Real-time detection and classification of underwater soundscape signals via a convolutional autoencoder

Cultured cetaceans

Different species of baleen whales display distinct spatial distribution patterns in the Scotia Sea during the austral summer. Passive acoustic and visual surveys for baleen whales were conducted aboard the RRS James Clark Ross in the Scotia Sea and around South Georgia in January and February 2003. Identified calls from four species were recorded during the acoustic survey including southern right (Eubalaena australis), blue (Balaenoptera musculus), fin (B. physalus) and humpback whales (Megaptera novaeangliae). These acoustic data included up calls made by southern right whales, downswept D and tonal calls by blue whales, two possible types of fin whale downswept calls and humpback whale moans and grunts. Visual detections included southern right, fin, humpback and Antarctic minke whales (B. bonaerensis sp.). Most acoustic and visual detections occurred either around South Georgia (southern right and humpback whales) or south of the southern boundary of the Antarctic Circumpolar Current (ACC) and along the outer edge of the ice pack (southern right, blue, humpback and Antarctic minke whales). Fin whales were the exception, being the only species acoustically and visually detected primarily in the central Scotia Sea, along the southern ACC front. In addition to identifiable calls from these species, two types of probable baleen whale calls were detected: 50Hz upswept and pulsing calls. It is proposed that minke whales may produce the pulsing calls, based on their similarities with minke whale calls recorded in the North Atlantic Ocean. There was an overlap between locations of fin whale sightings and recordings and locations of 50Hz upswept calls in the central Scotia Sea, but these calls were most similar to calls attributed to blue whales in other parts of Antarctica. More study is required to determine if baleen whales produce these two call types, and if so, which species. The efficiency of acoustics and visual surveys varied by species, with blue whales being easier to detect using acoustics, Antarctic minke whales being best detected during visual surveys and other species falling in between these two extremes.

To support optimal monitoring and enforcement investment, management aimed at minimizing disturbance to wildlife requires an understanding of how regulatory compliance might vary spatially as well as across species and human-user groups. In the Salish Sea, humpback whales (Megaptera novaeangliae) and two ecotypes (southern resident and Bigg's) of killer whales (Orcinus orca) now interact with a large and growing number of small commercial and recreational vessels that partake in whale watching. Those vessels often approach close to cetaceans and thus pose risk via collision, marine noise and pollution, exposure to which may result in disturbance, injury and death. The primary management tool for mitigating impacts is minimum distance regulations. Compliance, however, is poorly understood. We examined commercial and recreational small vessel compliance with viewing distances across two seasons (June–September 2018 and 2019) in over ≈404 h of on-water observation. Overall vessel compliance was nearly 80%, but several distinct patterns emerged. Recreational boats were significantly more likely to violate distance regulations and boaters were more likely to be non-compliant around killer whales. Compliance did not vary with day of week or time of day. Spatially, non-compliance was concentrated in waters closer to coastal communities. Collectively, these patterns suggest that optimal enforcement could be targeted to identify areas of high non-compliance, especially for killer whales, with effort spread across days and times. Finally, we discuss how investments in education could target recreational boaters at a time when multiple and interacting stressors are accumulating in the Salish Sea.

Pushing distributed acoustic sensing (DAS) methods from typical seismic frequency ranges into higher frequencies (500 Hz to 1 kHz) will greatly advance the tools available for broad spatial scales of marine mammal monitoring. Recordings from an Orcasound hydrophone stationed near a pre-existing fiber optic cable in Puget Sound were analyzed throughout the fall of 2024 for Southern Resident Killer Whale (SRKW) calls. Simultaneously, the cable was illuminated and interrogated to gather DAS data. The acoustic calls were corroborated with citizen science data of visual sightings to filter the calls spatially relevant to the DAS system. The hydrophone data was processed in a method designed to mimic the DAS system specifications. Determining the exact power spectra of each call for specific frequencies (0–1 kHz) provides a much stronger framework for identifying killer whale calls on pre-existing cables with unknown armoring or positioning. This study allows for exploration and expansion of DAS into higher frequencies to monitor SRKW and other marine mammals with limited fieldwork required compared to studies that may require laying a specialized cable. Ultimately, using hydrophone data will accelerate the movement of DAS into the marine mammal observation realm for species with higher-frequency vocalizations. [Work supported by Orcasound.]

Killer whales (Orcinus orca) have only infrequently been reported from Hawaiian waters, and most of what is known about killer whales worldwide comes from studies in coastal temperate waters. Here we present 21 records of killer whales from within the Hawaiian Exclusive Economic Zone between 1994 and 2004. Killer whales were recorded nine months of the year, most around the main Hawaiian Islands. Although there were more records than expected during the period when humpback whales are abundant around the Islands, there is likely an increase in sighting effort during that period. Killer whales were documented feeding on both a humpback whale and cephalopods, and two species of small cetaceans were observed fleeing from killer whales. Although it is possible that there are both marine mammal–eating and cephalopod-eating populations within Hawaiian waters, it seems more likely that Hawaiian killer whales may not exhibit foraging specializations as documented for coastal temperate populations. Saddle...

Understanding interactions between large ships and large whales is important to estimate risks posed to whales by ships. The coastal waters of Alaska are a summer feeding area for humpback whales (Megaptera novaeangliae) as well as a prominent destination for large cruise ships. Lethal collisions between cruise ships and humpback whales have occurred throughout Alaska, including in Glacier Bay National Park (GBNP). Although the National Park Service (NPS) establishes quotas and operating requirements for cruise ships within GBNP in part to minimize ship-whale collisions, no study has quantified ship-whale interactions in the park or in state waters where ship traffic is unregulated. In 2008 and 2009, an observer was placed on ships during 49 different cruises that included entry into GBNP to record distance and bearing of whales that surfaced within 1km of the ship's bow. A relative coordinate system was developed in ArcGIS to model the frequency of whale surface events using kernel density. A total of 514 whale surface events were recorded. Although ship-whale interactions were common within GBNP, whales frequently surfaced in front of the bow in waters immediately adjacent to the park (west Icy Strait) where cruise ship traffic is not regulated by the NPS. When ships transited at speeds >13 knots, whales frequently surfaced closer to the ship's midline and ship's bow in contrast to speeds slower than 13 knots. Our findings confirm that ship speed is an effective mitigation measure for protecting whales and should be applied to other areas where ship-whale interactions are common.

We examined the incidence of rake mark scars from killer whales Orcinus orca on the flukes of humpback whales Megaptera novaeangliae throughout the North Pacific to assess geo- graphic variation in predation pressure. We used 3650 identification photographs from 16 wintering or feeding areas collected during 1990 to 1993 to determine conservative estimates in the percentage of whales with rake mark scarring. Dramatic differences were seen in the incidence of rake marks among regions, with highest rates on wintering grounds off Mexico (26 vs. 14% at others) and feed- ing areas off California (20 vs. 6% at others), 2 areas between which humpback whales migrate. Although attacks are rarely witnessed, the prevalence of scars demonstrates that a substantial por- tion of animals are attacked, particularly those that migrate between California and Mexico. Our data also suggest that most attacks occur at or near the wintering grounds in the eastern North Pacific. The prevalence of attacks indicates that killer whale predation has the potential to be a major cause of mortality and a driving force in migratory behavior; however, the location of the attacks is inconsis- tent with the hypothesis that animals migrate to tropical waters to avoid predation. Our conclusion is that, at least in recent decades, attacks are made primarily on calves at the wintering grounds; this contradicts the hypothesis that killer whales historically preyed heavily on large whales in high- latitude feeding areas in the North Pacific.

Calls of killer whales, Orcinus orca, were analyzed using computed sound features to classify sound patterns and
\nidentify call similarties. Calls were classified and separated according to the podfamily group within clans identified
\npreviously by John Ford (U. BC) in the Vancouver whale populations. Acoustic characteristics of the same call type from
\ndifferent individuals were extremely similar, so that discriminating these different sounds was the goal. The WHOI
\nAcouStat program and associated database systems were used to define numerical statistics for each call, and then, these
\nwere compared to sort and classify the sounds. The results were in agreement with Ford's descriptions of the calls derived
\nfrom visual inspection of sound spectrograms of calls. The classification analyses demonstrated that although specific
\nshared calls from different killer whales were much alike, they could be sorted by the pod/subpod of the whales producing
\nthe calls. A typical analysis, for example, of the N4 call from Clan A (Vancouver, BC), classified 97% of the calls
\ncorrectly according to the pod/family of the whales producing the calls. Remaining calls were variant, and likely a result of
\nindividual differences in call sounds. Similar classification analysis were tested on unsorted, unalyzed recordings from
\ndifferent populations of whales, and these too could be distinguished, with 98.5% correct separation of the calls.

This work compiles a decade (2001-2010) of marine mammal sightings in the Malpelo Fauna and Flora Sanctuary (FFS) and the area between the island and mainland coast. Four separate sources of data were consulted, which used visual searching during cruising efforts while Malpelo’s surrounding waters were surveyed from a vantage point. Seven species were identified in the FFS: October and November were the months with higher species richness. Tursiops truncatus had the highest encounter rates (17.78 groups/100 h), followed by Megaptera novaeangliae (1.62) and Stenella attenuata (0.88). These species were usually within 6 km from the island. Other species seen around the island include Stenella coeruleoalba, Delphinus delphis, Stenella longirostris and Zalophus wollebaeki. On the other hand, thirteen species were identified during cruises, and March and April were the months with the highest species richness. Megaptera novaeangliae had the highest encounter rate (5.94), followed by T. truncatus (3.30), S. attenuata (3.08), D. delphis (3.08) and S. coeruleoalba, Globicephala macrorhynchus and Orcinus orca, each one with 0.66. Other species seen during the cruises include Steno bredanensis, Pseudorca crassidens, Grampus griseus, Peponocephala electra, Physeter macrocephalus and Ziphius cavirostris. Megaptera novaeangliae were associated to the continental shelf, T. trunctaus and D. delphis to oceanic waters, and S. attenuata to both the continental shelf and oceanic waters. Delphinus delphis was more abundant in intermediate waters, during the first trimester (January-March), while T. truncatus was the most abundant species around the Sanctuary during all seasons, suggesting that is the same population. The latter was also the only species found all year round in both zones (around the island and in oceanic waters), and encounter rates did not change across years. Megaptera novaeangliae had a seasonal presence (mostly June to November) with a higher abundance during the third trimester (July-September), both around Malpelo and the transect. The presence of humpback whales calves suggests that Malpelo is used for reproductive purposes. This new information about the marine mammals found in Malpelo FFS and the area between the island and the continent contributes to the understanding of these species in Colombian waters.

Comparing call-based versus subunit-based methods for categorizing Norwegian killer whale, Orcinus orca, vocalizations

The Cape Solander Whale Migration Study is a citizen science project that annually counts northward migrating humpback whales (Megaptera novaeangliae) off Cape Solander, Sydney, Australia. Dedicated observers have compiled a 20‐year data set (1997–2017) of shore‐based observations from Cape Solander's high vantage point. Using this long‐term data set collected by citizen scientists, we sought to estimate the humpback whale population trend as it continues to recover postexploitation. We estimated an exponential growth rate of 0.099 (95% CI = 0.079–0.119) using a generalized linear model, based on observer effort (number of observation days) and number of whales observed, equating to 10% per annum growth in sightings since 1997. We found that favorable weather conditions for spotting whales off Cape Solander consisted of winds <30 km/hr from a southerly through a north westerly direction. Incidental observations of other cetacean species included the endangered blue whale (Balaenoptera musculus) and data deficient species such as killer whales (Orcinus orca) and false killer whales (Pseudorca crassidens). Citizen science‐based studies can provide a cost‐effective approach to monitoring wildlife over the time necessary to detect change in a population. Information obtained from citizen science projects like this may help inform policy makers responsible for State and Federal protection of cetaceans in Australian waters and beyond.

This study presents a detailed seasonal comparison of the abundance and distribution of cetaceans within 100‐150 nmi (185‐278 km) of the California coast during 1991 and 1992. The results of a shipboard line‐transect survey conducted in July‐November 1991 (“summer”) were compared to those from aerial line‐transect surveys conducted in March‐April 1991 and February‐April 1992 (“winter”). Using a confidence‐interval‐based bootstrap procedure, abundance estimates for six of the eleven species included in the comparison exhibited significant (α= 0.05) differences between the winter and summer surveys. Pacific white‐sided dolphins (Lagenorhynchus obliquidens), Risso's dolphins (Grampus griseus), common dolphins (Delphinus spp.), and northern right whale dolphins (Lissodelphis borealis) were significantly more abundant in winter. The abundance of blue whales (Balaenoptera musculuss) and gray whales (Eschrichtius robustus) reflected well‐documented migratory patterns. Fin whales (B. physalus) were significantly more abundant during summer. No significant differences in seasonal abundance were identified for Dall's porpoises (Phocoenoides dalli), bottlenose dolphins (Tursiops truncatus), killer whales (Orcinus orca), sperm whales (Physeter macrocephalus), or humpback whales (Megaptera novaeangliae). Significant north/south shifts in distribution were found for Dall's porpoises, common dolphins, and Pacific white‐sided dolphins, and significant inshore/offshore differences were identified for northern right whale dolphins and humpback whales.

Data are presented on the distribution, measurements and weights, ecology, and other aspects of the biology of 12 species of cetaceans, one pinniped, and the manatee in Florida waters. Most of the records are for the period from 1953 through June 1963, although some for earlier years are included. Records of cetaceans based on strandings or specimens washed onshore, dead specimens seen or collected in offshore waters, and skeletal materials from beaches include a probable one of the little piked whale (Balaenoptera acutorostrata) 1 of the humpbacked whale (Megaptera novaeangliae), 3 of the goose-beaked whale (Ziphius cavirostris), 7 of the pygmy sperm whale (Kogia breviceps), 2 of the sperm whale (Physeter catodon), 1 of the rough-toothed dolphin (Steno bredanensis), 18 verified or probable records of the short-finned pilot whale (Globicephala macrorhyncha), 17 verified or probable records of the bottle-nosed dolphin (Tursiops truncatus), 2 of the common dolphin (Delphinus delphis), and 1 of the long-beaked dolphin (Stenella longirostris). The latter is the second record of this species in Florida waters and apparently the third for North America. Observations on living cetaceans include 16 and possibly 18 verified or probable records for the Atlantic right whale (Balaena g. glacialis), 1 for the humpback whale, 2 for the killer whale (Orcinus orca), 3 and a possible 4th for the short-finned pilot whale, and 10 for the bottle-nosed dolphin. The sight record of the humpback whale apparently constitutes the first record of this species in the Gulf of Mexico. External, skeletal measurements, or both, of Florida specimens of Megaptera, Kogia, Physeter, Steno, Globicephala, Tursiops, Delphinus, and Stenella longirostris are given. One pinniped record obtained was of the California sea lion. (Zalophus californianus), and another probably also refers to this species. Both undoubtedly involve animals escaped from captivity. One individual apparently moved approximately 100 miles from where it escaped in about 10 days . Twenty-five records of dead manatees (Trichechus manatus) and 5 sight records are summarized. Of the former, 10 individuals had apparently been killed by cold, 3 by boats, and 2 by shooting. The cause of death of the remainder is unknown, although the death of 7 individuals in the Fort Myers area in the spring of 1963 coincided with a red tide epidemic in the Gulf to the north. Further evidence of the occurrence of seasonal movements in Florida manatee populations is presented.

Many cetacean species are susceptible to mortality or serious injury from vessel collisions. Spatially explicit assessments of risk per whale can help identify potential problem areas to guide appropriate mitigation measures. Canada’s Pacific waters host high cetacean densities and intense maritime traffic, and the issue of vessel collisions has taken on a high priority in British Columbia (BC) recently due to several major industrial development applications. Spatially-explicit statistical modelling and Geographic Information System (GIS) visualisation techniques identified areas of overlap between shipping activity and waters used by humpback, fin and killer whales. Areas of highest risk were far removed from areas with highest concentrations of people, suggesting that many beach-cast carcasses could go undetected. With few exceptions, high-risk areas were found in geographic bottlenecks, such as narrow straits and passageways. Port expansion and a proposed pipeline for carrying oil from Alberta to BC’s north coast (with associated oil tanker traffic) would increase ship strike risk for all three species. The risk assessments illustrate where ship strikes are most likely to occur, but cannot estimate how many strikes occur. Propeller wounds on live killer whales are common in the region, and fatal collisions have been reported in BC for all three species. Procedures were used to estimate potential mortality limits in accordance with a wide range of quantitative management objectives from jurisdictions around the world. While the extent of under-reporting of ship strikes has not been evaluated, the few known cases of collisions involving fin whales suggest that mortality due to ship strike for this species may already be approaching or even exceeding mortality limits under the most risk-averse management objectives. It is hoped that risk maps may inform environmental impact assessments of marine traffic because it will be easier to plan new shipping lanes so that they avoid high-density areas for whales than it will be to move the lanes after they become entrenched.

Humpback whales wintering in the entire West Indies chain are widely treated as comprising a single breeding population. However, most areas outside of Silver Bank and Samana Bay, Dominican Republic, are poorly and sporadically studied. Data is presented on the timing and movement patterns of 262 whales from the southeastern Caribbean, extending from Antigua in the north to Trinidad and Tobago in the south. Whales from the area were re-sighted in all of the major North Atlantic feeding grounds. However, of the 43 individuals re-sighted in feeding areas, animals from eastern feeding grounds were significantly over-represented, while those from western feeding areas were under-represented. This is in direct contrast to the pattern previously demonstrated in the Dominican Republic. Supporting this finding, the proportion of whales showing visible scarring on the flukes from non-lethal attack by killer whales was similar to that previously shown for Norway; yet lower than that presented from western feeding areas. The seasonal pattern of distribution in the southeastern Caribbean shows a peak of occurrence about six weeks later than in the Dominican Republic, and there is little overlap in the periods of greatest use. Sightings are uncommon before February. The peak in abundance occurs during March and April, declining during May, with some sightings extending into June. This is consistent with the pattern of sightings from historical whaling records in the southeastern Caribbean. These results suggest that the humpbacks mating and calving in this region are not a representative subset of those that winter in the Dominican Republic. Further studies will be needed to examine the spatial nature of the pattern shown here and define the nature and limits of this group, but these results suggest that some part of this breeding area represents a previously un-described distinct population segment within the North Atlantic. Given this, the widely held idea that there is a single West Indies humpback whale distinct population segment is in need of reconsideration.