g., echolocation clicks) that can be detected using hydrophones. To date, PAM studies have mostly focused on occurrence with few studies addressing other research aims. To address this, emerging and novel approaches to analysing PAM data were applied to recordings collected at three sites off the Northumberland coast, UK, a relatively understudied area of growing interest due to increased common bottlenose dolphin (Tursiops truncatus) activity in recent years. The effect of environmental drivers on the occurrence of three sympatric species (common bottlenose dolphin, white-beaked dolphin (Lagenorhynchus albirostris) and harbour porpoise (Phocoena phocoena)) were identified using GAM-GEE analysis of species-specific click detections, identifying key drivers such as seasonality, location, and diel phase. Bottlenose dolphin area usage was investigated by incorporating click and whistle detections into a single hidden Markov modelling framework to identify latent “activity states”. Differences in activities interpreted as foraging and socialising were identified across each location, and across diel and tidal phases. Abundance estimates were produced for bottlenose dolphin from individual specific “signature whistles” using different modelling approaches. These estimates were compared to a separate estimate produced from photo-ID data collected at the same time from boat-based surveys