COSEWIC assessment and update status report on the wolverine Gulo gulo in Canada
- Assessment Summary
- Executive Summary
- COSEWIC Mandate, Membership and Definitions
- Lists of Figures and Tables
- Species Information
- Population Sizes and Trends
- Limiting Factors and Threats
- Special Significance of the Species
- Existing Protection or Other Status
- Summary of Status Report
- Technical Summary
- Acknowledgements and Literature Cited
- Biographical Summary of Contractor and Authorities Consulted
Population Sizes and Trends
- Yukon Territory
- Northwest Territories
- Nunavut Territory
- British Columbia
- Quebec and Labrador
- New Brunswick
- National Parks
- Canadian Wolverine Population Estimate
- Coterminous United States
The earliest harvests attributed to specific Canadian jurisdictions indicate that wolverine populations, based on harvest success, may have been declining in the Prairie Provinces in the 1920s and 1930s (Novak et al. 1987). Harvests in Quebec, Labrador and Ontario were already low by that time. Fur harvest is dependent on factors in addition to furbearer population levels, including weather and economic conditions (fur prices and alternative employment opportunities) which influence trapper effort.
Low densities, large home range sizes, and long-distance movements by dispersing individuals contribute to the wolverines’ vulnerability to trapping. Despite these factors, some local populations appear to have recovered from local overharvest. Improved fur harvest systems, such as registered traplines, and trapper education, applied over several decades, have contributed (Johnson 1990). Exceptions appear to be the eastern wolverine population, which was probably historically low and remains either very rare or extirpated. Populations in the Pacific region, including Vancouver Island and the lower mainland, have similarly become very rare or extirpated.
Wolverines are considered high priority for management attention in the Yukon, where the population is monitored using fur harvest statistics, winter track counts and an annual trapper questionnaire to obtain local knowledge about populations and trends from all licensed trappers (Slough et al. 1987, Slama, pers. com., 2002). All evidence points to a healthy and stable population in all regions of the territory over the past 20 years (Slama, pers. com., 2002). Trappers report wolverines to be common, and harvests are not declining relative to trapping pressure (Slama and Jessup, pers. com., 2002). Wolverine pelts must be sealed in the Yukon, including pelts held over or used in the territory, and those taken by First Nation’s trappers. Wolverines harvested by Inuvialuit trappers are not sealed and enter the Northwest Territories. Trappers have reported wolverines to be particularly abundant since the snowshoe hare population crash in 1999 (Jung, pers. com., 2002). Alternative hypotheses to numerical abundance include the increased trappability of wolverines when food supplies decrease (Banci 1987), or an increase in dispersing juveniles following reproductive success at the hare peak.
There are few long term data sets on wolverine abundance, other than inferences from fur harvest and trapper questionnaire data; however, long term winter track count data has been collected in the Yukon. O’Donoghue et al (2001) reported 0.69 to 15.83 wolverine tracks per 100 km-days (transect length in km x days since last snowfall) of transect in the Kluane area between 1986-87 and 1996-97. Wolverine tracks were most common during the second winter of hare population decline (1991-92; 5.68 tracks per 100 km-days) and during the subsequent hare increase phase (5.64 and 15.83 tracks in 1995-96 and 1996-97, respectively). In the Teslin burn, 300 km to the east, B. Slough (unpubl. data) recorded 0 to 20.44 wolverine tracks per 100 km-days between 1986-87 and 1993-94. The peak track density occurred in 1993-94, during the third winter of hare decline in that area.
Banci and Harestad (1990) estimated wolverine density to be 10.75/1000 km2 in continuous and saturated habitat in the south-central Yukon. The estimate using known wolverines (assuming variable habitat quality and not all habitat is saturated) was 5.65/1000 km2. The total population estimate for about 85% of the territory was a resident population of 2503 wolverine, and a fall population of 4171, including juveniles and transients (Banci 1987). The estimate was based on habitat suitability estimates (area trapped, trapper-years and numbers of wolverines harvested) which were biased. Banci (1987) assumed that “trapper-years” were a standard unit of effort among ecoregions; however, a trapper’s effort on small, easily accessible trapping concessions near communities and roads is typically much greater than that on large, remote, fly-in concessions. This is true in terms of numbers of traps set, area covered, and length of trapping sessions. Banci’s estimates based on this method are believed to be underestimates for the Yukon. Wolverine density estimates (e.g., Banci 1994) are remarkably consistent across North American ecozones when estimates based on habitat suitability ratings are disregarded. A more realistic wolverine population estimate for the Yukon, based on 100% coverage of the territory and more homogeneous densities in high quality habitat, is 3500 to 4000 residents.
Large scale wolf poisoning through the 1970s had a detrimental impact on wolverines. Poisoning not only removes wolves as a provider of carrion, but the non-target kill of wolverines is significant. Furthermore, wolverines scavenge poisoned wolves, although apparently do not consume stomachs containing poisoned bait (reported in Kelsall 1981). Most of Rausch and Pearson’s (1972) 198 Yukon wolverine carcasses were collected from predator control stations. Wolf control using sterilization techniques and community-based trapping programs continues as an ungulate management tool. Aerial wolf control was conducted in two large regions of the Yukon in the 1990s; however, the effect of these programs on wolverines is not known.
The Wildlife Management Advisory Council (North Slope) (1998) reported a harvest of about 10 to 13 wolverines per year by Inuvialuit trappers (from the Northwest Territories) on the Yukon’s North Slope (i.e., Arctic ecological area). The population size there is unknown and the only threat is harvest pressure. Wolverines are often observed in association with the Porcupine caribou herd.
It has generally been assumed that wolverine densities are variable in the Northwest Territories, being higher in the mountains of the west and in the taiga, and lower in the tundra habitats of the north and east (Figure 3) (Poole, pers. com., 2002). However, large herds of barrenground caribou and associated wolves in the taiga and tundra habitats suggest that wolverines may be more common in those habitats than was once believed (Mulders, pers. com., 2003). The Kitikmeot wolverine carcass study in Nunavut has changed attitudes about wolverine population densities in that region. Wolverines are absent from Banks Island.
Wolverine fur is valued by northerners given its ability to deflect or minimize frost build up. As a result, many wolverine pelts remain in the north and are used domestically for parka trim. The fur trade statistics for the Northwest Territories are based on furs exported to fur auction and not total harvest (Table 1). Based on carcass collection programs in Nunavut Territory (Table 1, discussed in the Nunavut Territory section below), the harvest of many communities may be greatly underestimated. Land claim agreements require harvest studies in three settlement areas in the Northwest Territories, Sahtu, Gwich’in, and Inuvialuit. All hunters and trappers in the communities are interviewed annually regarding their wildlife harvest. Harvests from the Sahtu region were reported to range from 5 to 12 wolverine (1998 to 2001; Bayha, pers. com, 2002) and in the Gwich’in settlement area the harvest ranged from 4 to 14 (Rose 2002). The Inuvialuit harvest ranged between 21 and 62 annually from 1986 to 2000, except in 1997 when the harvest by Inuvik hunters and trappers was estimated to be 62 wolverines, and the total harvest was 124 (0 to 5 in other years) (Fabijan 1991a, 1991b, 1991c, 1995a, 1995b, 1995c, 1995d, 1996, 1997, 1998, 2000; Pinard 2001). The Government of the Northwest Territories is considering establishing a territory-wide wolverine carcass collection program to obtain better data on regional harvest levels and patterns, and biological data (Mulders, pers. com., 2003).
Six communities in the Inuvialuit Settlement Region (Aklavik, Holman, Inuvik, Paulatuk, Sachs Harbour and Tuktoyaktuk) have adopted community conservation plans (e.g., Wildlife Management Advisory Council (WMAC-NWT) 2000a and 2000b). Wolverine conservation measures include the identification and protection of important habitats from disruptive land uses, avoidance of den disturbance, and discouraging hunting in summer.
The wolverine populations across the Northwest Territories are believed to be stable, but harvesting pressures and increasing levels of non-renewable resource development may lead to further habitat loss and fragmentation, which may adversely affect wolverine distribution and abundance in the future (Mulders, pers. com., 2003). Studies on the central barrens (Mulders 2000) have shown a healthy wolverine population in that region. There is a pilot project in the barrens using hair-snagging to collect DNA for individual identification and population estimation. GPS collars may also be deployed to assess the effectiveness of the hair sampling grid and to obtain movement data (Mulders, pers. com., 2003).
Studies of genetic variability of Northwest Territories and Nunavut wolverines have shown moderate male-mediated gene-flow among populations, and suggest that all western populations studied were connected (Wilson et al. 2000, Kyle and Strobeck 2001, Chappell 2002, Kyle and Strobeck 2002). Mitochondrial DNA analyses, however, show the genetic independence of populations caused by natal range fidelity exhibited by females (Wilson et al. 2000, Chappell 2002). The implications for wolverine conservation are that individual populations should be conserved and travel and dispersal corridors must be maintained in the matrix between population refugia.
Sources: Statistics Canada, Census of Wildlife Pelt Production, except:
BC – 2000/01 (B.C. Fur Harvest Return System).
ON – 2000/01 (Dawson, pers. com., 2002).
MB – 2000/01 (Berezanski, pers. com., 2002).
SK – 2000/01 (Arsenault, pers. com., 2002).
AB – 2000/01 (Kosinski, pers. com., 2002).
YT – 2000/01 (Slama and Jessup, pers. com., 2002).
NT – 2000/01 (Erasmus, pers. com., 2002).
NU – Carcass collections; Dumond, pers. com. 2003).
Note: Pelt production statistics represent each jurisdiction’s best estimate of total harvest based on furs exported to auction or furs sealed. Furs retained for domestic use are not included in the harvests of Northwest Territories and Nunavut although they represent a significant proportion of the total harvest. This is discussed further in the text.
1 Northwest Territories includes Nunavut to 1991/92.
2 Data collection began for Nunavut in 1992/93 for political reasons related to an Inuit land claim, in preparation for the establishment of Nunavut as a territory, which occurred on April 1, 1999. Data in parentheses is based on carcass collections.
3 Not including carcass data for Nunavut.
Wolverine densities are moderate in the west and low on the arctic islands and eastern Nunavut (Figure 3), where numbers are believed to be stable, but sensitive to harvest pressures. As in the Northwest Territories, wolverine fur is prized, and most of the pelts are used domestically for parka trim. The fur trade statistics for Nunavut, too, are based on furs exported to fur auction and not total harvest. Table 1 shows fur trade statistics compared with numbers of carcasses collected in the Kitikmeot wolverine carcass study (Dumond and Krizan, pers. com., 2002), and it serves to illustrate that the fur trade statistics underestimate harvest in Nunavut as in the Northwest Territories. The carcass numbers are believed to represent up to 90% of the total wolverine harvest (Mulders, pers. com., 2003). Nunavut became a territory on April 1, 1999, but pelt and carcass data are presented for the region from 1992-93.
Wolverines may be especially vulnerable on the arctic tundra, where visibility and snowmobile access are good. Moderate to high harvest may take place near communities and concentrations of caribou, depending on recruitment from neighbouring population refugia. Studies of wolverine metapopulation dynamics and the genetic and morphological variation within and among wolverine populations in Nunavut and other jurisdictions are summarized in the Northwest Territories discussion above.
Wolverine numbers are believed to be stable over much of British Columbia, where they have been estimated at 2089 to 3567 individuals (Lofroth, pers. com., 2003). Densities estimated by Quick (1953) in northeast British Columbia were 4.76/1000 km2. Wolverines may be declining in the southeast, where habitat and disturbance pressures are greatest; however, no trend data are available (Lofroth, pers. com., 2002). Krebs and Lewis (2000) and Krebs (pers. com., 2003) estimate an average density of 6.16/1000 km2 in southeastern British Columbia. Forestry, ranching, oil and gas development, human access and human settlement are negatively impacting habitats in British Columbia. Declining caribou herds, linked to logging activities, are a threat to wolverines. A Revelstoke area wolverine sample showed signs of genetic differentiation from the more northern populations (Kyle and Strobeck 2002).
The Vancouver Island population is believed to be very low, and possibly extirpated (Lofroth, pers. com., 2002) as there have been no sightings since 1992 (Appendix 1). There are many threats to wolverines on Vancouver Island including extensive clear-cut forest harvesting, human settlement, and human activity, including several transportation corridors. The Vancouver Island marmot, a potential summer food for denning females, is endangered (COSEWIC 2001).
A rabies control program in the 1950s, involving the non-selective poisoning of about 5500 wolves, is believed to have had a serious impact on wolverines which would have taken decades to recover from (Petersen 1997). Petersen (1997) believed that trapping pressure for wolverines was low, since the low density populations would be difficult to target, even with the incentive of high fur prices. Experience from the southern Yukon, where wolverines are common, suggests that they can be successfully targeted using selective trapping methods. The quota of one wolverine per trapline further limits trapper effort.
The extent of recent habitat loss and changes in ungulate populations is unknown, but land use activities such as agriculture, forestry, and oil and gas development are occurring within the wolverines’ present range. Recent opinion surveys of trappers in Alberta indicated a stable population north of 56°N latitude but declining elsewhere in 1987, and a province-wide declining population in 1994 (Petersen 1997). Wolverines are currently most abundant in the western areas. The current range and relative abundance of wolverines in Alberta (Figure 3) were inferred largely from fur harvest data. Poole and Mowat (2001) noted a decrease in the density of harvest locations over time, but low sample sizes hampered this analysis. A decline in wolverine harvests noted since the 1970s was due in part to restrictive quotas which were imposed in the late 1980s (Poole and Mowat 2001). Wolverine harvests were correlated with lynx harvests and lynx pelt price, suggesting the opportunistic harvest of wolverines while trapping lynx. The Alberta Research Council and the Alberta Fish and Wildlife Service are experimenting with the use of hair capture techniques, camera traps and snow tracking as long-term monitoring tools to assess wolverine distribution and abundance trends in boreal and montane Alberta (Mowat 2001, Besko and Wilkinson, pers. com., 2002, Fisher 2003).
Wolverine populations are not monitored in Saskatchewan, where they are believed to be rare in the southern boreal forest, but common in the north (total population < 1000) and possibly declining. The major threats are trapping pressure and habitat fragmentation from new roads within their range (Keith, pers. com., 2002). Habitat losses due to forestry and other land uses are also occurring.
Wolverines inhabit the northern part of the province, north of 53° latitude (Berezanski, pers. com., 2002). Highest densities appear to be in the northeast and northwest; however, the northcentral region receives less trapping effort and may represent a population reservoir (Berezanski, pers. com., 2002). The species was historically rare in southern Manitoba, where human developments forced the range limit north. The genetic structure of wolverine populations of Manitoba and northwestern Ontario is similar, but relatively distinct from other populations (Kyle and Strobeck 2002), suggesting that gene flow may be limited between these peripheral populations and other sources. Conservation efforts, including the re-establishment of gene flow between peripheral and core populations, were recommended. Chappell (2002) also found a high level of structuring of mitochondrial DNA among 9 wolverine populations in the Northwest Territories, Nunavut and Manitoba, indicating site fidelity among female wolverine and male-biased gene flow.
The wolverine is apparently not as rare as once thought in Manitoba (van Zyll de Jong 1972, 1975; Holbrow 1976). Van Zyll de Jong’s (1972) population estimate of 60 wolverines was based on fur sales records. A population increase likely followed a cessation of indiscriminate wolf poisoning in the mid-1970s, a subsequent increase in the wolf population, and finally, the adoption of a limited harvest season (Johnson 1990). Johnson (1990) estimated the Manitoba wolverine population at 500 to 800, and more recently the population has been estimated at 1200 to 1600 (Berezanski, pers. com., 2002). The population is thought to be stable to increasing at present and there have been many recent sightings across southern Manitoba. Wolverine populations in the northeast may be benefiting from the increase in the Pen Island caribou herd, which is having the same beneficial effect in Ontario (Dawson 2000). Current threats include logging. Wolverine harvest effort is expected to decline concurrent with a decline in the number of active trappers and a transfer of effort to more lucrative furbearers (Berezanski, pers. com., 2002).
Wolverines are found in small numbers in northwestern Ontario following declines since the 1800s (Dauphiné 1989). There is some evidence that numbers increased in the Fort Severn area since the 1970s concurrent with increases in caribou numbers (Dawson 2000). The Pen Islands caribou population may have declined since 1994, and wolverine harvests also declined in the 1990s (Dawson, pers. com., 2002). The small population of wolverines on the coast of James Bay in the Cape Henrietta Maria region may have been extirpated in the 1970s (Dawson 2000). Wolverine range in Ontario is still not clearly defined, but they appear to be recolonizing some areas in northwestern Ontario (Dawson 2000). Kelsall (1981) believed that the wolverines in Ontario were isolated from those in Quebec and Labrador, and there is no recent evidence to the contrary.
Timber harvesting is occurring in the southern portion of wolverine range (Dawson 2000) and is a threat to populations. A zero quota was applied to wolverine prior to the 2001-02 trapping season; however, 3 accidental and nuisance kills occurred (Dawson, pers. com., 2002). Treaties in northern Ontario allow for Aboriginal peoples to harvest in a manner to which they were accustomed to prior to treaty signing (Heydon, pers. com., 2002). This allows them to trap wolverine for their own use. These rights may be infringed upon for the protection of a threatened species; however, the degree to which a right may be infringed upon is evolving and not set in policy at present.
A camera trapping project was initiated in March 2001 in the Red Lake area to refine the southern limit of wolverine distribution. Aerial surveys conducted in February 2003 indicated a relatively continuous distribution in northwestern Ontario, where 103 tracks were intersected on 5700 km of transects (Magoun, pers. com., 2003). The area to the east will be surveyed in 2004. Additional studies include the collection of Aboriginal Traditional Knowledge (Dawson, pers. com., 2002).
Wolverines have not been confirmed to occur in Quebec since 1978, and the population status at the present time is uncertain (Fortin et al. 2002). There is a consensus among local biologists that the species is either extremely rare or extirpated. The species was never common. Historic wolverine harvest data for Quebec are considered unreliable as there is no proof that any of the pelts attributed to Quebec (or Labrador) came from that region, since fur trading companies operated over a very large area (Obbard et al. 1987). There have been close to 60 unconfirmed sightings from both jurisdictions since 1935, but these may be immigrants, rather than indicators of a local breeding population. Sightings are believed to be underreported. The last capture was in the Schefferville area in 1978.
Wolverines have not been verified in Labrador since the 1950s (Brazil, pers. com., 2002); however, about 16 sightings have been reported since then. A species at risk stewardship program has been initiated to inform communities about the status of wolverines and other species at risk (McNeill, pers. com., 2002). Aboriginal traditional and local knowledge, including historical and recent observations of wolverine are being collected. The most recent unconfirmed sighting was near Nain in April 2002, when E. Merkuratsuk observed and followed tracks (reported by McNeill, pers. com. 2002).
Population declines of the eastern wolverine are thought to be due to trapping and hunting in the late 19th C., dwindling caribou herds in the early 20th C., human encroachment on habitat, reductions in the numbers of wolves, and the indiscriminate use of poison baits. Hunting and trapping wolverines is now prohibited, and wolves and caribou have recovered substantially. Despite the reversal of these negative ecological factors, there has been no evidence of a coincidental recovery of wolverines. The accidental capture of wolverines in traps set for other species poses some risk to recovery. Logging and hydroelectric reservoirs consume wolverine habitat. The area of suitable habitat remaining has been estimated at 500 000 km2.
A draft recovery plan (Fortin et al. 2002) has the following 4 main objectives:
- achieve a population of 100 wolverines (a minimum viable population).
- maintain this population for 10 years (when natural increases should be realized).
- prevent wolverine losses attributable to human activities, and
- ensure that habitats are available in sufficient quantity and quality to attain the population objective
Evidence of the wolverine’s presence in New Brunswick is sketchy, consisting entirely of brief mention in early written works (Sabine, pers. com., 2002). Wolverines have been extirpated from New Brunswick, probably since the early 1800s. There have been no recent sightings, and there are no plans for reintroduction of the species to the province.
Parks Canada maintains a database on species at risk, including observations and population status assessments (Alvo, pers. com., 2002). Wolverine population assessments are generally not based on field studies per se. There is an ongoing study in Vuntut National Park in the Yukon to obtain information on the status and ecology of mustelids, in particular wolverine (Henry, pers. com., 2002; Henry 2003). The study uses winter track-transects, carcass collections, Aboriginal traditional knowledge, and local knowledge. There is a continuing wolverine track monitoring program in Kluane National Park and Reserve, Yukon (Henry, pers. com., 2002), and a wolverine track survey was conducted in the Chilkoot Trail National Historic Site, in British Columbia, in 2003 (Slough and Rivard 2003). Poll (reported by Alvo, pers. com., 2002) estimates the combined population of Mount Revelstoke and Glacier National Parks to be 25 wolverines (included in the British Columbia population estimate). Wolverine trapping by Aboriginal peoples is permitted in all national parks in the 3 northern territories, and in the Alberta portion of Wood Buffalo National Park, and Wapusk National Park in Manitoba (where licensed trapping is also permitted). Trapping effort is believed to be minimal, however.
Western wolverine population estimates (mature individuals) are available only for the Yukon Territory (3500 to 4000), British Columbia (2089 to 3567), and Manitoba (1200 to 1600). The available information suggests that populations in the remaining jurisdictions are approximately as follows: Alberta (1500 to 2000, wolverine range is relatively extensive, with some high density areas), Saskatchewan (1000, likely less than in Manitoba), Ontario (300 or less, based on small range area), Northwest Territories (3500 to 4000, similar to Yukon range area and relative abundance), and Nunavut (2000 to 2500). The total western Canadian wolverine population is therefore estimated at 15 089 to 18 967. Assuming that current harvests are sustainable, then the fall (pre-trapping) population estimate, including juveniles is 2.5% (Golden, pers. com. 2002) to 6% (J. Krebs, pers. com., 2002) or 8% (Gardner et al. 1993) higher than the harvested population; as high as 20 484 wolverines. Considering that average annual harvests of wolverines have exceeded 500 animals, or 2.5% of this population estimate, over the past 13 seasons (Table 1), it appears that the harvest of wolverines is currently sustainable. It is also likely that some jurisdictional wolverine population estimates are low.
The eastern wolverine population in Quebec and Labrador, based on the opinions of eastern jurisdictional biologists, is extremely low and close to extirpation. Recent unconfirmed sightings are still being investigated (Brazil, pers. com., 2003).
Wolverine densities in south-central Alaska have been estimated at 4.78/1000 km2 (Whitman and Ballard 1983), 4.69/1000 km2 (Becker and Gardner 1992), and 5.2/1000 km2 (Becker 1991) which are similar to an earlier estimate of 4/1000 km2 by Quick (1953). Densities on the north slope of Alaska have been estimated at 7.2 to 20.8 wolverines per 1000 km2 (Magoun 1985, based on home range size). An ongoing study in northwestern Alaska should provide further information on wolverine demographics (Shults, pers. com., 2002). A harvest assessment program indicates that wolverines are abundant and able to withstand a moderate to high harvest near villages, exploiting recruitment and immigration from refugia. There is no wolverine population estimate for Alaska (Golden, pers. com., 2003).
The total number of wolverines inhabiting the lower 48 states may be less than 750 (Predator Conservation Alliance 2001). Tenuous populations currently inhabit montane regions in Washington, Oregon (est. of 100 in the Cascade Mountains), California (unknown population), Idaho (est. of 300), western Montana (est. of 300), Wyoming (est. of 50) and Colorado (Banci 1994, estimates from Predator Conservation Alliance 2001). Remnant populations also exist in the southern Rockies (Colorado; Kahn and Byrne 1998) and, possibly, Michigan and Maine.
Estimates of wolverine density in Montana ranged from 15.38/1000 km2 in high quality habitats in northwest Montana (Hornocker and Hash 1981) to as low as 5 to 6.67/1000 km2 in fringe habitats (Hash 1987). Copeland (1996) estimated a density of 4.0-5.1/1000 km2 in Idaho. Without specific studies, estimates of population sizes and trends across most of the wolverine’s range in Washington, Oregon, California, Idaho, Montana, Wyoming and Colorado are difficult to assess, and rely largely on accidental captures and reported sightings. A wolverine conservation strategy is being developed by members of the Western Forest Carnivore Committee, starting with historical record evaluation (Quade, pers. com., 2002). Plans to reintroduce wolverines to Colorado are on hold (Wait, pers. com., 2002). Edelmann and Copeland (1999) recommend maintaining and enhancing movement corridors between mountainous habitats in Idaho and Oregon to ensure the colonization of all habitats and regional population persistence.
- Date Modified: