Search

Aerial images are increasingly adopted and widely used in various research areas. In marine mammal studies, these imagery surveys serve multiple purposes: determining population size, mapping migration routes, and gaining behavioral insights. A single aerial scan using a drone yields a wealth of data, but processing it requires significant human effort. Our research demonstrates that deep learning models can significantly reduce human effort. They are not only able to detect marine mammals but also track their behavior using continuous aerial (video) footage. By distinguishing between different age classes, these algorithms can inform studies on population biology, ontogeny, and adult-calf relationships. To detect beluga whales from imagery footage, we trained the YOLOv7 model on a proprietary dataset of aerial footage of beluga whales. The deep learning model achieved impressive results with the following precision and recall scores: beluga adult = 92%—92%, beluga calf = 94%—89%. To track the detected beluga whales, we implemented the deep Simple Online and Realtime Tracking (SORT) algorithm. Unfortunately, the performance of the deep SORT algorithm was disappointing, with Multiple Object Tracking Accuracy (MOTA) scores ranging from 27% to 48%. An analysis revealed that the low tracking accuracy resulted from identity switching; that is, an identical beluga whale was given two IDs in two different frames. To overcome the problem of identity switching, a new post-processing algorithm was implemented, significantly improving MOTA to approximately 70%. The main contribution of this research is providing a system that accurately detects and tracks features of beluga whales, both adults and calves, from aerial footage. Additionally, this system can be customized to identify and analyze other marine mammal species by fine-tuning the model with annotated data. [ABSTRACT FROM AUTHOR]

Very high-resolution (VHR) satellite imagery has proven to be useful for detection of large to medium cetaceans, such as odontocetes and offers some significant advantages over traditional detection methods. However, the significant time investment needed to manually read satellite imagery is currently a limiting factor to use this method across large open ocean regions. The objective of this study is to develop a semi-automated detection method using object-based image analysis to identify beluga whales (Delphinapterus leucas) in open water (summer) ocean conditions in the Arctic using panchromatic WorldView-3 satellite imagery and compare the detection time between human read and algorithm detected imagery. The false negative rate, false positive rate, and automated count deviation were used to assess the accuracy and reliability of various algorithms for reading training and test imagery. The best algorithm, which used spectral mean and texture variance attributes, detected no false positives and the false negative rate was low (<4%). This algorithm was able to accurately and reliably identify all the whales detected by experienced readers in the ice-free panchromatic image. The autodetection algorithm does have difficulty separately identifying whales that are perpendicular to one another, whales below the surface, and may use multiple segments to define a whale. As a result, for determining counts of whales, a reader should manually review the automated results. However, object-based image analysis offers a viable solution for processing large amounts of satellite imagery for detecting medium-sized beluga whales while eliminating all areas of the imagery which are whale-free. This algorithm could be adapted for detecting other cetaceans in ice-free water. [ABSTRACT FROM AUTHOR]

Beluga whales (Delphinapterus leucas) have a long history of subsistence harvest that has continued into the present and a history of exploitation through commercial harvest. The Eastern High Arctic – Baffin Bay (EHA) population of beluga whales has the northernmost distribution of any beluga whale population in Canada. Beluga whales from the EHA population have a single complete abundance estimate from 1996 and 3 partial abundance estimates in 1981, 2010, and 2012 from aerial surveys; reliable estimates from surveys are lacking in recent years, limiting the ability to determine population dynamics. In 2020 satellite imagery was used to estimate abundance for beluga whales in the EHA population for estuaries in the whales' summering area. We built a stochastic stock production model to estimate population dynamics from the start of reliably compiled harvest history data in 1977 to 2022, using abundance estimates from 1981, 1996, 2010, 2012, and 2020. We also extended the model 10 years into the future under 7 potential annual harvest scenarios (within the range of annual reported harvests) to calculate the probability of decline. We estimated the starting population in 1977 as 29,615 whales (95% CI = 20,765–46,251), and the estimate from 2022 was 16,495 whales (95% CI = 6,731–33,504). Landed catch of 0, 100, 200, 300, 400, 500, and 600 beluga whales annually resulted in <1%, 3%, 23%, 50%, 70%, 83%, and 90% probabilities of decline, respectively. There was a notable decrease in abundance over the time series, likely caused by harvest pressure. Harvest in recent years has ranged from 155–553 catches per year, and our results suggest that 108 landed catches per year equates to a 5% risk of decline, the low risk goal for precautionary fisheries management, which aligns with population conservation goals. Our model is a first step in understanding the EHA beluga whale population dynamics using what data are available and suggests that current harvests in some years may result in a population decline and should be monitored. [ABSTRACT FROM AUTHOR]

Assessing genetic structure and diversity in wildlife is particularly important in the context of climate change. The Arctic is rapidly warming, and endemic species must adapt quickly or face significant threats to persistence. Bowhead whales (Balaena mysticetus) and narwhals (Monodon monoceros) are two long‐lived Arctic species with similar habitat requirements and are often seen together in the Canadian Arctic. Although their ranges overlap extensively, bowhead whales experienced significantly greater commercial whaling mortality than narwhals over several centuries. The similar habitat requirements but different harvest histories of these two species provide an opportunity to examine present‐day genetic diversity and the demographic and genetic consequences of commercial whaling. We whole‐genome resequenced contemporary Canadian Arctic bowhead whales and narwhals to delineate population structure and reconstruct demographic history. We found higher genetic diversity in bowhead whales compared to narwhals. However, bowhead whale effective population size sharply declined contemporaneously with the intense commercial whaling period. Narwhals, in contrast, exhibited recent growth in effective population size, likely reflecting exposure to limited opportunistic commercial harvest. Bowhead whales will likely continue to experience significant genetic drift in the future, leading to the erosion of genetic diversity. In contrast, narwhals do not seem to be at imminent risk of losing their current levels of genetic variation due to their long‐term low effective population size and lack of evidence for a recent decline. This work highlights the importance of considering population trajectories in addition to genetic diversity when assessing the genetics of populations for conservation and management purposes. [ABSTRACT FROM AUTHOR]

The Cumberland Sound beluga whale (Delphinapterus leucas) population inhabits Cumberland Sound on the southeast side of Baffin Island, Nunavut. The population is listed as threatened under the Species at Risk Act. The last abundance estimate from an aerial survey was estimated at 1,381 (95% CI: 1,270-1,502) beluga whales in 2017 for an area covering 12,485 km². Since then, satellite imagery has been used as a remotely based non-invasive method to monitor these whales. Very High Resolution (VHR) satellite imagery covering 9,690 km² of water was collected from Cumberland Sound from August 30 to September 7, 2021, during the ice free season. Readers with previous imagery analysis experience analyzed the images and identified 704 certain detections. Abundance estimates were corrected for availability bias for whales that were too deep to be detected in the imagery (>2m). We present a total estimate of 1,690 (CV = 0.16; 95% CI: 1,241-2,301) beluga whales in Cumberland Sound (22,663 km²). This estimate covers a larger area and estimates a higher abundance than the 2017 aerial survey. Regular population abundance assessments are essential for understanding population dynamics and trends and we have shown here that satellite imagery is a comparable method to aerial surveys for estimating abundance. [ABSTRACT FROM AUTHOR]

Chemical composition of tissues can act as a biological tag to discriminate among groups of animals that inhabit different areas. In Canada, subsistence hunting of the Baffin Bay narwhal (Monodon monoceros Linnaeus, 1758) population is managed as stocks represented by summer aggregations. However, narwhals are highly mobile and are hunted during the migration while stocks mix. Thus, information that can help managers decipher the stock origin of hunted individuals to prevent overexploitation of animals adapted to particular summering grounds is needed. Stable isotope and trace element analyses were conducted on narwhal skin tissues from five stocks in the eastern Canadian Arctic from 1990 to 2015. Discriminant analysis showed a significant difference between Admiralty Inlet and Eclipse Sound stocks in the summer residency period and both differed from Jones Sound and Somerset Island. During the migration season, there was more overlap and less distinction among stocks, but 75% of animals were classified correctly to their defined stocks in both periods. Together stable isotope and trace element analyses are useful for delineating stocks and could be used to complement other stock discrimination approaches. Key words: Monodon monoceros, narwhal, trace elements, stable isotopes, [[delta].sup.13]C, [[delta].sup.15]N, discriminant analysis, stock management. La composition chimique des tissus peut servir d'étiquette biologique permettant de discriminer des groupes d'animaux habitant des régions différentes. Au Canada, la chasse de subsistance visant la population de narvals (Monodon monoceros Linnaeus, 1758) de la baie de Baffin est gérée en fonction de stocks représentés par des groupements estivaux. Les narvals sont cependant très mobiles et sont chassés durant la migration alors que les stocks se mélangent. Des renseignements pouvant aider les gestionnaires à déterminer le stock d'origine d'individus chassés sont donc nécessaires pour éviter la surexploitation d'animaux adaptés à des aires d'estivage données. Des analyses d'isotopes stables et d'éléments en traces ont été réalisées sur des tissus de peau de narvals prélevés de cinq stocks dans l'est de l'Arctique canadien de 1990 à 2015. L'analyse discriminante montre une différence significative entre les stocks du bras Admiralty et du détroit d'Éclipse en ce qui concerne la période de résidence estivale, et que ces deux stocks sont différents de ceux du détroit de Jones et de l'île Somerset. Durant la période de migration, il y avait plus de chevauchement et moins de différences entre les stocks, mais 75% des animaux ont été classés correctement dans leurs stocks définis durant les deux périodes. Combinées, les analyses d'isotopes stables et d'éléments en traces sont utiles pour délimiter les stocks et pourraient être utilisées comme compléments à d'autres approches de discrimination des stocks. [Traduit par la Rédaction] Mots-clés : Monodon monoceros, narval, éléments en traces, isotopes stables, [[delta].sup.13]C, [[delta].sup.15]N, analyse discriminante, gestion des stocks., Introduction Successful management of marine mammal populations relies on an understanding of their spatial distribution at all times of the year. Management goals of preserving the local abundance of marine [...]

Hunted animals are often managed as static management units, or stocks, specific to hunting regions. However, movement of animals between regions poses a particular challenge for management to ensure that the hunt of individual stocks is sustainable. The incorporation of genetic information in stock assessments can improve management decisions, but the resolution of genetics may not differentiate stocks, making the use of movement data necessary. The Joint Working Group of the North Atlantic Marine Mammal Commission (NAMMCO) and the Canada-Greenland Joint Commission on Conservation and Management of Narwhal and Beluga (JCNB) has developed a model that allocates catches in different hunting regions and seasons to different stocks based on movement data, local knowledge, and expert opinion. The model uses information on stock size, catches in different hunting areas/seasons, and a matrix which estimates the proportion of animals in each stock that are available to hunters in different regions and seasons. This matrix can be informed by quantitative data on stock structure (e.g., genetics, telemetry) or qualitative information (local knowledge, expert opinion, etc.). Uncertainty in the availability of animals and individual stock sizes is incorporated in a stochastic version. The model is presented using a case study of narwhals, which are managed as stocks based on their summer distribution in Canada and Greenland., Introduction Marine mammal populations or stocks often have large spatial distributions and follow seasonal migration patterns that expose them to taking by hunters or other risks at different times and [...]

The monodontids--narwhals, Monodon monoceros, and belugas, Delphinapterus leucas--are found in much of the Arctic and in some subarctic areas. They are hunted by indigenous subsistence users. In the past, some populations were substantially reduced by commercial hunting and culling; more recently, some populations have declined due to uncontrolled subsistence hunting and environmental degradation. Monodontids are impacted increasingly by human activities in the Arctic including ship and boat traffic, industrial development, icebreaking, seismic surveys, competition with fisheries, and alteration of habitat due to climate change. Since comprehensive reviews in the 1990's, substantial new information has become available on both species and on changes to their habitat as a result of human activities and climate change. Thus NAMMCO and partners undertook an updated review in 2017. The review recognized 21 extant beluga stocks, 1 extirpated beluga stock, and 12 stocks of narwhals. The available information on each stock regarding population size, depletion level, current and past removals, and trends in abundance was reviewed to determine status. Concern was expressed where the lack of information prevented reliable assessment, removals were thought to be unsustainable, or the population was deemed at risk of declining even without direct removals by hunting. Beluga stocks of greatest concern are the small stocks in Ungava Bay (possibly extirpated), Cook Inlet (ca 300), St. Lawrence Estuary (ca 900), and Cumberland Sound (ca 1,100), and the stocks with uncertainty in Eastern Hudson Bay and the Barents-Kara-Laptev Seas. Narwhal stocks of greatest concern are those in Melville Bay and East Greenland., Introduction The family Monodontidae is comprised of the narwhal, Monodon monoceros, and the beluga or white whale, Delphinapterus leucas (Rice, 1998). These two species are found in much of the [...]

Less than 1,500 beluga whales (Delphinapterus leucas) are estimated to be in Cumberland Sound. These whales, considered as a single population of beluga, have been listed as Threatened under the Species at Risk Act and recently assessed as Endangered by the Committee on the Status of Endangered Wildlife in Canada. This beluga whale group has been defined based on satellite tracking data, contaminants, morphometrics, and previous genetic analyses. However, questions regarding whether multiple populations of beluga whales visit Cumberland Sound, their genetic stock discrimination, and the timing of migration and distribution of whales from different groups within Cumberland Sound remain. A recent reexamination of a long haplotype of the mitochondrial DNA (mtDNA) control region showed greater discrimination among beluga whale populations in Eastern Canada, including a small subsample of whales harvested from Cumberland Sound. In this study, we reexamined the genetic distinctiveness of beluga whales hunted in Cumberland Sound (N = 208) compared with other Eastern Arctic whales (N = 657), analyzing all samples collected from this area with the long haplotype of mtDNA. We also genotyped a subsample of whales harvested in Cumberland Sound (N = 27) and Western Hudson Bay (N = 121) using 12,370 nuclear DNA (nDNA) single nucleotide polymorphisms (SNPs) to investigate their distinctiveness. Our results using mtDNA confirmed that approximately 35% of beluga whales harvested in July and August from Cumberland Sound had haplotypes private to this region. The rest of the harvested whales had haplotypes shared with other populations from the Hudson Bay-Strait Complex. Nuclear DNA results also suggested the presence of two populations in Cumberland Sound during summer with approximately 74% of the whales belonging to the CSB population. The degree of differentiation between the CSB and WHB populations was low (FST=0.014), but this is expected given the recent colonization of the Hudson Bay-Strait Complex. Our results support that there are two populations of beluga whales that summer in Cumberland Sound. Based on all current information, managing beluga whales inhabiting Cumberland Sound in the summer as a single stock comprised of two genetic populations is the most precautionary approach. [ABSTRACT FROM AUTHOR]

Belugas from the Hudson Bay-Strait Complex are harvested by hunters from Nunavik and Nunavut communities. In past studies, a genetic mixture analysis (GMA) was used to determine the contribution of animals from the Western Hudson Bay (WHB) and Eastern Hudson Bay (EHB) populations to harvests in the different management units. The population definition of WHB and EHB relied on short haplotypes from the mitochondrial (mt) DNA control region. However, studies with long haplotypes have shown that four populations could be identified in the Hudson Bay-Strait Complex. Here, we aim to 1) revisit the definition of populations within this area by resequencing and comparing short and long haplotypes for the mtDNA control region of 2 861 belugas, and 2) update the relative contribution of the newly-defined reference groups in management units using a GMA. Long haplotypes confirmed the existence of four previously defined populations, WHB, EHB, James Bay (JAM), and Cumberland Sound (CSB). They also allowed for the identification of a fifth population in the Belcher Islands (BEL). For the second objective, we tested the validity of the five populations using a leave-one-out approach and observed a high rate of erroneous assignments between EHB and BEL populations. Misassignments were due to shared genetic matrilineages and possible admixture during summer. Assignments were improved when EHB and BEL populations were combined in as a single reference named BEL-EHB stock. The GMA with the four new reference groups (WHB, JAM, CSB populations and BEL-EHB stock) led to higher proportions of the BEL-EHB stock in the fall harvest in Hudson Strait (5.6%) and Northeastern Hudson Bay (14.9%), and year-round in Sanikiluaq (3.3 to 61.2%), compared to previous studies using only two reference groups. Changes in reference groups, and not resequencing, likely caused most of the variation in the proportional estimate of the BEL-EHB stock in management units. Results from this study have increased the accuracy of the proportional contribution from reference groups to the harvest by the Nunavik and Nunavut (Sanikiluaq) communities. [ABSTRACT FROM AUTHOR]