Song repertoire analysis of bowhead whales (Balaena mysticetus) based on passive acoustic monitoring data from mooring ARKF05-17, recorder SV1088, in Fram Strait from July 2016 to July 2017

• Main Author: Thomisch, Karolin
• Format: Dataset
• Language: English
• Published: PANGAEA 2022
• Subjects: Arctic Ocean; ARK-XXX/2; GN05; Bowhead whales; DATE/TIME; F5-17; FRAM; FRAM Moorings; FRontiers in Arctic marine Monitoring; MOOR; Mooring; Mooring (long time); MOORY; North Greenland Sea; passive acoustic monitoring (PAM); Polarstern; Presence/absence; PS100; PS100/019-1; MSM76_198-1; song type analysis; Type; vocal repertoire; Arctic; Greenland; Balaena mysticetus; bowhead whale; Fram Strait; Greenland Sea; North Greenland
• Online Access: https://doi.pangaea.de/10.1594/PANGAEA.945404; https://doi.org/10.1594/PANGAEA.945404

Data on song type repertoire of bowhead whales (Balaena mysticetus) and temporal trends in the occurrence of these song types were obtained from passive acoustic monitoring (PAM) data collected by passive acoustic recorder SV1088 of type Sono.Vault (manufactured by develogic GmbH, Hamburg, Germany) at 79° 00.02' N, 005° 40.12' E, mooring ARKF05-17, in Fram Strait. Passive acoustic data were collected as part of the Frontiers in Arctic Marine Monitoring (FRAM) observatory in Fram Strait from July 2016 to July 2017 (recording period) by SN1088 (deployment period from July 2016 to September 2018). The recorder was moored at 808 m depth and scheduled to record continuously at a sample rate of 48,000 Hz. For the assessment of the song repertoire of bowhead whales within a one-year period, spectrograms were visually checked for the presence of bowhead whale songs, based on hourly presence information obtained from automated detection of bowhead whale vocalizations using the LFDCS ('Low-Frequency Detection and Classification System') software (Baumgartner and Mussoline, 2011, https://doi.org/10.1121/1.3562166), using a user-developed call library, an SNR ('signal-to-noise- ratio') threshold of 8 dB and a Mahalanobis distance threshold of 1.5. The term 'song' comprised both call sequences and true songs, thereby following the differentiation of Stafford et al. (2012; https://doi.org/10.3354/esr00444). Other bowhead whales signals recorded, such as constant calls, moans or down- and upsweeps that did not show any repetitive pattern, were not included in the song repertoire analysis. Only true songs and call sequences that were clearly distinguishable against the background noise and repeated at least three times within a day were considered for song repertoire analysis. Classification of songs was based on descriptive song characteristics, such as spectral structure of units, the arrangement of units and their frequency range. Song types were numbered in their order of first occurrence, and variants of the song types are ...