Application of long-term passive acoustic monitoring to evaluate spatial and temporal patterns of multi-species acoustic assemblages of marine mammals in the western North Atlantic Ocean

Passive acoustic monitoring (PAM) is an efficient method for large-scale marine mammal monitoring. PAM technologies concurrently sample multiple soniferous marine mammal species, and when coupled with verified detectors, provide information that can be used to evaluate community composition. This analysis used data collected by an array of high-frequency PAM recorders at eight sites between 30°0’N and 42°0’N along the shelf-break in the western North Atlantic Ocean. Daily acoustic presence of 13 known marine mammal species and grouped unknown odontocetes was determined for all months between June 2016 and May 2019. Classification trees were generated from monthly summaries of daily detections to identify community composition dissimilarities. Detections of Gervais’s beaked whale ( Mesoplodon europaeus) represented the root node and split sites by latitude, grouping the three sites south of 34°0’N and the five sites north of 38°0’N. Tree nodes were further divided by other odontocetes. The presence of mysticetes also varied by site and season. Distinctive communities were identified for each site, with odontocetes being resident and mysticetes more migratory. The root and secondary splits were all driven by beaked whale species, demonstrating the importance of identifying these whales to species instead of aggregating them as is common for visual survey data.