Estimation of walrus populations on sea ice with infrared imagery and aerial photography

Population sizes of ice‐associated pinnipeds have often been estimated with visual or photographic aerial surveys, but these methods require relatively slow speeds and low altitudes, limiting the area they can cover. Recent developments in infrared imagery and its integration with digital photography could allow substantially larger areas to be surveyed and more accurate enumeration of individuals, thereby solving major problems with previous survey methods. We conducted a trial survey in April 2003 to estimate the number of Pacific walruses (Odobenus rosmarus divergens) hauled out on sea ice around St. Lawrence Island, Alaska. The survey used high altitude infrared imagery to detect groups of walruses on strip transects. Low altitude digital photography was used to determine the number of walruses in a sample of detected groups and calibrate the infrared imagery for estimating the total number of walruses. We propose a survey design incorporating this approach with satellite radio telemetry to estimate the proportion of the population in the water and additional low‐level flights to estimate the proportion of the hauled‐out population in groups too small to be detected in the infrared imagery. We believe that this approach offers the potential for obtaining reliable population estimates for walruses and other ice‐associated pinnipeds.     

Regional walrus abundance estimate in the United States Chukchi Sea in autumn

Human activities (e.g., shipping, tourism, oil, gas development) have increased in the Chukchi Sea because of declining sea ice. The declining sea ice itself and these activities may affect Pacific walrus (Odobenus rosmarus divergens) abundance; however, previous walrus abundance estimates have been notably imprecise. When sea ice is absent from the eastern Chukchi Sea, walruses in waters of the United States usually rest together onshore at a single Alaska coastal haulout, where they can be surveyed more easily than when they rest on dispersed offshore ice floes. We estimated the number of walruses on land (herd size) at this haulout from 13 unoccupied aircraft system (UAS) surveys flown within a 10-day period in each of 2018 and 2019. We estimated population size of walruses using the haulout over the course of the surveys by combining herd size data with data from satellite-linked transmitters that indicated whether tagged walruses were in or out of water during each survey. Our estimates of the population size of walruses using the haulout during each year's survey period were similar to each other and more precise than historical walrus abundance estimates: posterior means (95% credibility intervals) were 166,000 (133,000–201,000) for 2018 and 189,000 (135,000–251,000) for 2019. Auxiliary observations support using these estimates to represent the size of the population using the eastern Chukchi Sea in autumn during the surveyed years. Our study site was the only substantial Chukchi Sea coastal haulout in the United States during the survey periods and study-specific tracking data (consistent with known distribution and movement patterns) indicated tagged walruses remained in eastern Chukchi waters during the survey periods. In addition, the imagery, telemetry, and analytical methods developed for this study advance the prospect for precise range-wide walrus population size estimates.     

AERIAL CENSUS OF PACIFIC WALRUSES IN THE CHUKCHI SEA, 1985

The U.S. Fish and Wildlife Service conducted a survey of the walruses in the pack ice of the Chukchi Sea between 16 September and 2 October 1985, as part of a joint effort with the Soviet Union to estimate the size of the Pacific walrus population. American observers conducted censuses from two aircraft along randomly selected north–south lines over the pack ice. The observers counted walruses within a constant viewing angle that corresponded to a total strip width of 1.38 km at an altitude of 152 m.
In nine days of flying, 15,312 walruses were observed, of which 10,140 were on 5,764 km² of census strips. Few walruses were observed east of 161°W longitude or west of 170°W longitude, hence the census effort was stratified. Walrus concentrations between 161° and 170° shifted slightly westward during the 2-wk duration of the censuses. The differences among days in observed walrus density were due to changes in the numbers of walruses on the ice within 37 km of the ice edge. The number of observable walruses in pack ice of the eastern Chukchi Sea was estimated to be 62,177 (SD = 19,480), based on censuses conducted on 29 and 30 September and 1 October. At that time there were also at least 15,238 in Bristol Bay, Bering Sea. The Soviets counted 39,572 on the shores of the western Chukchi and Bering seas and estimated 115,531 in pack ice of the western Chukchi Sea. Summing U.S. and Soviet estimates, the total population of Pacific walruses in 1985 was 232,518. This number was comparable with earlier estimates from censuses conducted jointly by the U.S. and the Soviets. However, information on fraction hauled out, segregation, and movements is needed for more precise estimates.     

Aerial survey of Atlantic walruses (Odobenus rosmarus rosmarus) in the Pechora Sea, August 2011

Basic knowledge of the population biology of Atlantic walruses throughout the eastern parts of their range, including the Pechora Sea, is scarce or nonexistent. Herein, we present the first estimate of walrus numbers from the Pechora Sea based on an aerial survey of 2,563?km of coastline, using a combination of infrared techniques and digital imagery. Hauled out walruses were found at three sites (Vaygach Island and two sites on Matveyev Island). A total of 968 animals were counted on aerial photographs; all of the animals appeared to be males. Crude measurements of dorsal curvilinear lengths of a subset of the photographed animals (N?=?504) showed that many were adults, but 14.5?% belonged to younger age classes (shorter than 225?cm). Using an adjustment factor developed for male walruses in Svalbard, to account for animals at sea during the survey, the number of walruses occupying this area was estimated to be 3,943 (95?% CI, 3,605–4,325). No females with calves were seen in this survey, implying that the population that uses the Pechora Sea during summer has a distributional area that is larger than the survey area. Extensive oil exploration, development and production are currently taking place in the Pechora Sea. Risks posed to walruses and their prey by these industrial activities should be assessed immediately, and the genetic delineations of this population should be clarified.     

Pacific Walrus (Odobenus rosmarus divergens) Resource Selection in the Northern Bering Sea

The Pacific walrus is a large benthivore with an annual range extending across the continental shelves of the Bering and Chukchi Seas. We used a discrete choice model to estimate site selection by adult radio-tagged walruses relative to the availability of the caloric biomass of benthic infauna and sea ice concentration in a prominent walrus wintering area in the northern Bering Sea (St. Lawrence Island polynya) in 2006, 2008, and 2009. At least 60% of the total caloric biomass of dominant macroinfauna in the study area was composed of members of the bivalve families Nuculidae, Tellinidae, and Nuculanidae. Model estimates indicated walrus site selection was related most strongly to tellinid bivalve caloric biomass distribution and that walruses selected lower ice concentrations from the mostly high ice concentrations that were available to them (quartiles: 76%, 93%, and 99%). Areas with high average predicted walrus site selection generally coincided with areas of high organic carbon input identified in other studies. Projected decreases in sea ice in the St. Lawrence Island polynya and the potential for a concomitant decline of bivalves in the region could result in a northward shift in the wintering grounds of walruses in the northern Bering Sea.     

Large numbers of marine mammals winter in the North Water polynya

The importance of the North Water polynya in Smith Sound as an overwintering area for marine mammals has been questioned. One way to address the issue is to assess the abundance of selected marine mammals that are present during winter in the North Water. Visual aerial surveys involving double observer platforms were conducted over the eastern part of the North Water polynya in April 2014. Four species of marine mammals were included in strip-census estimation of abundance. Perception bias was addressed using a double-platform survey protocol, a Chapman mark–recapture estimator for whales, seals and walruses (Odobenus rosmarus) on ice and a mark–recapture distance sampling estimation technique for walruses in water. Availability bias was addressed by correcting abundance estimates by the percentage of time animals detected in water that were available for detection at the surface. The resulting estimates suggested that 2544 walruses (95 % CI 1513–4279), 6005 bearded seals (Erignathus barbatus, 95 % CI 4070–8858), 2324 belugas (Delphinapterus leucas, 95 % CI 968–5575) and 3059 narwhals (Monodon monoceros, 95 % CI 1760–5316) wintered in the eastern part of the North Water polynya in April 2014. The walrus estimate is larger than previous summer estimates, and it emphasizes the importance of the habitat along the Greenland coast as a walrus wintering ground. The estimate of belugas is likely negatively biased due to the partial coverage of the potential habitat. The estimate of narwhals is large compared to the few previous observations of narwhals in winter in the North Water, and it demonstrates that large numbers of narwhals winter there. The overall conclusion is that the North Water is indeed an important wintering area for at least walruses, belugas, narwhals and bearded seals.     

Indication of two Pacific walrus stocks from whole tooth elemental analysis

The Pacific walrus (Odobenus rosmarus divergens) is considered to be a single panmictic population for management purposes. However, studies on population structuring in this species are limited; in part, because portions of the population’s range are often inaccessible. Therefore, alternative and complementary methods for investigating stock structure in the Pacific walrus are of particular interest. We used measures of elemental concentrations in whole tooth sections from ICP-MS in a discriminant analysis to investigate evidence of stock separation between walruses from two of three known breeding areas (S.E. Bering, St Lawrence, and Anadyr Gulf). Elemental compositions of teeth from female and male walruses from the S.E. Bering and St Lawrence breeding areas were significantly different, providing evidence of separate stocks. We also obtained insights into the potential relation of walruses from non-breeding areas to walruses from these breeding groups based on similarities in their dental elemental profiles.     

Migration of Walruses (Odobenus rosmarus) in the Svalbard and Franz Josef Land area

Thirty-four walruses ( Odobenus rosmarus ) were fitted with satellite transmitters (PTTs) from 1990 to 1993 in order to study the distribution of the population in the Svalbard area. Twenty-eight were caught at Svalbard and six at Franz Josef Land. All were males except one female caught at Franz Josef Land. At Svalbard, one walrus was caught on the west coast of Spitsbergen, while the others were caught at southern Edgeøya. All walruses were caught in the period from mid-July to early September. The PTTs provided information on location for periods ranging from 0 to 212 days. The results of the satellite trackings show that there is a migration of male walruses between most of the walrus areas at Svalbard and Franz Josef Land. In particular, it seems that migration of males from southern Edgeøya to Kvitøya, Viktoria Island, and Franz Josef Land is common. The walruses winter in the southern parts of Svalbard, as well as within the winter pack-ice of north-eastern Svalbard, which contains numerous open leads. The only walrus at Franz Josef Land that was followed to mid-winter stayed in the area and therefore supports the view that walruses also winter in that area. It is assumed that the majority of walruses at Svalbard are males from one common Svalbard-Franz Josef Land stock. The walrus in the Svalbard-Franz Josef Land area today belong to a recovering population. Their current distribution and behaviour may therefore differ from that found in Svalbard in former times.     

Population structure and gene flow of the Atlantic walrus (Odobenus rosmarus rosmarus) in the eastern Atlantic Arctic based on mitochondrial DNA and microsatellite variation

The population structure of the Atlantic walrus, Odobenus rosmarus rosmarus , was studied using 11 polymorphic microsatellites and restriction fragment length polymorphism detected in the NADH-dehydrogenase ND1, ND2 and ND3/4 segments in mtDNA. A total of 105 walrus samples were analysed from northwest (NW) Greenland, east (E) Greenland, Svalbard and Franz Joseph Land. Two of the 10 haplotypes detected in the four samples were diagnostic for the NW Greenland sample, which implied that the group of walruses in this area is evolutionary distinct from walruses in the other three areas. One individual sampled in E Greenland exhibited a Pacific haplotype, which proved a connection between the Pacific walrus and walruses in eastern Greenland. The Franz Joseph Land, Svalbard and E Greenland samples shared the most common haplotype, indicating very little differentiation at the mtDNA level. Gene flow ( Nm ) estimates among the four areas indicated a very restricted exchange of female genes between NW Greenland and the more eastern Atlantic Arctic samples, and a closer relationship between the three samples composing the eastern Atlantic Arctic. The genetic variation at 11 polymorphic microsatellite loci grouped individuals into three populations, NW Greenland, E Greenland and a common Franz Joseph Land–Svalbard population, which were connected by moderate gene flow.     

Proportion of higher trophic-level prey in the diet of Pacific walruses (Odobenus rosmarus divergens)

During nutritionally stressful situations, Pacific walruses (Odobenus rosmarus divergens) may switch from preying on benthic invertebrates to higher trophic-level prey (HTLP) (e.g., pinnipeds and/or seabirds). We applied a Bayesian mixing model to stable isotope (C and N) data from analyses of various tissues (tongue and lumbar muscle, skin, and liver) to quantify the proportional contribution of HTLP to walruses (n = 293 individuals). The mode contribution of HTLP to walrus diet was ~22 % (±10 %) based on muscle mixing models, which is consistent with results from contaminant studies of Atlantic walruses (Odobenus rosmarus rosmarus), but higher than estimates based on historical stomach content analyses of Pacific walruses. A broader range in the proportion of HTLP (0–60 %) shown by mixing models using stable isotope data from liver and skin of walruses indicated they pursue an opportunistic foraging strategy. Data from the HTLP-consuming walruses were comparable with our stable isotope data of a known “seal-eating” walrus. No significant difference was evident between the estimated contributions of HTLP to the diet of male versus female walruses (P > 0.01). This finding suggests that changes in diet base for walruses are not influenced by the sex of the predator.     

Killer whales (Orcinus orca) hunting for walruses (Odobenus rosmarus divergens) near Retkyn Spit, Chukotka

Group hunting by killer whales for walruses was observed in August 18, 2008, in the littoral area (3 km from the haulout of walruses, Retkyn Spit, Chukotka). The group of killer whales consisted of seven adults (one adult male did not participate in attacks) and two calves. Based on prey type, these killer whales were mammal-eating. The total duration of their hunt activity was not less than 95 min. The hunt consisted of three phases. The first phase was an attack on the group of walruses and choice of individual prey; the second phase was attacks on the chosen walrus; and the third (final) phase was a decrease in activity of killer whales and leaving group with walrus from sea shore. The main behavioral patterns of killer whales during the hunt were discerned. Two killer whales tried to kill walruses by chasing them and jumping out of the water on the shore. The video analysis of the ??attack phase?? showed that killer whales made 55 attacks on the walrus during 17.3 min. On average, each killer whale attacked the walrus seven times. The attack tactics of killer whales, the number of movements, and the location of killer whales (adults and calves) relative to each other and to the walrus were described. Well coordination of their movements and group actions was observed.     

Density and habitat use of lions and spotted hyenas in northern Botswana and the influence of survey and ecological variables on call-in survey estimation

Top predators significantly impact ecosystem dynamics and act as important indicator species for ecosystem health. However, reliable density estimates for top predators, considered necessary for the development of management plans and ecosystem monitoring, are challenging to obtain. This study aims to establish baseline density estimates for two top predators, spotted hyena and lion, in the Okavango Delta in northern Botswana. Using calling stations, we surveyed free-ranging populations of the two species and investigated methodological variables that might influence results about distributions and densities, including habitat type, seasonality, and different types of playback sounds. Calling stations were distributed over a survey area of approximately 1,800 km² characterized by three major habitat types: mopane woodland, floodplain and mixed acacia sandveld. Results indicate spotted hyenas were evenly distributed independent of habitat type and season throughout the survey area with an overall density estimate of 14.4 adults/100 km². In contrast, lion distribution and density varied significantly with habitat and season. Lion density in the prey-poor mopane woodland was near zero, while in the comparatively prey-rich floodplains it was estimated at 23.1 individuals/100 km² resulting in a weighted average density of 5.8 individuals/100 km² across the entire study area. In testing the effect of varying playback sounds we found that both species were significantly more likely to respond to calls of conspecifics. Our results show how several methodological variables may influence density estimates and emphasize the importance of standardized calling-station survey methods to allow consistent replication of surveys and comparison of results that can be used for landscape-scale monitoring of large predator species.     

Body growth in Atlantic walruses (Odobenus rosmarus rosmarus) from Greenland

Morphometric data were collected from 105 Atlantic walruses ( Odobenus rosmarus rosmarus ) in northwestern Greenland in the periods 1977–78 and 1989–91. Of these 21 walruses were subjected to a detailed study on body composition.
The asymptotic maximum standard body length of Atlantic walruses in NW Greenland was 269 cm for females and 314 cm for males. This is similar to Pacific walruses, but significantly longer than Atlantic walruses from Hudson Bay. Despite this, walruses from NW Greenland apparently do not attain the same total body mass as Pacific walruses ( O.r.divergens ).
The asymptotic maximum body weights for walruses in NW Greenland were estimated to be 720 kg for females and 1114 kg for males.
Total body surface area was proportional to the two-thirds power of total body weight.
The percentage proportion of blubber and viscera were both negatively correlated to body mass, while skin and muscles constituted a nearly fixed proportion. On average, blubber constituted 19% of total body mass of adult females, 15% of adult males and 24% of subadults of both genders. The average walrus consisted of 18% blubber, 12% skin, 12% viscera and 58% blood, muscle and skeleton. Muscles were estimated to constitute 44% of total body weight.
Allometric functions for weight of internal organs relative to body mass are presented.     

RESPONSE OF PACIFIC WALRUSES TO DISTURBANCES FROM CAPTURE AND HANDLING ACTIVITIES AT A HAUL-OUT IN BRISTOL BAY, ALASKA

Observations were made on herds of the Pacific walrus ( Odobenus rosmarus divergens ) to study their response during the capturing and handling of adult males in summer 1995 at a haul-out at Cape Peirce in southwestern Alaska. Three behaviors (alertness, displacement, and dispersal) were quantified from 16 capture sessions. Herd sizes ranged from 622 to 5,289 walruses. Handling of an immobilized walrus consisted of attempts to attach telemetry devices to the tusks and collect various biological samples. Handling activities resulted in an average of about 10-fold or greater levels of behavior in alertness, displacement, and dispersal than during precapture and darting periods. High levels of behavior usually occurred within the first 45 min of handling. In 8 of 10 capture sessions, walruses returned to predisturbance levels of behavior within 40 min of cessation of the handling disturbance. Alertness and displacement were moderately and negatively correlated with herd size during the handling period, which may reflect an effect of a threshold distance from the point of disturbance to responding individuals. Observations of walruses tagged with VHF radio transmitters indicated that the activities from a given capture session did not preclude tagged walruses from using the haul-out over a subsequent 11 -wk monitoring period. Moreover, non-tagged walruses continued to extensively use the haul-out during and after the period in which capture sessions were conducted.     

Population densities and habitat use of the golden jackal (Canis aureus) in farmlands across the Balkan Peninsula

The main objective of this study was to analyze the habitat use and population densities of the golden jackal in four countries across lowland regions of the Balkan Peninsula, known as the core area of the species' distribution in Europe. Using indirect (acoustic) method for detecting territorial golden jackals, we surveyed jackal presence and densities on 331 monitoring sites in four countries, covering area an of 4,296 km² in total during April and May 2007–2012. We used GIS to assess landscape and environmental characteristics in a 2-km circular buffer (12.6 km²) around calling stations. Average population density of golden jackals in the study areas ranged between 0.6 and 1.1 territorial groups/10 km² (mean ± SE, 0.6?±?0.06 groups/10 km²), with several high-density areas with up to 4.8 territorial groups/10 km². Analysis of habitat use showed that for both jackal occurrence and number of jackal groups, the only significant parameter was the interaction between country and intensity of agriculture, indicating that jackals adapt their habitat selection patterns in relation to the habitat availability. We observed that selection of the more suitable habitats (shrub–herbaceous vegetation/heterogeneous agricultural vegetation) increased with lower proportion of these habitat types in the study area. Our study confirms high habitat plasticity of the golden jackal and offers explanation for its recent range expansion, which might be connected with the land use changes during the last decades in the Balkan Peninsula.     

Impact of road construction on giant panda’s habitat and its carrying capacity in Qinling Mountains

The Qinling giant panda (Ailuropoda melanoleuca) is an endangered endemic species to China. Despite ongoing efforts to ensure its conservation, concerns about maintaining its populations persist. We used GIS fed with data on land use including road network of 2001, third national giant panda survey, and a digital elevation model (DEM) to assess the impact of road construction on giant panda habitat, and estimate the carrying capacity of the Qinling Mountain area. We assessed habitat suitability with a mechanistic model, and conducted correlation analysis to evaluate relationship between the extent of giant panda habitat and amount of sites occupied by pandas within of 5 km × 5 km grid. We also estimated the carrying capacity of the Qinling Mountainous Area.
Our results revealed a significant correlation (R² = 0.447, P < 0.01) between the number of sites with signs left by giant panda and the extent of habitat within of 5 km × 5 km grid. The minimum habitat area that can support one panda was 10 km². Before the road network construction, the area of habitat suitable for the panda amounted about 1561 km² and that of marginally suitable habitat about 1499 km². The corresponding carrying capacity represented about 240 individuals. After the road network construction, the suitable habitat area was reduced by nearly 30% to 1093 km². Marginally suitable habitat and unsuitable habitat have both increased by 17% and 1%, respectively. As a result, the potential population size which the habitat could support was reduced to 217 individuals. The study results also suggested that most impacts on habitat from road construction took place in the high elevation areas above 1500 m. However, regarding the impact on the giant panda habitat, road networks developed much more inside the current nature reserves than outside of them.     

Impact of road construction on giant panda’s habitat and its carrying capacity in Qinling Mountains

The Qinling giant panda (Ailuropoda melanoleuca) is an endangered endemic species to China. Despite ongoing efforts to ensure its conservation, concerns about maintaining its populations persist. We used GIS fed with data on land use including road network of 2001, third national giant panda survey, and a digital elevation model (DEM) to assess the impact of road construction on giant panda habitat, and estimate the carrying capacity of the Qinling Mountain area. We assessed habitat suitability with a mechanistic model, and conducted correlation analysis to evaluate relationship between the extent of giant panda habitat and amount of sites occupied by pandas within of 5 km × 5 km grid. We also estimated the carrying capacity of the Qinling Mountainous Area.
Our results revealed a significant correlation (R² = 0.447, P < 0.01) between the number of sites with signs left by giant panda and the extent of habitat within of 5 km × 5 km grid. The minimum habitat area that can support one panda was 10 km². Before the road network construction, the area of habitat suitable for the panda amounted about 1561 km² and that of marginally suitable habitat about 1499 km². The corresponding carrying capacity represented about 240 individuals. After the road network construction, the suitable habitat area was reduced by nearly 30% to 1093 km². Marginally suitable habitat and unsuitable habitat have both increased by 17% and 1%, respectively. As a result, the potential population size which the habitat could support was reduced to 217 individuals. The study results also suggested that most impacts on habitat from road construction took place in the high elevation areas above 1500 m. However, regarding the impact on the giant panda habitat, road networks developed much more inside the current nature reserves than outside of them.     

Pacific walrus: Benthic bioturbator of Beringia

The dependency of walruses on sea ice as habitat, the extent of their feeding, their benthic bioturbation and consequent nutrient flux suggest that walruses play a major ecological role in Beringia. This suggestion is supported by several lines of evidence, accumulated during more than three decades of enquiry and leading to the hypothesis that positive feedbacks of walrus feeding strongly influence productivity and ecological function via benthic bioturbation and nutrient flux. Walruses annually consume an estimated 3 million metric tons of benthic biomass. Walrus prey species inhabit patches across the shelf according to sediment type and structure. Side-scan sonar and our calculations indicate that the area affected by walrus feeding is in the order of thousands of square kilometers per year. Annual to long-term walrus bioturbation results in significant, large-scale changes in sediment and biological-community structure, and magnifies nutrient flux from sediment pore water to the water column by about two orders of magnitude over wide areas. The combined effects of walrus feeding must be placed in the context of long-term, regional climate changes and responses. Should sea ice continue to move northward as a result of climate change, the walrus' ecological role could be diminished or lost, the benthic ecosystem could be fundamentally altered and native subsistence hunters would be deprived of important resources.     

The influence of ice conditions on terrestrial haulouts of the pacific walrus Odobenus rosmarus divergens Illiger, 1815 in the Gulf of Anadyr,Bering Sea

The boundaries of the area of appropriate depths for feeding by the Pacific walrus were determined. Based on the ice maps for the period of 1997–2011, the long-term dynamics of ice breakup and formation in the Gulf of Anadyr (Bering Sea) were analyzed. There was a tendency towards a reduction in the ice-cover period in the Gulf of Anadyr, especially during 2007–2011. In 2008, the number and duration of stay of walruses at terrestrial haulout sites on Meeskyn Spit Island and on Retkyn Spit continued to decrease. Walruses also used other haulouts located in the Gulf of Anadyr at Cape Gek and Cape Retkyn. The age and sex structure of the walrus herd changed: the proportions of adult males and young animals decreased, while the proportion of females with calves increased.     

The Laptev Sea walrus Odobenus rosmarus laptevi: an enigma revisited

The walrus ( Odobenus rosmarus ) is in some current systematic schemes divided into three subspecies: O. r. rosmarus in the North Atlantic, O.   r. divergens in the North Pacific and O.   r. laptevi in the Laptev Sea. These three subspecies have been described as differing in body size, but the taxonomic status of O.   r. laptevi is disputed. The current study applies molecular and morphometric methods to assess the taxonomic status of O.   r. laptevi and to analyse the systematic and phylogeographic relationships between the three purported walrus subspecies. Tusk length and tusk circumference were measured from the few skulls available of O.   r. laptevi , and the obtained values were within the ranges reported for Pacific walruses. Thus, morphologically, subspecies status for O.   r. laptevi is not supported according to the Amadon–Mayr '75% rule'. Phylogenetic analyses and haplotype networks based on mitochondrial nucleotide sequence data of NADH dehydrogenase 1, 16S rRNA, cytochrome oxidase I and the d -loop of the control region of the historic O.   r. laptevi bone material and contemporary O.   r. rosmarus and O.   r. divergens showed that the Laptev Sea walrus groups with individuals from the North Pacific. Thus, the mitochondrial sequence data do not support the recognition of three walrus subspecies as reciprocally monophyletic evolutionary units with independent evolutionary histories. Only O.   r. rosmarus and O.   r. divergens meet this criterion with the present sampling. Accordingly, we recommend that Odobenus r. laptevi be abandoned and the Laptev walrus instead be recognized as the westernmost population of the Pacific walrus, Odobenus r. divergens. However, further research is recommended to assess whether the Laptev walrus could be considered as a significant unit in terms of conservation and management, since it is unique in several ecological parameters.