Acoustic monitoring is an essential tool for investigating animal communication and behavior when visual contact is limited, but the scalability of bioacoustic projects is often limited by time-intensive manual auditing of focal signals. To address this bottleneck, we introduce orcAI—a novel deep learning framework for the automated detection and classification of a broad acoustic repertoire of killer whales (Orcinus orca), including vocalizations (e.g., pulsed calls, whistles) and incidental sounds (e.g., breathing, tail slaps). orcAI combines a ResNet-based Convolutional Neural Network (ResNet-CNN) with Long Short-Term Memory (LSTM) layers to capture both spatial features and temporal context, enabling the model to classify signals and to accurately determine their temporal boundaries in spectrograms. Trained on a comprehensive dataset from herring-feeding killer whales off Iceland, the framework was designed to be adaptable to other populations upon training with equivalent data. Our final model achieves up to 98.2% accuracy on test data and is delivered as an open-source tool with an easy-to-use command-line interface. By providing a ready-to-use model that processes raw audio and outputs annotations, orcAI serves as a useful tool for advancing the study of killer whale vocal behavior and, more broadly, for understanding marine mammal communication and ecology.
Eumetopias jubatus (Steller sea lions) that breed on the Oregon coast are generally smaller than those found in Alaska. During the breeding season, male Zalophus californianus (California sea lions) move past E. jubatus rookeries in Oregon on their way to breeding rookeries in Southern California. Introgression by male Z. californianus could potentially explain the difference in size of Oregon E. jubatus relative to other E. jubatus populations. To assess whether morphologically detectable hybrids occur between Z. californianus and E. jubatus in Oregon, we used traditional skull morphometrics and Principal Component Analysis to identify individuals with intermediate morphological characteristics. Multiple range comparisons were used to determine which groups differed significantly. The comparisons revealed significant cranial morphological variation within each species' clusters and identified two male Z. californianus that overlapped with the male E. jubatus cluster, as well as two female E. jubatus that overlapped with the male Z. californianus cluster, suggesting potential hybrid status. These results support the possibility that long-term introgression between these two species may account for the observed size difference of the Oregon and Alaskan E. jubatus.
Monitoring live and stranded cetaceans can be expensive and logistically challenging, resulting in knowledge gaps. Very high-resolution (VHR) optical satellites are considered a potential solution to addressing some of these gaps. Despite success at detecting live and stranded cetaceans, satellites have only been trialed on restricted spatiotemporal scales. This project presents a framework for assessing the feasibility of using VHR optical satellite-based monitoring of cetaceans at high temporal frequency and local to global scales, focusing on the UK and UK Overseas Territories as a case study. We assess the primary environmental conditions necessary for the successful application of this technology: cloud cover and wind speed. Five-year monthly median “Total cloud cover” and “10m wind speed” ERA5 global reanalysis data were analyzed to map the spatial feasibility of satellite monitoring. We found that for the United Kingdom, VHR optical satellites could complement existing monitoring methods to achieve greater spatial and temporal coverage of live cetacean surveys, particularly, offshore, during the boreal spring and summer. However, satellites cannot address gaps in UK live cetacean monitoring in winter due to high wind speeds reducing whale detection probability. Based on environmental conditions, the tropics hold the greatest promise for achieving year-round satellite-based cetacean monitoring. In the Falkland Islands, particularly, the remote, unpopulated coastlines of West Island, satellites have the potential to improve strandings monitoring, opportunistically complementing existing stranding monitoring efforts.
Baleen whales are among the largest animals ever and engineer marine ecosystems by transporting nutrients both vertically through the water column and across vast oceanic distances. Hidden underwater, their feeding habits often remain unseen and hence must be studied indirectly based on stomach contents and stable isotopes. Here, we explore the potential of bone-bound calcium isotopes as a chemical proxy for extant baleen whale foraging ecology. Unlike bulk nitrogen and various trace elements (Mg, Sr, and Ba), calcium isotopes clearly distinguish gulp (n = 4 species) from skim feeders (n = 4 species) and provide a powerful tool to infer diet and feeding strategies. The difference in calcium isotope composition plausibly reflects incidental ingestion of seawater, which may be more pronounced in skim feeding whales than previously thought.
Ringed seals (Pusa hispida) in Svalbard, Norway, are suspected to be declining due to a significant reduction in land-fast sea ice, which serves as an essential breeding habitat, but were last surveyed in 2002. We address this data gap by conducting Uncrewed Aerial Systems (UAS) aerial surveys throughout Isfjorden (including adjacent small fjords) on the island of Spitsbergen, in the Svalbard Archipelago, during the molting season in 2023. We also provide an assessment of the size/age structure of the ringed seal population within Isfjorden using body lengths measured from aerial images. We show a decrease of 77% in land-fast sea ice area and a concomitant decline in ringed seal abundance of approximately 46% since 2002. During the same period, the density of hauled-out seals increased from 1.86 seals/km2 in 2002 to 2.41 seals/km2 in 2023. Length-based estimates of age suggest that 65% of the contemporary population consists of immature individuals. Together, the results from this updated population survey underscore the need for more frequent monitoring of ringed seals in Svalbard to better understand the drivers of population decline and highlight the broader implications of environmental change on Arctic marine biodiversity. The innovative use of UAS technology in this study demonstrates its utility in wildlife research, offering a less invasive, more cost-effective, efficient, and safer method for surveying marine mammals in polar regions compared to crewed aircraft.
Photo-identification (photo-ID) is a widely used, non-invasive method for monitoring individual animals, including humpback whales (Megaptera novaeangliae; HBWs), and has provided valuable insights into their population dynamics, movement patterns, and social structures. Traditional identification relies on the trailing edge and ventral pigment patterns of the tail fluke (fluke-ID); however, not all whales present their flukes, limiting identification and re-sighting opportunities. We developed a novel aerial-identification (aerial-ID) approach using drone imagery to identify individual HBWs based on the arrangement of two features, tubercles (TB) and cookiecutter shark scars (CCS). Between January and March 2022, we sampled 1498 HBWs, including repeated individuals, capturing fluke-ID images for 772 and aerial-ID images for 1437. Fluke-ID yielded 164 re-sightings (76 lactating females, 88 others), while aerial-ID yielded 372 (249 and 123, respectively), representing a 227% increase for lactating females and 40% for others. We extended this approach to a multi-year, cross-regional dataset (2018–2025) of 54 individuals verified with fluke-ID. All were matched using aerial-ID, with the longest re-sight spanning 2737 days (6.5 years), representing the maximum interval within our study period. Aerial-ID thus offers a powerful complement to fluke-ID, expanding demographic coverage, increasing re-sighting rates, and enabling long-term, cross-regional monitoring.
Goose-beaked whale (Ziphius cavirostris), a cosmopolitan odontocete, is commonly found in the Mediterranean Sea. This study, the first to examine the stomach contents of the goose-beaked whale in the Aegean Sea and Levantine Basin, reveals key aspects of its diet. The stomach contents of one whale from Seferihisar, Türkiye, in 2016, and five from a mass stranding on Cyprus in February 2023 were analyzed. A total of 1,448 lower beaks from 16 cephalopod species were identified, with six species—Octopoteuthis sicula, Ancistrocheirus lesueurii, Histioteuthis reversa, Chtenopteryx sicula, Chiroteuthis veranii, and Todarodes sagittatus—comprising the primary diet. Four species—Loligo forbesii, Pyroteuthis margaritifera, Pterygioteuthis giardi, and Brachioteuthis riisei—were recorded for the first time in goose-beaked whale stomachs. Their prey preference is generally pelagic cephalopods, though benthic or demersal species are occasionally eaten. The high species diversity and the presence of species not found in other studies may be due to differences in cephalopod fauna characterized by seabed and other oceanographic features of the eastern Mediterranean. The Mediterranean population is genetically divided into western and eastern subpopulations, and their diets appear to differ.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：