Humpback Whale Song Research Articles

Comparing the underwater soundscape of the Hawaiian Islands Humpback Whale National Marine Sanctuary and potential influences of the COVID-19 pandemic

Passive acoustic monitoring is an effective technique for long-term monitoring of the soundscape in marine protected areas. Ocean noise is a key concern for the U.S. Office of National Marine Sanctuaries and has been identified as a research priority. The Sanctuary Soundscape Monitoring Project (“SanctSound”) was implemented to support efforts to address ocean noise across seven U.S. sanctuaries using a comprehensive and standardized approach. In this study, acoustic recordings were collected in the Hawaiian Islands Humpback Whale National Marine Sanctuary during the humpback whale seasons (November-May) from 2018-2022. Data encompassed 14 deployments across four sites in the main Hawaiian Islands: Hawaiʻi, Maui, O’ahu, and Kauaʻi. The soundscape was dominated by biological sources, most prominently the seasonal detection of humpback whale song. Third octave level monthly medians ranged from 70.4-105 dB re 1 µPa across sites with distinct peaks from January to April particularly at both Hawaiʻi and Maui sites. Overall, we reported relatively low vessel detection rates, with Maui having the highest daily average of vessel detections (x = 19.16). No COVID-19 impact could be observed acoustically using soundscape metrics which was likely due to the dominance of humpback whale chorusing. However, vessel detections and AIS data revealed a reduction in vessel activity after the onset of the pandemic at the Maui and Hawaiʻi sites. This study demonstrates that standardized metrics are a useful tool for obtaining long-term, baseline soundscape levels to understand the various contributions to the underwater soundscape and potential changes within marine protected areas in Hawaiʻi.

Development of a machine learning detector for North Atlantic humpback whale song.

The study of humpback whale song using passive acoustic monitoring devices requires bioacousticians to manually review hours of audio recordings to annotate the signals. To vastly reduce the time of manual annotation through automation, a machine learning model was developed. Convolutional neural networks have made major advances in the previous decade, leading to a wide range of applications, including the detection of frequency modulated vocalizations by cetaceans. A large dataset of over 60 000 audio segments of 4 s length is collected from the North Atlantic and used to fine-tune an existing model for humpback whale song detection in the North Pacific (see Allen, Harvey, Harrell, Jansen, Merkens, Wall, Cattiau, and Oleson (2021). Front. Mar. Sci. 8, 607321). Furthermore, different data augmentation techniques (time-shift, noise augmentation, and masking) are used to artificially increase the variability within the training set. Retraining and augmentation yield F-score values of 0.88 on context window basis and 0.89 on hourly basis with false positive rates of 0.05 on context window basis and 0.01 on hourly basis. If necessary, usage and retraining of the existing model is made convenient by a framework (AcoDet, acoustic detector) built during this project. Combining the tools provided by this framework could save researchers hours of manual annotation time and, thus, accelerate their research.

Acoustic diversity, bearing estimation, and localization of sound sources in shallow water and deep ocean off the U.S. Northeast coast using a coherent hydrophone array

An in-house developed 160-element large-aperture hydrophone array from Northeastern University was utilized for ocean monitoring in both the shallow waters of the Great South Channel (GSC) and the deep-water area off the continental slope south of Rhode Island in September 2021. Utilizing passive ocean acoustic waveguide remote sensing (POAWRS) technology, the system enabled real-time, broad-area surveillance of the oceanic environment. A novel multistage clustering was introduced for annotating the vast volume of underwater acoustic data, drawing inspiration from the human cognitive process of categorizing sounds. Initially, sounds are separated into broad groups, akin to a first human listening pass. Subsequent stages of clustering refine these categories, utilizing different sets of features to achieve a granularity of annotation that closely mirrors expert human annotation processes. It was found that the acoustic landscape of the GSC was primarily composed of diverse marine mammal sounds, such as fin whale 20 Hz pulses, humpback whale songs, minky whale buzz sequences, sperm whale and dolphin echolocation clicks, and unidentified whale calls. Conversely, the deep-water area off the continental slope featured predominantly fish sounds, chorus-like sounds, and dolphin vocalizations. Detailed analyses of the most prevalent vocalizations are presented, including their bearing-time trajectories and localizations.

Humpback whale song vocalization behavior and temporo-spatial distributions in the Norwegian and Barents Sea observed with a coherent hydrophone array

The vocalization behavior of humpback whales in the Norwegian and Barents Seas is examined based on recordings of a large-aperture, densely-populated coherent hydrophone array system. The passive ocean acoustic waveguide remote sensing (POAWRS) technique is employed to provide detection, bearing-time estimation, time-frequency characterization and classification of the humpback whale vocalizations. The song vocalizations, composed of highly structured and repeatable set of phrases, were detected throughout the diel cycle between February 18 to March 8, 2014. The beamformed spectrograms of the detected humpback vocalizations are classified as song sequences based on inter-pulse intervals and time-frequency characteristics, verified by visual inspection. The song structure is compared for humpback whale vocalizations recorded at three distinct regions off the Norwegian coast, Alesund, Lofoten and Northern Finmark. Multiple bearing-time trajectories for humpback songs were simultaneously observed indicating multiple singers present at each measurement site. Humpback whale received call rates and temporo-spatial distributions are compared across the three measurement sites. Geographic mapping of humpback whale calls from their bearing-time trajectories is accomplished via the moving array triangulation technique.

Linear discriminant analysis based hidden Markov model for detection of Mysticetes’ vocalisations

Mysticetes’ produce distinctive vocalisations which are used for echolocation, communication, and other marine functions. These cryptic vocalisations are studied by marine scientist to determine the behavioural patterns and movement of this suborder of cetaceans within their ecosystem. In practice, these vocalisations are gathered using passive acoustic monitoring over days, weeks, months, and even years. Therefore, it is complex to study these sounds using traditional visual inspection techniques because the gathered datasets are huge. Machine learning (ML) tools such as Gaussian mixture models (GMMs), support vector machines (SVMs), and hidden Markov models (HMMs) have been adopted in recent times to proffer analytic solutions to automatically detect and study these cryptic vocalisations. Notwithstanding, the feature extraction techniques employed play a vital role in determining the performance of these ML tools. In most cases, the performance of the feature extraction technique is directly proportional to the performance of the ML tools. Thus, the method of linear discriminant analysis (LDA) is introduced in this article as a feature extraction technique that can be adapted with the HMMs (LDA-HMM) to seamlessly detect the vocalisations of Mysticetes. The performance of the proposed LDA-HMM detector is compared with other recent detectors for Mysticetes’ vocalisations in the literature using two different species: Humpback whale songs and Bryde’s whale pulses. Experimental results show that the developed LDA-HMM detector is a performance-efficient alternative in comparison to the recent detection techniques studied in this article. Besides, the LDA-HMM detector offers less computational time complexity; as such, it is more suitable for real-time applications.

Diel spatio-temporal patterns of humpback whale singing on a high-density breeding ground.

Humpback whale song chorusing dominates the marine soundscape in Hawai'i during winter months, yet little is known about spatio-temporal habitat use patterns of singers. We analysed passive acoustic monitoring data from five sites off Maui and found that ambient noise levels associated with song chorusing decreased during daytime hours nearshore but increased offshore. To resolve whether these changes reflect a diel offshore-onshore movement or a temporal difference in singing activity, data from 71 concurrently conducted land-based theodolite surveys were analysed. Non-calf pods (n = 3082), presumably including the majority of singers, were found further offshore with increasing time of the day. Separately, we acoustically localized 217 nearshore singers using vector-sensors. During the day, distances to shore and minimum distances among singers increased, and singers switched more between being stationary and singing while travelling. Together, these findings suggest that the observed diel trends in humpback whale chorusing off Maui represent a pattern of active onshore-offshore movement of singers. We hypothesize that this may result from singers attempting to reduce intraspecific acoustic masking when densities are high nearshore and avoidance of a loud, non-humpback, biological evening chorus offshore, creating a dynamic of movement of singers aimed at increasing the efficiency of their acoustic display.

Abrupt change in humpback whale song from Brazil suggests cultural revolutions may occur in the South Atlantic

Abrupt change in humpback whale song from Brazil suggests cultural revolutions may occur in the South Atlantic

First song description of the humpback whales, Megaptera novaeangliae (Balaenopteridae: Artiodactyla), breeding off Nicaragua

Introduction: Humpback whales belonging to the Central American (CA) Distinct Population Segment breed off the Pacific coast of Nicaragua. Knowledge on this endangered population and its conservation status is limited.
 Objective: The aim of this study is to provide the first description of the CA humpback whale song off Nicaragua, which helps further understanding on the population’s dynamics.
 Methods: Acoustic recordings of songs were obtained during dedicated boat-based surveys at two locations on the Pacific coast of Nicaragua in 2018. Recordings were made from the boat using a portable system for a total of 23 hours and 56 minutes over 32 days from January to April 2018. A total of nine recordings of high enough quality for the song analysis were identified during this period from three different days at Padre Ramos (PR) (northern site) and four different days at San Juan del Sur (southern site). Song structure was described using standard humpback whale song elements, i.e. themes, phrases, and units.
 Results: A total of seven themes, seven phrases, and 19 unit types were identified. Three of the themes were common and frequently repeated in a song cycle while the others were less common in the repertoire and were recorded only during the middle of the season. Song theme order was variable, both within and across song sessions.
 Conclusions: This study provides the first song description of humpback whales along the Pacific coast of Nicaragua. Comparison of these songs with other datasets from the CA population and other breeding areas in the Northern Hemisphere could help understand the migratory patterns of these animals and the level of connectivity among populations since song can be socially learnt. Future data collection of humpback whale song recordings in Nicaragua is necessary to gain further understanding on the song structure variation within this population and the mechanisms of song transmission and dynamics across populations in the region.

Humpback whale: Yes/no? Challenges with passive acoustic monitoring of this well-described species

Humpback whales sing song and produce non-song sounds. Dozens of research articles have described their acoustic repertoire—even if only considering Australian waters. While song structure changes slowly over seasons, individual notes may exhibit multi-decadal stability. Irrespectively, automatically detecting humpback whale vocalizations remains an ongoing development activity while manual validation of automated detections continues to be problematic. While analysing archival recordings from Australian waters using the NOAA/Google convolutional neural network pretrained for humpback whale song [https://tfhub.dev/google/humpbackwhale/1], we found numerous instances of a particular type of sound with spectrographic features resembling those of humpback whale vocalizations. These sounds lacked the context of a song and occurred in chorus-like patterns, withthe same sound recorded at varying received levels and overlapping in time—indicative of many animals simultaneously producing nothing but this specific call. To investigate these phenomena further, we (1) built a custom detector for these specific calls, (2) ran the detector on archival recordings from North-western Australia, and (3) compared geographic and seasonal presences and absences to known patterns of humpback whale occurrence. The likelihood of these scenarios being humpback whales is discussed.

Does female choice for song complexity drive sexual selection in humpback whales?

Male humpback whales (Megaptera novaeangliae) sing a long, stereotyped, and culturally transmitted song display. While song likely functions in sexual selection, whether it is primarily directed at females and/or males is still debated. Most males within a single population sing the same, slowly evolving song type. However, multiple humpback whale song revolutions (where a song introduced from a neighboring population rapidly and completely replaces the existing song) have spread across the South Pacific region from the east coast of Australia to French Polynesia. How cultural evolution and sexual selection interact in this complex vocal display is not well understood. We investigated song-mediated female mate choice in humpback whales using 20 years of paternity data and song recordings from individually identified males in the small, South Pacific population of New Caledonia. As mating has never been observed in humpback whales, we assessed male reproductive success (calf paternity) and song complexity (unit repertoire and song length) of individual males. We found that males who successfully sired offspring had on average higher song complexity than males who did not sire offspring. This suggests females may select mates at least partially based on song complexity, and thus, song being primarily directed at females.

Building bridges: Doug Cato’s remarkable career spanning underwater acoustics and animal bioacoustics

Doug Cato’s research includes pioneering studies of the underwater acoustic environment surrounding Australia and numerous contributions to our understanding of marine mammal acoustic behavior. Moreover, his relaxed avuncular manner allowed him to assemble collaborations spanning disparate disciplines while mentoring the next generation of acoustics researchers in Australia. Throughout his career, Doug built bridges between underwater acoustics and animal bioacoustics, and between Australian and international researchers. This talk highlights two international collaborations applying signal processing techniques to humpback whale songs and leopard seal calling bouts. These projects estimated the information entropy and structure of the sequences of sounds in these marine mammal vocalizations. Comparing sliding window entropy estimates with Markov models supports Doug’s earlier observations that Australian humpback song follows the hierarchical structure proposed by Payne & McVay for North Atlantic humpbacks [Miksis-Olds et al., JASA, 2008]. The same comparison of entropy estimators for leopard seal calling bouts reveals that most of the sequential structure of in these animals’ calls is captured by Markov models.

Arabian Sea Humpback Whale (Megaptera novaeangliae) Singing Activity off Netrani Island, India

Arabian Sea humpback whales (Megaptera novaeangliae), listed as “Endangered” on the International Union for Conservation of Nature’s Red List, remain resident throughout the year in the waters of the Arabian Sea and constitute a genetically isolated population. In the eastern Arabian Sea, information on humpback whales off the Indian coast has largely been limited to stranding records, local ecological knowledge, and opportunistic visual sighting data. These data, along with information from a long-term study off Oman, suggest that humpback whales migrate across the Arabian Sea into Indian territorial waters from October to March. To study the presence of Arabian Sea humpback whales in Indian waters more comprehensively, passive acoustic monitoring (PAM) was initiated along the west coast of India in 2019. Male humpback whales produce complex songs with a stereotyped structure; these songs are shared within a population, and song patterns are known to evolve progressively over time. In this article, a structural analysis of humpback whale song recorded over four days off the coast of Netrani Island, Karnataka, India, in December 2019 is presented. Time-frequency features of 2,641 individual call units were analysed. Call units had a fundamental frequency bandwidth ranging from 149.98 to 541.65 Hz, with a duration ranging from 1.19 to 5.5 s. The call units were used to identify phrases and themes required to construct the structure of the song, which can potentially help identify the population to which singing individuals belong. This study indicates the need for a long-term PAM program across the Arabian Sea to compare whale songs across the region. Simultaneous recordings over multiple seasons will best assess population connectivity, seasonal occurrence, and movement patterns within and between populations across the Arabian Sea and the Indian Ocean.

Variability in humpback whale songs reveals how individuals can be distinctive when sharing a complex vocal display.

Individually distinctive acoustic signals in animal vocal communication are taxonomically widespread, however, the investigation of these signal types in marine mammals has focused only on a few species. Humpback whale songs are a stereotyped, hierarchically structured vocal display performed by males, and hence thought to be sexually selected. Within a population, whales conform to a common version of the song despite the song constantly evolving. While humpback songs have been studied extensively at the population level, individual level variation has been rarely described, with inconclusive results. Here, we quantified inter- and intra-individual variability at different levels in the song hierarchy using songs from 25 singers across two song types from the eastern Australian population song of 2002 (12 singers), and the revolutionary song introduced in 2003 (13 singers). Inter-individual variability was found heterogeneously across all hierarchical levels of the song structure. In addition, distinct and individually specific patterns of song production were consistently recorded across song levels, with clear structural differences between the two song types. These results suggest that within the constraints of song conformity, males can produce individually distinctive patterns that could function as an advertisement to females to convey individual qualities.

Humpback whale song variations in La Reunion Island from 2016 to 2018 and in Madagascar in 2018

AbstractMale Megaptera novaeangliae produce complex and structured songs which are shared at the population level. Song patterns are culturally transmitted and evolve progressively through time, both over the breeding season and among years. The songs also undergo periods of relatively rapid change, termed “revolutions.” Acoustic monitoring was conducted from 2016 to 2018 in Reunion and throughout 2018 in Madagascar to assess spatiotemporal variation in song structures and population connectivity. A total of 46 high‐quality song samples were selected, representing 2,760 min of recordings in Reunion. In Madagascar, 12 samples representing 240 min of recordings were analyzed. Analysis of songs revealed 11 phrases and their variants. Low levels of temporal variations were observed over the breeding season. Songs recorded in June were very similar to those recorded in September. Greater variation was observed between years, and some phrases identified in 2018 may have evolved from phrases recorded in 2017. More variants were described for each phrase type in 2018 compared to 2016. All themes recorded in Reunion were shared with Madagascar, suggesting a high degree of population connectivity during the breeding season.

Adaptive signal adjustments in humpback whale songs

The use of vocal flexibility by humpback whales (Megaptera noavaengliae) often described as “singing” is a highly sophisticated communicative process, possessing a multilayered hierarchical structure across the various acoustical parameters. A deeper spectral analysis of songs made available through Google's Pattern Radio website studied in the month of February 2015 display several adaptive signal adjustments in humpback whale songs. As single singer variations in song properties become more clear, multi-singer variations in song properties remain hypothetical. This raises an important question of incorporating a temporal analysis approach to determine the multi-singer variations in song properties.

Source separation and arrival angle estimation using a single acoustic vector sensor

Extracting individual time signatures from two angularly-separated sources overlapping in time and bandwidth typically requires coherent array processing on multiple hydrophones, a process whose performance is degraded by limited aperture and spatial aliasing. Vector sensors, which measure both acoustic pressure and particle velocity, can measure an acoustic field’s directionality in a single, compact sensor package. Standard beamforming techniques can be applied to the outputs of a vector sensor to suppress one source’s signal, but its efficacy is limited by the angular insensitivity of the resulting cardioid beam pattern. Here we use fundamental physical principles to demonstrate that measurements on a 2-D vector sensor can completely determine the azimuths, amplitudes, and relative phases of two horizontal plane waves of the same frequency, achieving source separation and direction-of-arrival estimates for both. Interference between the waves creates a reactive intensity whose orientation is perpendicular to the bisector of the two wavenumber vectors. This observation leads to closed-form inversion formulas. Data examples of the technique are illustrated using overlapping humpback whale songs off Hawaii, and on weak airgun and bowhead whale signals embedded within intense directional noise from a nearby drilling site in the Beaufort Sea. [Work sponsored by ONR TFO.]

Regularities in spectral entropy variations within humpback whale song sessions

Consecutive songs produced by humpback whales recorded off the coast of Hawaii show recurrent patterns of acoustic variation. Analyses of spectral entropy, a measure of the uniformity of energy distribution in the frequency domain, revealed that spectral entropy cycles in predictable patterns within song sessions. At time scales spanning tens of minutes, variations in spectral entropy appear to track changes in received intensity levels, which are correlated with song duration and may reflect the diving behavior of singers. Intra-individual variations in spectral entropy suggest that singing humpback whales may also vocally control entropic features of songs, both by gradually modulating the acoustic characteristics of units and by varying the amount of time they spend producing specific patterns of units. Preliminary analyses suggest that variations in spectral entropy within song cycles are indicative of the amount of time and energy singers spend producing constant-frequency (CF) elements versus more frequency-modulated/broadband elements. CF elements of units are generally more difficult to spatially localize, but easier to detect over long distances, suggesting that singing humpback whales may be varying the production of consecutive songs in ways that affect auditory spatial processing.

Status and distribution of humpback whales (Megaptera novaeangliae) in northern Luzon, Philippines

Since the verification of the occurrence of humpback whales in the Babuyan Islands, Philippines in 1999, studies have been carried out on their status and distribution. Boat-based surveys of the waters around the islands were conducted from 2000-03. In 2003, a survey was also conducted off the eastern coast of Northern Sierra Madre. Once the whales were sighted, the location, group composition and behaviour were noted, fluke photo-identifications taken, sloughed skin and biopsy samples collected and songs recorded. A total of 367 humpback whale sightings were documented in four seasons around the Babuyan Islands. However, a cow-calf pair sighting off the coast of Northern Sierra Madre indicates that this breeding area may extend farther south. A total of 69 individuals have been photo-identified from 1999- 2003. Re-sightings across years are as follows: one individual was sighted for three consecutive years (2001-03); two individuals for two consecutive years (2001-02 and 2002-03); while three individuals were sighted in two different years (1999 and 2002; 2001 and 2003). Twelve of the 40 individuals photo-identified in the Philippines from 1999-2002 were matched to humpback whales identified in Ogasawara and Okinawa, Japan, including one whale that moved between these regions in one season. These matches indicate that the Philippine humpback whales are part of this greater western Pacific population. A subjective comparison between humpback whale songs recorded in the Philippines and Hawaii in 2002 indicated marked similarities. A total of nine distinctive themes were identified; seven of these were common to songs from both regions. The similarity in songs suggests humpback whales in the Philippines mix to some degree, at least acoustically, with the whales in Hawaii and the rest of the Pacific basin.

The occurrence of humpback whales across the Hawaiian archipelago revealed by fixed and mobile acoustic monitoring

The occurrence of humpback whales (Megaptera novaeangliae) across the 2600 km of Hawaiian archipelago, which include the remote atolls, banks, and seamounts of Papahānaumokuākea Marine National Monument (PMNM), remains poorly understood. Previous surveys for humpback whales beyond the main Hawaiian Islands have been scarce due to limited access and the challenging winter conditions typically found in PMNM when whales are present. To overcome these limitations, a combination of moored acoustic recorders and a Wave Glider autonomous surface vehicle were used to acoustically monitor eight locations and survey approximately 1500 km of the Hawaiian archipelago for the occurrence of humpback whale song during the 2019-2020 breeding season. Relative song prevalence was established using a machine learning tool and by quantifying the level of song chorusing. A generalized additive model framework was applied to understand the associations between habitat variables and humpback whale song occurrence, and sound propagation modeling was performed to examine whether acoustic propagation influenced observed patterns. Whale song was recorded at all monitored and surveyed locations across the archipelago, albeit in varying amounts. Among the locations monitored with moored recorders, the highest and most sustained seasonal chorusing levels were measured off Maui followed by French Frigate Shoals (Kānemilohai), Hawaii Island, Middle Bank, Oahu, Kauai, Gardner Pinnacles (Pūhāhonu) and Pearl and Hermes Reef (Holoikauaua), respectively. The Wave Glider mission to PMNM revealed that song prevalence was highest at Middle Bank and gradually decreased further to the northwest, reaching a minimum at Gardner Pinnacles (Pūhāhonu). However, song occurrence increased again at Raita Bank, remaining high between Raita Bank and the Northampton Seamounts. The results reveal that nearly the entire Hawaiian archipelago is exploited by humpback whales during the winter and early spring months. Moreover, song occurrence patterns suggest that there may be more structure in the distribution of whales in PMNM than previously known and raises questions about whether multiple subpopulations occur across the archipelago.

Soft-output signal detection for cetacean vocalizations using spectral entropy, k-means clustering and the continuous wavelet transform

Underwater passive acoustic monitoring systems record many hours of audio data for marine research, making fast and reliable non-causal signal detection paramount. Such detectors assist in reducing the amount of labor required for signal annotations, which often contain large portions devoid of signals.Cetacean vocalization detection based on spectral entropy is investigated as a means of vocalization discovery. Previous techniques using spectral entropy mostly consider time–frequency enhancement of the entropy measure, and utilize the short time Fourier transform (STFT) as its time–frequency (TF) decomposition. Spectral entropy methods also requires the user to set a detection threshold manually, which call for knowledge of the produced entropy measures.This paper considers median filtering as a simple, effective way to provide temporal stabilization to the entropy measure, and considers the continuous wavelet transform (CWT) as an alternative TF decomposition. K-means clustering is used to determine the threshold required to accurately separate the signal/no-signal entropy measures, resulting in a one-dimensional, two-class classification problem. The class means are used to perform pseudo-probabilistic soft class assignment, which is a useful metric in algorithmic development. The effect of median filtering, signal-to-noise ratio and the chosen TF decomposition are investigated.The accuracy and specificity measures of the proposed detection technique are simulated using a pulsed frequency modulated sweep, corrupted by a sample of ocean noise. The results show that median filtering is particularly effective for low signal-to-noise ratios. Both the STFT and CWT prove to be effective TF analyses for signal detection purposes, each presenting with different advantages and drawbacks. The simulated results provide insight into configuring the proposed detector, which is compared to a conventional STFT-based spectral entropy detector using manually annotated humpback whale (Megaptera novaeangliae) songs recorded in False Bay, South Africa, July2021.The proposed method shows a significant improvement in detection accuracy and specificity, while also providing a more interpretable detection threshold setting via soft class assignment, providing a detector for use in development of adaptive algorithms.