Atlantic walrus (Odobenus rosmarus rosmarus) COSEWIC assessment and status report: chapter 11

• Previous page
• Table of contents
• Next page

Limiting Factors and Threats

• Subsistence Hunting
• Contaminants
• Industrial Development
• Disturbance from Noise or Ecotourism
• Climate Change

Walruses are gregarious and valuable, and they have a narrow trophic niche and restricted seasonal distribution that makes them relatively easy for hunters to locate and vulnerable to environmental changes (Born et al. 1995). At present, hunting is the main limiting factor and threat to the Atlantic walrus in Canada. Contaminant uptake, industrial development, noise disturbance, and climate change are minor threats by comparison. Climate change may affect walruses most by exposing them to increased hunting pressure. The species’ susceptibility to disease is unknown.

Subsistence Hunting

Data on the takes of walruses in Canada are incomplete and vary widely in quality. They have been collected by different agencies, using various methods and for different purposes. Between 1980 and 1985 harvest estimates by the Baffin Region Inuit Association (BRIA) and DFO differed widely for some communities (Pattimore 1983a, b, 1985; Guinn and Stewart 1985; Donaldson 1988; J. Pattimore, pers. comm. 1986). In some instances they were simply educated guesses. Few data are available on hunter effort. Wide uncertainties in many of the data points--sometimes 200%--coupled with many missing values preclude meaningful analysis. But they do provide information on the seasonality of the hunts and a sense of the magnitude of the harvest by a community or from a population. Efforts are under way to educate hunters about the importance of reporting both landed and lost animals), and DFO and its co-management partners are working with the communities to improve the reporting system (J. Galipeau, NWMB, pers. comm. 2004).

Between 1977 and 1987 walrus harvest records for Nunavut were collected from DFO, Government of the Northwest Territories (GNWT), Royal Canadian Mounted Police (RCMP) and Hudson’s Bay Company records (Strong 1989). From 1988 to 1996 records are from annual harvest summaries prepared by DFO (DFO 1991, 1992a, b, 1993, 1994, 1995, 1996, 1997, 1999). These summaries include information from hunts monitored by Fishery Officers or GNWT Renewable Resource Officers, reported by Government Liaison Officers, calculated from long-term averages, or estimated using sales slips and trade records. Records from 1997 to 2003 are those reported by community Hunters and Trappers Organizations (HTOs) (P. Hall, DFO, pers. comm. 2003).

In Nunavik, community agents hired by DFO collected the walrus harvest data for 1977–97 (Brooke 1997) and 1997–99 (D. Baillargeon, DFO Quebec, pers. comm. 2003). The 1999–2003 records are from Makivik’s Scientific Research Centre, where samples from each walrus that is taken in Nunavik should be sent for Trichinella assays before the animal is eaten (D. Baillargeon, DFO Quebec, pers. comm. 2003). Where differences were found between the DFO and Makivik data in 1997–2003, the higher value was used. Additional hunting data are contained in the Nunavut Wildlife Harvest Study (Priest and Usher 2004) for the years 1996-2001. The reported harvest data were not corrected for hunting losses and do not include information on the age or sex composition of the harvest. Takes of young animals have on occasion gone unreported (Freeman 1969–70).

When the removal rates of walruses are determined, uncertainties in the reported landed harvest are compounded by uncertainty in loss rates. Few estimates of loss rates exist for subsistence hunts, and none for sport hunts. Loss rates can be high when walruses are killed in the water (Beaubier 1970; Orr et al. 1986), as they sink quickly. Hunters prefer to kill them on land or ice where they are easier to retrieve and butcher. To reduce losses, animals in the water may be harpooned before they are shot, wounded so they can be harpooned before being killed, or killed in shallow water where they can be retrieved with grappling hooks or at low tide. Harpooning a walrus is dangerous, since animals must be approached to within about 7.5 m, and wounded walruses become very aggressive and can sink or capsize canoes or small boats.

Orr et al. (1986) observed a loss rate of 32% at summer hunts in Foxe Basin, and Mansfield (1958) estimated the rate at 30% for the Canadian Arctic. Freeman (1969–70) observed a loss rate of about 30% in two summer (late August) hunts near Walrus Island, and of 38% (Freeman 1974–75) in an autumn (late September) hunt near Southampton Island. These loss rates are higher than those observed at open water hunts in the Avanersuaq (Thule) area of (north-west) Greenland (15-25%; Born and Kristensen 1981 cited in Born et al. 1995) and lower than those of the Alaskan walrus hunts, which averaged 42% over the period 1952 to 1972 (Fayet al. 1994). About 55% of the animals struck and lost in Alaska died immediately and most of the wounded died shortly after being struck. Improvements in the weapons used for the hunts over the period did not alter loss rates but increased the proportion of outright kills among the lost animals. Inuk hunters believe loss rates to be lower (~5%)(DFO 2000). None of these loss rates consider the indirect mortality of calves that are orphaned while still dependent on their mother’s milk.

The hunting mortality that Atlantic walrus populations can sustain is not exactly known. Estimates of sustainable yield range from 3 to 5% for a population that is between 59 and 93% of carrying capacity (DeMaster 1984 cited in Born et al. 1995). It is not known where Canadian stocks stand in relation to their carrying capacities. In the absence of information, specific to walrus, on sustainable yield, DFO (2000) uses yield rates of 2–5% inferred from cetaceans with similar life histories and reproductive patterns. Predictions of sustainable removal are unlikely to be accurate since the population estimates they rely upon are dated and incomplete.

Inuit and Cree from around Hudson Bay who participated in the Hudson Bay program of traditional knowledge studies reported that they “knew walrus better when they were still using dog teams” (Fleming and Newton 2003). This supports assertions below that hunting patterns changed and harvests declined with introduction of the snowmobile and reduction in the need for dog food. However, Davis (1981) suggested that demand for ivory ca. 1980 might, conversely, lead to increased hunting, but it is not known whether this has occurred. While the value of raw tusks has been small relative to hunting costs and the value of the meat (Anderson and Garlich-Miller 1994), the economic benefits of carving tusks to create value-added products may be quite significant.

South and East Hudson Bay Population

Hunters from Inukjuak, Kuujjuarapik, Umiujaq, and Sanikiluaq kill walruses from the South and East Hudson Bay population (Figure 5; Table 1). Hunters from Kuujjuarapik and Umiujaq land the occasional animal in some years, typically in the Sleeper Islands or along the coast of the Hudson Bay Arc, while those from Sanikiluaq and Inukjuak land a few on average each year, mostly from the offshore islands (JBNQNHRC 1988; Strong 1989; Olpinski 1990, 1993; Portnoff 1994; Brooke 1997).

Figure 5: Reported Landed Harvest from the South and East Hudson Bay Walrus Population, 1977 to 2002

See Table 1 for data.

Sources: Sanikiluaq: 1977-78 to 1987-89 (Strong 1989) 1989-90 (DFO 1991), 1990-91 (DFO 1992a), 1991-92 (DFO 1993), 1992-93 (DFO 1994), 1993-94 (DFO 1995), 1994-95 (DFO 1996), 1995-96 (DFO 1997), 1996-97 (DFO 1999), 1997-2002 (DFO unpubl. data); Kuujuarapik, Umiujaq and Inukjuaq: 1977-78 to 1996-97 (Brooke 1977), 1997-98 to 2004-05 larger harvest reported by DFO or Makivik (unpubl data from D. Baillargeon, DFO Quebec, pers. comm. 2003; L Cooper, DFO Ottawa, pers. comm.).

^a Unless a community quota is stated the allowable annual harvest is 4 walrus per Inuk.
^b n/d = no data.
^c nc = no community.

Historically, walruses in southeastern Hudson Bay and James Bay were mainly hunted at uglit, in the open water season (Twomey 1939; May 1942; Manning 1946, 1976; Freeman 1964; Olpinski 1990; Reeves 1995; Fleming and Newton 2003). They were also killed in winter and spring at the floe edge or in spring as they slept on floating ice pans. Most recent hunting in this region has taken place in late summer and fall (September and October) at the Sleeper Islands (Manning 1976; Schwartz 1976; Olpinski 1990, 1993; Portnoff 1994; Brooke 1997; Fleming and Newton 2003). In 1992 and 1993, hunters from Inukjuak visited the Ottawa Islands where they failed to locate walruses before travelling to the Sleepers for successful hunts (Olpinski 1993; Portnoff 1994). Indians in James Bay and southern Hudson Bay seldom travel offshore to hunt walruses (Johnston 1961) but hunted them occasionally in the past (Fleming and Newton 2003).

Over the period 1977-78 through 1989-90 the reported landed harvest from this population averaged 10.6±3.3 animals per year (simple mean of reported values ± 95% CI, n = 13); over the period 1990-91 through 2002-03 it averaged 8.5 ± 2.5 (n = 13). The effects on these averages of changes in the quality of the harvest data and of uneven reporting from Sanikiluaq could not be assessed. Assuming a loss rate of 32% (Orr et al. 1986), hunters may have removed about 15 walruses on average from the population each year since 1989–90.

Northern Hudson Bay–Davis Strait Population

Over 20 Canadian communities hunt walruses from the Northern Hudson Bay–Davis Strait population (Figure 6, Table 2).

Figure 6: Reported Landed Harvest from the Northern Hudson Bay-Davis Strait Walrus Population, 1977 to 2002

See Table 2 for data.

Sources: Nunavut: 1977-78 to 1987-89 (Strong 1989) 1989-90 (DFO 1991), 1990-91 (DFO 1992a), 1991-92 (DFO 1993), 1992-93 (DFO 1994), 1993-94 (DFO 1995), 1994-95 (DFO 1996), 1995-96 (DFO 1997), 1996-97 (DFO 1999), 1997-2002 (DFO unpubl. data); Nunavik: 1977-78 to 1996-97 (Brooke 1997), 1997-98 to 2004-05 larger harvest reported by DFO or Makivik (unpubl data from D. Baillargeon, DFO Quebec, pers. comm., L. Cooper, DFO Ottawa, pers. comm.). For summary statistics for Nunavut 2003-4 and 2004-5 and for Nunavik 2003-4, see Table 6.

^a Unless a community quota is stated the allowable annual harvest is 4 walrus per Inuk.
^b n/d = no data.

Communities along the west coast of Hudson Bay hunt animals from the northern Hudson Bay portion of the Hudson Bay–Davis Strait population (Figure 4). These harvests increase from south to north and hunters often have to travel north into the Coats Island area to find walrus herds (Figure 3; Table 2). There are no recent reports of animals being killed at Churchill, Manitoba. They are rarely taken at Arviat and irregularly at Whale Cove, but more commonly further north (Welland 1976; Gamble 1988, Strong 1989; Fleming and Newton 2003). Timing of the hunts varies among communities. All of the communities hunt animals at the ice edge but the largest harvests are typically taken during the open water season in the Repulse Bay (September–October) and Coral Harbour–Coats Island (July–September) areas (Gamble 1984, 1987a, b, 1988).

Residents of Puvirnituq, Akulivik, Ivujivik, and Salluit regularly take walruses from the Hudson Strait portion of the Hudson Bay–Davis Strait population (Roy 1971; Olpinski 1990, 1993; Portnoff 1994; Brooke 1997). Most animals are killed during the open water season--often in September and October, near Nottingham and Salisbury islands. However, hunters from Ivujivik, Puvirnituq and Inukjuak have landed walruses on occasion at Mansel Island. Hunters from Kangiqsujuaq and communities around Ungava Bay take most of their walruses in August through October at Akpatok Island.

Timing of walrus hunts in the Baffin Region of Nunavut varies among communities. Walruses are typically hunted year-round by Cape Dorset hunters, from February through November or December by Kimmirut and Iqaluit; from May through November or December by Pangnirtung and Qikiqtarjuaq; and in May, July and August by Clyde River (Pattimore 1983a & b, 1985; J. Pattimore, pers. comm. 1986). The hunting patterns of the Cape Dorset people changed around 1970 with the introduction of snowmobiles, which reduced the need to feed dog teams, and with concentration of people at the community (Kemp 1976). Nottingham and Salisbury islands had been important as hunting areas but are now seldom visited, and most walruses are landed along the southern coast of Foxe Peninsula (Kemp 1976; Orr and Rebizant 1987). Kemp (1976) reported a similar reduction in the hunting range and harvest by hunters from Kimmirut and Iqaluit. Fall walrus hunts from Iqaluit have apparently been constrained in recent years by the long distance to the hunting area and the lack of boats of suitable size (DFO 2000).

Over the period 1977–78 through 1989–90, the reported landed harvest from this population averaged 231±32 animals per year (simple mean of reported values ± 95%CI , n = 13); from 1990-91 through 2002-03 it averaged 147±30 (n = 13). Sequential five-year averages of the overall harvest from the population for the periods 1977–82, 1982–87, 1987–92, 1992–97, and 1997–2002 showed an abrupt decline in the reported harvest in the most recent period (i.e. mean±95%CI respectively of 172±54, 183±54, 145±31, 129±30, 68±14). Most of this decline was in Nunavut. The effects on these averages of changes in the quality of the harvest data and of uneven reporting could not be assessed. The decline is less but the pattern remains the same when long-term means are substituted for missing values. The Nunavut Wildlife Harvest Study provides additional data on community reported numbers of walrus killed from 1996 to 2001 (Priest and Usher 2004). Assuming a loss rate of 32% (Orr et al. 1986), and substituting long-term means for missing values, Canadian hunters may have removed about 247 walruses on average from the population each year since 1989–90. The vulnerability of animals from this population to hunting in Greenland waters is unknown.

Foxe Basin Population

Hunters from Igloolik and Hall Beach (Sanirajak), and visiting sport hunters, kill walruses from the Northern Foxe Basin population year-round, but most animals are landed from June through October (Guinn and Stewart 1986; Figure 7; Table 3). The landed harvest appears to be biased towards males, which are larger and have larger tusks. Only 31 of 98 landed animals examined between 1982 and 1984 were female (Orr et al. 1986). From 1977–78 through 1989–90 the landed harvest from this population averaged 218±37 animals per year (simple mean of reported values ± 95%CI, n = 13); from 1990–91 through 2001–02 it averaged 179±42 (n = 12). This difference was due to changes in the recorded takes at Hall Beach, as the means for Igloolik were the same for both periods. The effects on these averages of changes in the quality of the harvest data could not be assessed. Assuming a loss rate of 32% (Orr et al. 1986), and substituting long-term means for missing values, hunters may have removed, on average, 276 walruses from the population each year since 1989–90.

Figure 7: Reported Landed Harvest from Northern Foxe Basin Walrus Population, 1977 to 2001

See Table 3 for data.

Sources: Nunavut: 1977-78 to 1987-89 (Strong 1989) 1989-90 (DFO 1991), 1990-91 (DFO 1992a), 1991-92 (DFO 1993), 1992-93 (DFO 1994), 1993-94 (DFO 1995), 1994-95 (DFO 1996), 1995-96 (DFO 1997), 1996-97 (DFO 1999), 1997-2002 (DFO unpubl. data).

^a Unless a community quota is stated the allowable annual harvest is 4 walrus per Inuk.
^b n/d = no data.

Baffin Bay (High Arctic) Population

Hunters from Resolute, Arctic Bay, Pond Inlet, and Grise Fiord kill walruses from the Baffin Bay population (Figure 8; Table 4). Timing of the hunt varies among communities but most animals are taken from May through September (Pattimore 1983a+b, 1985; J. Pattimore, pers. comm. 1986). Hunters from Grise Fiord also land a few walruses in October, November, February and April. There was a sharp decline in the walrus harvest by Resolute ca. 1965-74 (Riewe 1976) and Grise Fiord ca. 1967 (Riewe and Amsden 1977) when snowmobiles were introduced and less meat was needed to feed sled dogs. However, recent re-introduction of dog teams in Grise Fiord has increased interest in hunting walruses (D. Akeeagok cited in Born et al. 1995).

Figure 8: Reported Landed Harvest from the Baffin Bay (High Arctic) Walrus Population, 1977 to 2002

See Table 4 for data.

Sources: Nunavut: 1977-78 to 1987-89 (Strong 1989) 1989-90 (DFO 1991), 1990-91 (DFO 1992a), 1991-92 (DFO 1993), 1992-93 (DFO 1994), 1993-94 (DFO 1995), 1994-95 (DFO 1996), 1995-96 (DFO 1997), 1996-97 (DFO 1999), 1997-2002 (DFO unpubl. data).

^a Unless a community quota is stated the allowable annual harvest is 4 walrus per Inuk.
^b n/d = no data.

Over the period 1977–78 through 1989–90 the landed harvest from this population averaged 17.6±4.9 animals per year (simple mean of reported values ± 95%CI, n = 13); over the period 1990-91 through 2001-03 it averaged 16.2±5.7 (n = 13). The average annual harvest during the period 1997-2002 (9.2±4.7) was roughly half that of the previous 5-year periods. This pattern did not change when long-term means were substituted for missing values. This change may be real or an artefact of missing data and/or changes in the quality of the harvest data. Assuming a loss rate of 32% (Orr et al. 1986), and substituting long-term means for missing values, Canadian hunters may have removed about 24 walruses on average from the population each year since 1989-90.

Hunters in north-west Greenland take many walruses from the Baffin Bay population (Born et al. 1995). The estimated annual landed harvest in Avanersuaq is 250 walruses and at Upernavik is about 10, which at a loss rate of 25% may remove about 312 animals from the population each year. The vulnerability of animals that summer in Canadian waters to hunting in north-west Greenland waters is unknown, but may be significant since some of the animals are migratory and hunters in Canada find bullets and harpoon heads from Greenland in their kills and vice versa. On occasion, hunters may also travel between jurisdictions to hunt. On 1 March 2004, an executive order on the protection and hunting of belugas and narwhals came into effect in Greenland (Olsen 2004). Whether the resultant quota restrictions will cause harvesters to shift their focus onto other species such as the walrus is unknown, as are any potential effects on shared walrus populations.

NAMMCO (2006) compiled numbers of walrus killed from 1996-2001 using reports obtained from DFO, Greenland, and the Nunavut Wildlife Harvest Study. They estimated an average of 124 animals was removed each year (110 north water, 4 west Jones Sound, 8 Penny Strait / Lancaster Sound). NAMMCO (2005) concluded that the levels of harvest were probably unsustainable. They drew a similar conclusion in 1995.

Limited information suggests the current population is small and continues to be hunted at unsustainable levels in the North Water area of Canada and northwest Greenland. Satellite tracking and genetic information suggests that some animals in this population are resident in the Canadian Archipelago (west Jones Sound and Penny Strait / Lancaster Sound) and are not exposed to over-hunting.

Sport Hunting

From 1928 through 1994, only Inuit could hunt Atlantic walruses in Canada. Now a limited hunt has been opened for non-resident hunters in order to benefit communities with walrus populations and hunts have been approved annually since 1995 (Table 5; P. Hall, DFO, Winnipeg, pers. comm. 2003). Licences for these hunts require annual approval from the Nunavut Wildlife Management Board (NWMB). Non-resident hunters can take the tusks but must leave the meat in the village. Hunts are primarily arranged through outfitters in Igloolik, Coral Harbour and Hall Beach, but licences have also been approved for Cape Dorset, Salluit, and Qikiqtarjuaq. They are becoming increasingly popular and are advertised widely over the Internet. Most animals are taken from northern Foxe Basin and some from northern Hudson Bay. Sport hunters landed a total of 9.4 ± 1.8 (95%CI) walrus each year on average, over the 5-year period 1997 to 2001.

^a Numbers in brackets indicate (licences requested:hunts approved), na = not available.

Source: L. Cooper, DFO, pers. comm.

Contaminants

Cadmium (Cd) concentrations in the soft tissue of Atlantic walruses from Foxe Basin and northeast Hudson Bay are high relative to other mammals (0.03-130.9 µg·g –1 wet wt; Outridge et al. 1994; Wagemann and Stewart 1994), as are their lead (Pb) concentrations (0.02–0.58 µg·g –1 wet wt; Wagemann and Stewart 1994). Mercury (Hg) levels in the muscle ranged from 0.02 to 1.34 µg·g –1 wet wt (Wagemann and Stewart 1994; Wagemann et al. 1995). They were higher on average (0.11 µg·g –1 wet wt, SD 0.13) than those sampled from the Thule area of Greenland in the late 1970s (0.06 µg·g –1 wet wt, SD 0.03; Born et al. 1981). The source of cadmium, lead, and mercury in Atlantic walruses in northern Foxe Basin appears to be natural rather than anthropogenic (Outridge et al. 1997, 2002). Metal concentrations in the tissue of these walruses paralleled that in the tissue of local clams, with the exception of cadmium (Wagemann and Stewart 1994).

Levels of organochlorine contaminants in walruses are generally low, reflecting the low lipid content of their prey. Because they feed low in the food web, walruses typically have 4–10 times lower concentrations of organochlorine contaminants than belugas from the same area, but a similar pattern of residues (Norstrom and Muir 2000). The highest levels are found in individuals that are thought to eat seals, which accumulate these contaminants in their fat (Muir et al. 1995).

The direct and indirect effects of oil on walruses have not been studied. Born et al. (1995) believed that several aspects of the species’s ecology may make it vulnerable to oil pollution, in particular, its gregariousness, which may spread oil from animal to animal, its preference for coastal areas and loose pack ice where oil may be more likely to accumulate, and its reliance on benthic molluscs which may accumulate petroleum hydrocarbons or succumb to the oil.

The effects of chemical contaminants on walruses are unknown (Wagemann and Stewart 1994; de March et al. 1998; Fisket al. (ed.) 2003).

Industrial Development

Commercial fisheries that overlap the walrus’s range could adversely affect them by competing directly for food, disrupting feeding habitats, and causing disturbance. Scallop dragging has been tried in Cumberland Sound, Hudson Strait, and Ungava Bay, and along the Nunavik coast of Hudson Bay, but these fisheries were uneconomical and are no longer operating (Stewart et al. 1993); and walruses feed preferentially on bivalves buried in soft bottoms, not on scallops or on the hard bottoms where they occur. However, the effects of such operations on walrus populations should be assessed if it is proposed to re-open them. Open water trawl or drag fisheries for shrimp, turbot, cod, or other species will not compete directly for food but may disturb walruses or their feeding habitats. Ship noise could displace walruses from their uglit and interfere with their communication (Salter 1979a; Born et al. 1995; Stewart 2002).

Threats posed to walruses in Canadian waters by hydrocarbon and mineral development and exploration are low at present. Indeed, there is less hydrocarbon exploration going on in the high Arctic today than there was in the 1970s and 1980s. (D.G. Wright, DFO Winnipeg, pers. comm. 2002) Both of Canada’s high-Arctic metal mines closed in September 2002 (M. Wheatley, pers. comm. 2003). This will reduce the effects of ice breaking activities on walruses entering Lancaster Sound in the spring. It will also reduce seismic and noise disturbances related to mining activities and the risk of hydrocarbon and heavy metal pollution. However, future developments could reverse this trend. For example, current plans for a diamond mine west of James Bay could increase levels of shipping and other disturbances; other effects, including alteration of the flows and water quality in rivers, could affect marine habitats. The recovery of walrus stocks in southern Hudson Bay and James Bay could be affected.

Industrial developments inland, including dams that would alter seasonal flows in rivers, might possibly affect marine habitats near the mouths of affected rivers. However, there are many hydro dams in the world, and the cumulative evidence that altered flows have had significant effects in the marine environment is not large.

Disturbance from Noise or Ecotourism

Machine noise, particularly from aircraft, disturbs walruses and can stampede them into the water, causing significant calf mortality and spontaneous abortion of foetuses (Salter 1979a; Bornet al. 1995). The response depends upon many factors related to the characteristics of the aircraft and its flight, environmental conditions, and the demography of the affected walruses. There is concern that frequent aerial tourist traffic may disturb walruses hauled out along the Ontario coast of Hudson Bay (C. Chenier, DNR Cochrane, pers. comm. 2003). In Russia, aircraft are prohibited within 50 km of most walrus haul-out sites, except by special permit (U.S. Fish and Wildlife Service 1993 cited in Born et al. 1995).

The reactions of walruses to vessel noise vary depending upon their previous experiences (Born et al. 1995). Animals from hunted populations tend to be skittish when approached by boats but when asleep can sometimes be approached within 10-20 m. Ice breaking activities cause Pacific walruses to enter the water: females and calves when the ship is within 500–1000 m and males when it is within 100–300 m. They move 20–25 km away from the disturbance if it continues, but will return after it stops. Intensive vessel traffic may have a negative effect on walrus distributions. In Russia, uglit are protected by regulations that limit vessel access, usually within 20 km of the site (U.S. Fish and Wildlife Service 1993 cited in Born et al. 1995).

International tours bring visitors from the United Kingdom to view walruses at uglit on Southampton, Bencas, Coats, and Walrus islands in July and August (http:/www.windowsonthewild.com/wow/Walrus/walrus.htm). The extent and effects of walrus ecotourism has not been documented but there is concern among Inuit and scientists that these disturbances could drive herds further into the pack ice or away from their traditional uglit, or cause stampedes (Stewart 2002; Dueck 2003; C. Chenier, DNR Cochrane, pers. comm. 2003).

The effects of pulsed noise from seismic exploration on walruses are unknown, as is the ability of walrus to habituate to non-threatening noises. Underwater noise might disrupt the transmission of songs during the breeding season.

Climate Change

The direct effects of climatic warming or cooling on walruses are likely limited and not necessarily negative. Sea ice cover does not appear to be a critical determinant of walrus populations, since various areas of their habitat are seasonally ice-free (R. Stewart, DFO Winnipeg, pers. comm. 2004) and their pristine distribution extended far south of its present limits. Bornet al. (2003) hypothesized that a decrease in the extent and duration of Arctic sea ice in response to warming might increase food availability for walruses, by increasing bivalve production and improving access to feeding areas in shallow inshore waters. If sea ice were not available walruses would be more likely to haul out on land. Behavioural and physiological responses to changes in air temperature suggest that Pacific walrus calves can maintain their body temperature at an air temperature of 18°C in still air and shade or under equivalent conditions (Fay and Ray 1968; Ray and Fay 1968). Above this temperature they withdraw into the water to avoid over heating. Air temperatures at or above this level for an extended period might disrupt normal feeding, moulting, and calving schedules.

NAMMCO (2006) noted that hunting pressure on walruses will likely increase as the amount and duration of ice cover in Arctic regions declines. Predation by killer whales and polar bears may also increase in the absence of ice as walrus are forced to use terrestrial sites. Killer whales may also remain in the Arctic for longer periods if there is less ice to entrap or restrict their movements.

The fossil record suggests that between 9000 and 1000 y BP, when walrus occupied areas along the east coast of Canada, the summer surface water temperatures of the area may have ranged from 12 to 15°C (Miller 1997). Whether these animals summered in these relatively warm waters or moved north into cooler waters is unknown.

The indirect effects of climate change may pose a greater threat to walruses than the change itself. In the event of warming, human populations in the north might increase and expand into hitherto unpopulated areas; in the event of cooling, walruses may be forced southward closer to existing communities. In either case, a reduction in suitable habitat is likely and they would become more vulnerable to hunting and disturbance.