Skip booklet index and go to page content

COSEWIC Assessment and Update Status Report on the Atlantic Walrus in Canada

Population Sizes and Trends

As the number of whales in the Canadian Arctic and Hudson Bay dwindled towards the end of the 19th century, whalers turned their efforts to hunting other species, including walruses (Low 1906; Degerbøl and Freuchen 1935). Commercial and sport hunting of walruses was banned in Canada in 1928 by an order in the Privy Council (P.C. 1036) (see Mansfield 1973). There are no estimates of the level of exploitation, but walrus populations in the eastern Canadian Arctic are apparently still much reduced (Manning 1946; Schwartz 1976; Reeves 1978; Davis et al. 1980; Richard and Campbell 1988; Born et al. 1995). There is no comprehensive compilation of historical harvest records that might be used to estimate initial levels of these populations. 

There are no complete or recent estimates of numbers for walrus populations in Nunavut (Stewart 2002) or Nunavik. There are index estimates for some populations but they have broad confidence intervals (Cosens et al. 1993), so that changes in population size would have to be very large to be detectable. These estimates cannot be corrected for animals that are submerged beyond view or for haul-out dynamics, which determine the proportion of a herd that is on land or ice or in the water at the time of the survey. Data on the number of walrus present in an area often consist simply of counts of animals hauled out at known areas of concentration, or opportunistic sightings, and cannot be used to estimate the total Canadian walrus population (Richard and Campbell 1988).

Greenland walrus populations--West Greenland, North Water, and East GreenlandFootnote 3--were recently assessed using a density-dependent population dynamics model with assumed life-history parameters, fitted by Bayesian methods to catch series extending back to the early 1900s or late 1800s, and a single survey estimate for each population (Witting and Born 2005). Conclusions from the analyses of the western Greenland populations (i.e. West Greenland and North Water) were that their pristine sizes had been very much larger, they were now overhunted, and further depletion or extinction was likely. Conclusions were dependent on the survey estimates and the assumptions about life-history parameters--and the data contained no information that could update the prior distributions of life history variables. A review of these assessments by NAMMCO in 2005 recognized that prior distributions of life history variables were not updated by the data. Concerns were also raised about the quality of the surveys. The likely connection between walruses in Hudson Strait and northern Hudson Bay stock with those in West Greenland was also not considered in the assessment by Witting and Born (2005). However, Witting and Born’s (2005) findings agreed with the general conclusion that these populations (West Greenland and North Water) are substantially depleted despite the reservations of the NAMMCO Scientific Committee concerning details of the assessment.


South and East Hudson Bay Population

A comprehensive survey of the walrus population in eastern and southern Hudson Bay and James Bay has not been conducted. In the spring of 1955, a herd estimated at over 1000 animals was seen hauled out on the sandspit at Cape Henrietta Maria (Clarke in Loughrey 1959) and in September 1955, the captain of the Fort Severn saw a herd off the coast at Winisk (Loughrey 1959). Born et al. (1995) have questioned the accuracy of the former estimate, which was told to Clarke, and cautioned against its use in determining population trend. The Ontario Department of Natural Resources (DNR) has recorded opportunistic observations of walruses at shoals off the mouth of the “Brant River” since 1957 (C. Chenier, DNR, pers. comm. 2003). Walruses have been seen on the shoals between 20 July and 18 October. The number of animals varies widely and no trend is apparent in the population. High counts recorded by Ontario DNR have ranged from 310 on 5 October 1978 (J.P. Prevett), to 204 on 9 September 1983 (K.F. Abraham) and from to 330 on 8 September 1986 (K.F. Abraham), to about 221 in August 1999 (C. Chenier).

In the late 1930s, Twomey and Herrick (1942) saw and hunted a herd they estimated at over 400 animals in the northern Sleeper Islands. Since then only smaller herds have been reported from the region. On 4 August 1971, Manning (1976) saw 75 walruses near the south end of the Sleeper Islands, and the next day saw 25 off the west coast of Kidney Island, largest of the Sleeper Islands. A herd of about 30 animals was seen at the Sleepers in October 1996; nine of these animals were harvested (Brooke 1997). In the summer of 1993, a herd of about 30 animals was seen in the Belchers directly north of Sanikiluaq during an aerial survey (J. Desrosier, pers. comm. 2003). Hunters report that there are fewer walruses near the community and on neighbouring islands now, than in the past (DFO 2000).

In the early 1990s, walrus were reportedly numerous along the Ontario coast of Hudson Bay west to the Winisk (Peawanuck) area, and have been seen in July near Fort Severn (Fleming and Newton 2003). This coastline may provide a refuge for the population, since Cree hunters do not have a strong tradition of hunting walruses and take few animals (Johnston 1961).

Richard and Campbell (1988) and Born et al. (1995) estimated the size of this population at 410+ and “500?” animals, respectively. Both of these estimates were tentative and based on a few sightings in a wide geographical area over a long period. Taking the largest direct counts of the past decade or so, as did Richard and Campbell (1988), yields an updated estimate of 270+ animals. The data are too few to assess whether a decline has occurred.


Northern Hudson Bay–Davis Strait Population

Aerial survey counts of walruses in the northern Coats Island, Walrus Island and southeast Southampton Island area of northern Hudson Bay were conducted in July or August of 1954 (Loughrey 1959), 1961 (Mansfield 1962), 1976-77 (Mansfield and St. Aubin 1991) and 1988-90 (Richard 1993). They produced maximum counts, respectively, of 2900 (Loughrey 1959: 80), 2650 (Mansfield 1959; 46), 2370 (Mansfield and St. Aubin 1991: 97), and 1376 (Richard 1993: 7) walruses. The real population size is likely greater, as these counts are not based on systematic surveys of the entire area and were not corrected for animals missed by the observers. While they suggest a declining trend, care must be taken in interpreting these data given differences in survey methods and ice coverage, and wide fluctuations in the numbers of animals hauled out at any particular time. Richard’s (1993) counts in 1988–90, for example, were above the average of the daily counts in 1976 and 1977. On Coats Island, Gaston and Ouelett (1997) counted about 600 animals at Cape Pembroke on 7 August 1992 and about 500 at Cape Prefontaine on 31 July 1995. Hunters from Coral Harbour have reported an increase in the number of walruses near their community over the past 10 years (DFO 2000).

Walruses were more common and numerous along the west coast of Hudson Bay between Arviat and Chesterfield Inlet in the past (Loughrey 1959; Born et al. 1995). They are now found mostly in the area north of Chesterfield Inlet. No counts are available for this region.

Cape Dorset hunters have reported walrus herds of between 500 and 1000 animals on the ice or at uglit along western Foxe Peninsula in summer, particularly between Cape Dorset and Cape Dorchester, with a similar number in the area between Salisbury and Nottingham islands (Orr and Rebizant 1987). They also report an increase in the number of walruses near the community over the past 30 years (DFO 2000). An aerial survey in August 1990 counted 461 walruses on Nottingham Island (Richard 1993).

Surveys of the southeast Baffin coast conducted in the summer and fall of 1977 through 1979 located a number of uglit (MacLaren Atlantic Limited 1978; MacLaren Marex Inc. 1979, 1980a+b; Smith et al. 1979). The number of walrus at each ugli was typically between a few and several hundred, with the exception of a small island near Lady Franklin Island, where 600 to 700 animals were seen on 15 August 1979 (McLaren Marex Inc. 1980b).

Currie (1963: 22) estimated the population of walruses that visited Akpatok Island in spring around 1960 as at least 1000 and perhaps over 2000. Indeed, before the 1950s hunters landed 800 animals there in one season, and during the 1950s there were annual harvests of 150 to 200 animals. Residents of the communities around Ungava Bay still hunt walruses each year at Akpatok Island (Olpinski 1990, 1993; Portnoff 1994; Brooke 1997) but the number of animals now present is unknown.

Richard and Campbell (1988) and Born et al. (1995) estimated the size of this population at 4850–5350 and 6000 animals, respectively. Both of these estimates were tentative and based on a few sightings in a wide geographical area over a long period. They have very wide uncertainty, but cannot be updated as there are no new data.


Foxe Basin Population

Because Foxe Basin was not a major whaling ground or shipping route its walrus population was not hunted intensively by whalers when larger prey became scarce. While its initial size is unknown the population is likely somewhat reduced by the substantial subsistence harvests. Orr et al. (1986) counted 2722 walruses during a reconnaissance helicopter survey of northern Foxe Basin in August 1983. In August 1988 and July 1989, systematic visual strip transect surveys were flown over central Foxe Basin (Richard 1993). In 1988, the number of animals at the surface was estimated at 5200 (95%CI 900-30500) based on a direct count of 440; in 1989 it was estimated at 5500 (95%CI 2700-11200), based on a direct count of 475. These surveys did not cover the whole of northern Foxe Basin or correct for animals that were submerged beyond view. They provide an index against which change can be measured rather than an overall estimate of the population.

Richard and Campbell (1988) estimated the size of this population at 2725+ animals based on direct counts by Orr et al. (1986); Born et al. (1995) based their estimate of 5500 animals on Richard’s (1993) survey data. The latter estimate has wide uncertainty limits but stands as the best available. It suggests a minimum population of well over 2700 animals in 1989. No trend in the population can be established from the data available.


Baffin Bay (High Arctic) Population

No comprehensive survey of the North Water and adjacent areas occupied by this population has been conducted, so only piecemeal information from different seasons and years is available for estimating the population size. Finley and Renaud (1980) counted about 700 walruses in the North Water in March 1979, but saw few during a survey the previous year. During a systematic survey of the area in late March 1993, Richard et al. (1998) saw 13 animals, widely dispersed. Davis et al. (1978) surveyed terrestrial haul-out sites and adjacent waters in the central Canadian High Arctic in 1977. Between 21 and 23 August they counted 303 animals in the Goose Fiord–Walrus Fiord area of southwest Ellesmere Island and Norfolk Inlet–Arthur Fiord areas of northern Devon Islands; 118 on the east side of Penny Strait, 281 in the Crozier Strait-Pullen Strait-Milne Inlet area, and 71 along the southwest coast of Devon Island. In the last area, a polar bear was present at one of the haul-outs and swimming animals were not counted. The distance that separates these areas and the brief counting period makes it unlikely that animals were double counted. Based on these counts at least 773 walruses were alive in the region in 1977.

These and other data were reviewed in detail by Born et al. (1995), who suggested that summering walruses might number: 100 in the Kane Basin region; 300 in the Buchanan Bay (78°58'N, 75°11' W) and Princess Marie Bay (79°20'N, 76°00' W); 300-600 in Jones Sound and along eastern Ellesmere Island south of Pim Island (78°44'N, 74°25' W); and 1000 in the Lancaster Sound area and along southern Devon Island. These numbers were educated guesses because the data were insufficient for an accurate, current assessment. Since then, late summer surveys of the coastal waters of Jones Sound and northern Lancaster Sound over 4 consecutive years (1998-2001) produced an average count of about 350 walruses (DFO 2000; R. Stewart, DFO pers. comm. 2004). This does not alter the estimates but confirms that at least 350 walruses were alive in the region in recent years.

An aerial survey was conducted in August 1999 along the coasts of eastern Ellesmere Island and in Jones Sound, south Devon Island and Cornwallis Island – Grinnell Peninsula. A total of 452 walruses were counted. Witting and Born (working paper cited in NAMMCO 2006) applied a correction factor to this count and estimated there were about 1000 animals in the surveyed area. They further assumed the unsurveyed areas contained about 500 animals. NAMMCO (2006) noted that the total estimate of 1500 for the North Water should not be used in assessments until further information is presented on the survey design and analysis.


Nova Scotia–Newfoundland–Gulf of St Lawrence (Maritime) Population

Originally, this population is believed to have numbered in the tens of thousands. There has been no evidence for the past 200 years of the population’s re-establishment.

Footnote 3

Note, however, that the West Greenland population is connected in some degree to the Canadian population defined here as the Northern Hudson Bay–Davis St population, while the North Water population is thought to be the same as at least the eastern-Jones-Sound part of the Baffin Bay population.

Return to footnote 3 referrer