Skip booklet index and go to page content

Arctic Ecozone+ Status and Trends Assessment

Description of the Condition of the Arctic Ecozone+

Ecosystem structure

Changes in the extent and quality of important biomes

Status: land cover classification and biomes

There are several classification systems that define biomes slightly differently. This report uses a generalized land cover classification system derived from the Millennium Ecosystem Assessment (Millennium Ecosystem Assessment, 2005; Frisk, 2011). In the Arctic Ecozone+, polar barrens, polar tundra, and wetlands cover the largest areas (Figure 57). Lakes and rivers, which comprise a small proportion of the total area of the ecozone+, were excluded from the analysis. The five broadly defined biomes in the Arctic ecozone+ are defined as:

  1. Polar tundra – treeless regions with greater than 50% vegetation cover
  2. Polar barren – treeless region with less than 50% vegetation cover
  3. Wetlands – areas saturated for sufficient time to promote wetland or aquatic processes
  4. Mountains – steep and high lands with specific criteria related to elevation and slope
  5. Snow/ice/glaciers – masses of perennial snow and ice with definite lateral limits, typically flowing in a particular direction

Climatically, “Arctic” is defined as the region north of the 10°C summer isotherm; ecologically, it is the area north of the treeline.

Figure 57. General land cover classes (biomes) in the Arctic Ecozone+.

See Appendix 2: Detailed land cover classes for more information.

Pie chart
Source: Ahern (2010); Frisk (2011); Ahern et al. (2011), using data from the Canada Centre for Remote Sensing
Long description for Figure 57

This pie graph shows the percentage of the five general land cover classes in the Arctic Ecozone+.  Polar barren covers 56%, polar tundra covers 24%, wetlands cover 10%, snow/ice/glacier covers 7%, and mountain/alpine covers 3%.

Top of Page

A more detailed analysis of land cover in the Arctic Ecozone+ (Olthof et al., 2005) has been developed by the Canadian Centre for Remote Sensing and was analysed for this report (Ahern, 2010) (Figure 58). This analysis uses 14 land cover classes, 12 of which are found in the Arctic Ecozone+ (see Appendix 2: Detailed land cover classes for more information) (Ahern, 2010; Frisk, 2011; Ahern et al., 2011). This analysis classifies the Arctic Cordillera as mostly snow and ice, mountains, and sparsely vegetated bedrock or till colluvium. The Northern Arctic is dominated by barren ground or sparsely vegetated bedrock, till colluviums, or barren ground. The Southern Arctic is dominated by low and dwarf shrubs.

Figure 58. Detailed land cover in the Arctic Ecozone+.
Source: Olthof et al. (2005) with additional analysis provided by Ahern (2010)
Long description for Figure 58

This map shows the locations of the different land cover types in the Arctic Ecozone+. The Arctic Cordillera is mostly snow and ice, mountains, and sparsely vegetated bedrock or till colluvium. The Northern Arctic is dominated by barren ground or sparsely vegetated bedrock, till colluviums, or barren ground. The Southern Arctic is dominated by low and dwarf shrubs.

Top of Page

Circumpolar perspective

Analyses conducted for the production of the circumpolar Arctic vegetation map (Walker et al., 2005) produced the following summary results:

Within the Arctic (total area = 7.11 x 106 km2), about 5.05 x 106 km2 is vegetated. The remainder is ice-covered.

About 26% of the vegetated area is erect shrublands, 13% peaty graminoid tundra, 13% mountain complexes, 12% barrens, 11% mineral graminoid tundra, 11% prostrate-shrub tundra, and 7% wetlands.

Canada, among Arctic nations, has the largest portion of Arctic terrain (36%) and has by far the most in the High Arctic ecosystem categories (63% of the global total), mostly associated with abundant barren types and prostrate dwarf-shrub tundra, whereas Russia has the largest area in the Low Arctic, predominantly low-shrub tundra.

Habitat diversity is also an important component of ecosystem structure. Some habitat types assume a particular importance because they are rare and are important, even critical, to the life cycle of particular species. The following examples illustrate the patchy distribution and importance of small habitat elements in a large landscape:

  • Eskers occupy a small proportion of the Arctic but may be critically important as denning habitat for wolves, grizzly bears, and ground squirrels (Mueller, 1995; McLoughlin et al., 2004).
  • Grizzly bears in the Mackenzie Delta can only den in river cut-banks in consolidated, upland soils of Pleistocene origin (or, rarely, pingos), not in more recent fluvial deposits of Holocene origin (Harding, 1976).
  • Nesting shorebirds and geese tend to associate with wetland and riparian habitats that are unevenly distributed on the coastal plain (Brown et al., 2007c).

Top of Page

Trends in polar biomes
Reduction in tundra

There is evidence of approximately a 20% decrease over the past 25+ years in the extent of Tundra Climate, a measure of the cold temperature and low precipitation conditions that support the polar tundra, barrens and ice and snow biomes (Figure 59) (Wang and Overland, 2004). The steady decline in the area of Tundra Climate is matched by reductions in the areas with a tundra signature in primary productivity as measured by NDVI and the strongest changes in Tundra Climate have occurred in Northwest Canada in the past 25+ years (Wang and Overland, 2004; Bhatt et al., 2010, and clearly visible in Figure 63).

Figure 59. Changes in area of the tundra climatic zone around the circumpolar Arctic since 1982, based on NDVI and Köppen Climate Classification System.

"NCEP" and "CRU" are two datasets of monthly surface air temperatures. "NDVI" is a calculation based on light reflected by vegetation, as detected by satellite (see report section on Primary productivity).

Source: Wang and Overland (2004)
Long description for Figure 59

This line graph shows the changes in area of the tundra climatic zone around the circumpolar Arctic since 1982, based on NDVI and Köppen Climate Classification System.  The graph shows the area of tundra climate declined steadily from 1980 to 2000.

Top of Page

Figure 60. Spatial distribution of difference in areas of tundra climate using the Köppen climate classifications between the two years 1978 and 1998.

Based on CRU climate dataset.

Source: Wang and Overland (2004)
Long description for Figure 60

This figure has three maps of the circumpolar Arctic.  The first two maps show the spatial distribution of difference in areas of tundra climate using the Köppen climate classifications between the two years, 1978 and 1998. These maps compare the area of the following climate classifications: temperate group, boreal group, tundra group, and frost group.  The third map shows the area of tundra lost in this decade.  These maps show a significant loss of tundra and a northward expansion of the boreal group between 1978 and 1998.

Top of Page

The warming climate may have been the cause of increased biomass (inferred from increased NDVI) since 1986 on the Porcupine caribou range in northern Yukon and adjacent Alaska (Henry et al., 2012) (Figure 61). Structural changes south of the Arctic Ecozone+ also influence ecosystems within the ecozone+. For example, increased fire frequency is believed to be the reason for a decrease in the cover of forest tundra woodland in the winter range of the Bathurst caribou herd (Chen et al., 2013). As discussed in the section on Primary productivity (page 71), increases in biomass have been measured since 1980 in permanent plots at a High Arctic site (Figure 48 and Figure 49). These increases in biomass are likely responses to the warming climate (Figure 50).

Figure 61. Changes in NDVI across the annual range of the Porcupine Caribou Herd in northern Yukon and adjacent Alaska, 1986 to 2005.

The Yukon portion of this analysis includes part of the Southern Arctic and part of the Taiga Cordillera immediately to the south. Change is relative to a 2000 baseline.

Source: Henry et al. (2012) (Supplementary data, Figure 3) based on Chen et al., unpublished data
Long description for Figure 61

This figure has two heat maps showing the changes in NDVI across the annual range of the Porcupine Caribou Herd in northern Yukon and adjacent Alaska from 1986 to 2005.  The maps showbiomass (inferred from increased NDVI) has increased since 1986 on the Porcupine caribou range in northern Yukon and adjacent Alaska.

Top of Page

Comparisons of historical and contemporary aerial photographs also provide evidence that Arctic vegetation has undergone significant shifts in recent decades. Increased shrub cover has been confirmed in two repeat photography studies in northern Alaska (Sturm et al., 2001; Tape et al., 2006) and one on Herschel Island in Canada (Myers-Smith et al., 2011b). A study in the Mackenzie Delta region of Canada’s Southern Arctic showed that there has been an increase in deciduous shrubs in the region over the past 40 years (Lantz and Kokelj, 2008). A similar dramatic increase in shrub cover was documented by Tremblay et al. (2012) for the region around the community of Kangisuujuaq in Nunavik east of Ungava Bay (Figure 62).

Figure 62. Changes in shrub cover in an area near Kangiqsualujjuaq, Nunavik, Quebec, 1964 to 2003, based on air photo analysis.

Values are percentages of the total study area (7.2 km2). Arrows show direction of change from one vegetation cover type to another, accompanied by the amount of change.

Source: Tremblay et al. (2012)
Long description for Figure 62

This figure shows representative photographs of three different vegetation cover types--shrubless/treeless (<10% cover), discontinuous (10–90% cover), and continuous (>90% cover)--and the percent of the study area that was covered by each in 1964 and 2003. Shrubless/treeless had 45.1% shrub cover in 1964, compared to 29.3% in 2003. Discontinuous had 20.7% shrub cover in 1964, compared to 26.5% in 2003, and Continuous had 34.2% shrub cover in 1964 compared to 44.2% in 2003.

Top of Page

Plot-level data offer a more detailed view of vegetation changes than either remote sensing or aerial photography. A synthesis of resampled plots in tundra ecosystems around the world showed changes in vegetation over the past 10–30 years that are consistent with responses to warming, but with considerable variation among the sites (Elmendorf et al., 2012b). The responses were strongly controlled by the initial conditions of the site and the changes in soil moisture (Callaghan et al., 2011c; Elmendorf et al., 2012b). Resampling of plots established in the 1970s in Alaska yielded results consistent with a warming and drying trend, in which moist and wet community types tended to be replaced by dry community types through time (Hinzman et al., 2005). After resampling of control plots from a 13-year experiment at Toolik Lake, Alaska, Shaver et al. (2001) found that graminoids, mosses, and lichens were decreasing and evergreen shrubs were increasing in abundance. In a similar resampling study at a High Arctic site on Ellesmere Island, Hill and Henry (2011) found a significant increase in biomass of wet sedge tundra between 1980 and 2005 (see the section on Primary production (biomass) in tundra ecosystems on page 75).

Increased growth and cover by shrubs and infilling by tree species also threatens tundra at its southern margins in the forest-tundra ecotone. Correlations between temperature and treelines throughout the paleoecological record strongly support the notion that climate warming will advance treelines. For example, during the Little Ice Age, shrub tundra increased while treelines retreated (Tinner et al., 2008). As the northern boundary of tree distributions is often temperature limited, northern migration of trees might seem almost inevitable. However, caution in predicting the rate and locations of treeline expansion is warranted, as rates of treeline expansion may lag in some areas due to seed availability, disturbance frequency, permafrost changes, and moisture constraints (Hofgaard and Harper, 2011; Walker et al., 2012; Henry et al., 2012).

White spruce trees along the northern Quebec–Labrador treelines show different responses according to their position relative to the sea. Along the coast, invading spruce are taking hold above the current treeline, while in the interior recent warming has not been strong enough to change the regressive tree line trajectory (Payette, 2007). Treelines in the forest-tundra areas of Quebec have risen slightly either through establishment of white spruce from seed or through growth of stunted spruce already established on tundra hilltops (Gamache and Payette, 2005; Caccianiga and Payette, 2006).

Treeline is advancing in other regions of the circumpolar Arctic that are increasing in productivity (Ims et al., 2013). Expansion of white spruce forests into areas previously occupied by tundra has been documented in numerous locations in Alaska (Suarez et al., 1999; Lloyd and Bunn, 2007), while recent studies have also shown that moisture is an important limiting factor in treeline advance (Ohse et al., 2012). Treeline advance into the tundra has also been documented in mountainous areas of Siberia (Kharuk et al., 2010; Kirdyanov et al., 2012) and northwestern Siberia, where forests have expanded into tundra at the rate of 3 to 10 m per year (Kharuk et al., 2006).

Vegetation is a key component in surface energy balance, regulating climate at regional and global scales. The observed and projected increases in the cover and density of shrub species and the increasing density of trees in forest tundra areas (Sturm et al., 2001) will alter the surface energy balance of northern terrestrial systems by decreasing the albedo and increasing absorption of solar radiation (Chapin III et al., 2005; Sturm et al., 2005). The increased absorption of solar radiation will increase the heating of the atmosphere. Chapin et al. (2005) estimated that a change from tundra to boreal forest will increase the heating of the atmosphere by approximately 4 W/m2, which is similar to the effect of doubling the CO2 concentration in the atmosphere. This will be an increasingly important positive feedback.

Top of Page

Species assemblages and plant communities

Changes in algal and invertebrate species assemblages in lakes and ponds

Striking and often unprecedented changes in the biota of Arctic lakes since the1850s have been linked to ecological shifts consistent with climate warming (Smol et al., 2005). Many algal and invertebrate communities of Arctic lakes and ponds have crossed ecological thresholds and a plausible explanation is recent warming (Smol and Douglas, 2007b). Other potential causes including ultraviolet radiation, nutrient enrichment, and atmospheric transport of pollutants cannot account for the scale, the nature, and the timing of the observed changes in algal communities. The best explanation is climatic warming leading to longer ice-free growing seasons and associated limnological changes including thermal stratification, habitat availability, and lakewater chemistry (Antoniades et al., 2005; Smol and Douglas, 2007b).

Analysis of sediments to identify and measure types of fossil diatoms, as well as other indicators, has detected major changes in species assemblages starting in the mid-19th century. Sediments deposited before this time show that diatom communities were relatively stable over centuries, and in some cases millennia, compared to the recent changes. Analysis of cores from 55 circumpolar Arctic lakes (Smol et al., 2005) showed that lakes have become more productive with more diverse algal and invertebrate communities. Lakes in the Canadian High Arctic experienced abrupt shifts in favour of diatoms that thrive in littoral habitats and mossy substrates, consistent with longer growing seasons and habitat changes under warming conditions. In deeper lakes on Baffin and Ellesmere islands and in lakes throughout the subarctic, the earlier onset and longer duration of thermal stratification becomes increasingly important, with pronounced increases in planktonic diatoms that fare better under conditions of less ice cover and greater thermal stratification, and concurrent decreases in benthic diatom taxa associated with cold conditions and extensive ice cover. In contrast, no major shifts in diatom assemblages were observed over the last two centuries in lakes in Labrador and northern Quebec, areas that have experienced little warming until very recently.

Local topographic and geologic factors also influence the response to warming (Smol et al., 2005). Sediments from lakes on Ellef Ringnes Island showed a transition to different, more diverse diatom communities starting about 1850 (Figure 63), typical of results reported from previous research (Smol et al., 2005). The same shift occurred in lake sediments near Alert, Ellesmere Island, but later, a difference that is attributed to varying sensitivity of the lakes to environmental change (Antoniades et al., 2005). For example, as reviewed by Smol and Douglas (2007b), lakes of different size would be expected to change at different time scales to climatic changes, with larger lakes typically responding more slowly than shallow ponds.

Figure 63. Diatom stratigraphy from a sediment core from a pond on Ellef Ringnes Island, showing the marked changes in species assemblages since 1850, especially in very recent years.

The core represents approximately 1300 years of sediment deposition. The three columns on the right are characteristics inferred from the diatoms (right to left: pH, specific conductivity, and analysis of change in diatom assemblage structure by detrended correspondence analysis).

Source: Antoniades et al. (2005)
Long description for Figure 63

This bar and line graph shows the diatom stratigraphy from a sediment core from a pond on Ellef Ringnes Island, showing the marked changes in assemblages of 13 species.  Prior to 1850, relative abundance of Pinnularia intermedia was high, at approximately 40%. Post 1850, Pinnularia intermedia became less dominant, and diatom communities became more diverse.

Top of Page

In Canada’s most northerly lake, Ward Hunt Lake, located at 83°N latitude and with perennial ice cover, diatoms appear only in the top 2.5 cm of sediment, corresponding to about the past 200 years of the 8,500 years of sediment records examined (Antoniades et al., 2007). This was accompanied by an increase in photosynthetic pigments (a measure of the standing crop of algae) of two orders of magnitude.

This pronounced shift in diatom assemblage composition is not a trend exclusive to the Canadian Arctic, or even just in northern latitudes. A synthesis of diatom-based paleolimnological studies of over 200 lake sediment records from around the Northern Hemisphere (Rühland et al., 2008) showed a general pattern of species-specific shifts in diatom assemblage composition since the mid-19th century that often followed millennia of relative stability in the diatom communities. Their analysis found that these changes occurred significantly earlier (by about 100 years) and often more dramatically in Arctic lakes than in lakes from temperate latitudes. They attributed this earlier and more rapid shift in Arctic lakes to the greater degree of warming that has occurred in the Arctic. The diatom species shifts observed at all latitudes were remarkably similar and occurred in conjunction with observed changes in freshwater habitat and structure that were linked to substantially warmer temperatures and changes in ice cover.

Remains of invertebrates in sediments show that the changes in lake conditions and algal communities have been transmitted to higher trophic levels. For example, aquatic insect (chironomid) populations in three Ellesmere Island ponds greatly expanded and diversified at the same time as diatom communities changed (Quinlan et al., 2005). Changes in both algae and invertebrate species assemblages are most pronounced in areas that have warmed the most, such as parts of the Canadian High Arctic (Smol et al., 2005).

Top of Page

Changes in tundra plant communities

Species diversity was found to decline in warming experiments across the tundra biome after up to six years of warming (Walker et al., 2006) and remained low after more than 10 years (Elmendorf et al., 2012a). The major factor in the decline in diversity was the loss of bryophyte and lichen species, which likely were shaded by the increasing cover and height of the vascular plants (Figure 47).

As species composition changes along gradients, and as local gradients in soil moisture and exposure can be relatively short spatially, tundra plant communities can be relatively distinct (Bliss and Matveyeva, 1992). The local topography imposes exposure and moisture gradients by influencing snow depth. At the regional scale, the effects of latitude and distance from the ocean are the major gradients affecting plant community structure (Walker et al., 2005).

Snow bed tundra plant species have been identified as potentially at risk if snowfall becomes less and melt is more rapid, leading to longer growing seasons with drier soils (Bjork and Molau, 2007). Increased precipitation during the summer may also affect the species composition of tundra plant communities, especially in the dry polar semi-desert areas. The trajectory of precipitation regimes in the Arctic, however, is difficult to predict (ACIA, 2005) and there have been very few experimental studies altering both winter and summer precipitation in combination with warming. An experimental study in High Arctic tundra plant communities showed that watering with about 13 L every 10 days for three seasons did not affect net primary production (Henry et al., 1986).

Tundra plant communities are showing changes across the biome that are consistent with responses to warming (see also the section on Primary production (biomass) in tundra ecosystems on page 75). In a study of 158 plant communities across 46 sites that were resampled between 1980 and 2010, Elmendorf et al. (2012b) found increases in canopy height and height of most vascular growth forms, increased abundance of shrubs and litter, and decreased cover of bare ground (Figure 64). However, the responses varied greatly among climate zones and depended strongly on moisture and permafrost conditions. Overall, the responses were very similar to a biome-wide analysis of warming experiments (Elmendorf et al., 2012a) and provide plot-level support to the changes in reflectance measured by NDVI (Bhatt et al., 2010).

Figure 64. Biome-wide changes in vegetation height and abundance, circumpolar sites.

Height (a) is the mean height of the canopy for each category; abundance (b) is the probability of increase. Error bars show ± SE and are bold when mean change rates were significant (P<0.05). Sample sizes (number of studies, number of locations) are on the x-axis.

Source: Elmendorf et al. (2012b)
Long description for Figure 64

This figure consists of two bar graphs which show the biome-wide changes in vegetation height and abundance at circumpolar sites.  The graphs shows increases in canopy height and height of most vascular growth forms, increased abundance of shrubs and litter, and decreased cover of bare ground.

Top of Page

Major human stressors on ecosystem structure

Climate change

As with the linked trends in ecosystem processes and functions, climate change is the main human-induced stressor on ecosystem structure in the Arctic Ecozone+.

Top of Page

Stressors related to industrial development, settlements, and human activity
Fragmentation and disturbance

Overall, the degree of human-induced fragmentation in the Arctic Ecozone+ is extremely low, as there are few roads and other linear features. Fragmentation, however, is a stressor of concern at the regional level, with the potential to become more widespread with increasing human population and industrial activity in the Arctic.

Disturbance from roads, pipelines, and other linear facilities, and from vehicle and aircraft traffic, has been widely hypothesized to affect wildlife. These hypotheses have been tested experimentally and by behavioural observations in many Arctic and subarctic settings (Slaney, 1975; Harding and Nagy, 1980; Gunn, 1984; James and Stuart-Smith, 2000; Gnieser, 2000; Dyer et al., 2002; Lunn et al., 2004). In general, the studies show modest impacts on animal energetics related to fleeing industrial activities, modest habitat losses from avoidance of areas with industrial facilities and activities, and little or no reluctance of migrating animals (for example, caribou) to cross linear facilities (Nellemann and Cameron, 1998; Wolfe et al., 2000b). In some cases, wild animals are attracted to such facilities. For example, caribou were attracted to the tailings pond and airstrip at Lupin Gold Mine, Contwyoto Lake (Gunn et al., 1998). It is, however, clear that some species, including caribou, grizzly bears, wolverines, and possibly muskoxen, do avoid areas of intensive activity and disturbance, at least for some life history requirements (Harding and Nagy, 1980; Dumond, 2006; Krebs et al., 2007).

A study in the Southern Arctic and Taiga Shield examined the cumulative impacts of human activities and associated infrastructure (mainly related to diamond mine development) on the distribution of caribou, wolves, grizzly bears, and wolverines over an area of 190,000 km2, 400 km northeast of Yellowknife, from 1995–2000 (Johnson et al., 2005). The study was conducted as part of the West Kitikmeot/Slave Study (see Case study on environmental governance: Kitikmeot on page 183). Mines and other major developments had the greatest effect on species occurrence, followed by exploration activities and outfitter camps, but there was not a uniform response, with some carnivores being attracted to these facilities. Modeling techniques were used to assess the reduction in habitat effectiveness (i.e., to what degree the animals selected poorer quality habitats in response to proximity to disturbances). Grizzly bears and wolves showed the strongest negative response, followed by caribou and wolverines. Seasonally, however, the biggest impact was on post-calving caribou: models suggested a 37% reduction in the area of the highest quality habitats and an 84% increase in the area of lowest quality habitats.

Effects of human infrastructure on predator habitat

Human infrastructure (roads and airstrips, often with culverts, telecommunications poles and towers, navigation beacons, and buildings) greatly enhance habitat structure for some species by providing new denning and nesting opportunities (Liebezeit et al., 2009; Reid et al., 2011; Wilson et al., 2013). These new structures generally benefit predators (foxes and raptors) and thereby change the spatial distribution of the foraging areas of these predators. This can result in changes to the reproductive success of various prey species (notably shorebirds, (Liebezeit et al., 2009)), and the nesting distribution of competing raptors (Reid et al., 2011). Mitigation measures include minimizing construction of such structures, designing structures to reduce raptor-nesting potential, and minimizing the extent of road and pipeline networks.

Effects of food supplementation on trophic interactions

Food supplementation results when garbage disposal is centralized, creating feeding opportunities in a few locations for scavengers such as foxes and gulls. The scavengers benefit, with potential spillover effects on other species. The expansion of the red fox into Arctic Canada coincided with centralized community development for Inuit in the 1930s through 1960s, and was likely related to this enhancement of food resources at certain limited sites (Marsh, 1938; Macpherson, 1964). Red fox outcompete Arctic fox, so the red fox expansion can result in loss of Arctic foxes from ecosystems. The two species co-exist on the Yukon North Slope, where there is currently no food supplementation (Gallant et al., 2012).

Case study: Cruise ship tourism, potential emerging stressor on ecosystem structure

Historically, ice conditions have prohibited most commercial shipping in the Canadian Arctic, and it was not until 1984 that cruise tourism got underway with the first cruise ship voyage through the Northwest Passage, by the MS Explorer (Jones, 1999). Increasing reduction of summer sea ice (see section on Sea ice on page 36) makes most of the Canadian Arctic increasingly accessible to cruise ship tourism (Stewart et al., 2007; Stewart et al., 2010).

From 1984 to 1991, activity was sporadic. From 1992 on, both numbers of cruises and diversity of cruise routes increased steadily. The number of planned cruises in Arctic Canada doubled between 2005 and 2006 (from 11 to 22) and continued to increase from 2007 to 2010 at a rate of approximately 10% per year (Dawson, 2012). Cruise ship activity has now begun to level off, likely due to the 2009 economic recession, limited number of appropriate ships available, and the fluctuating tourism market for Arctic cruises (Dawson, 2012). Lack of infrastructure (such as port facilities and ground transport), as well as few tourist sites and community activities for groups of visitors, also impose limitations on the growth of the cruise ship industry.

During the 2007 season, 23 separate cruises, run by six different companies, brought approximately 2,110 visitors to the Canadian Arctic (Stewart et al., 2010). Many of these ships made planned, and occasionally unplanned, stops in national parks such as Auyuittuq National Park and Sirmilik National Park on Baffin Island, and Quttinirpaaq National Park on Ellesmere Island, as well as to nearby locations (Table 6) (Stewart et al., 2008).

Table 6. Cruise ship visits to Arctic National Parks and nearby destinations
Pond Inlet1299
Tanquary Fjord111
Ellesmere Island111
Auyuittuq National Park432

Source: Stewart et al. (2008)

Top of Page

Most cruising expeditions encourage tourists to hike in the parks. Without regulated access and trails, large number of tourists hiking in parks and in other accessible coastal sites could have enduring negative impacts on the surrounding ecosystems. During investigations of abandoned settlements on three islands in the eastern Canadian Arctic Archipelago, Forbes (1996) found that flora of the Arctic has a limited number of species able to respond to disturbance and anthropogenically disturbed patches may be extremely persistent. Even relatively low-intensity, small-scale disturbances have immediate and persistent effects on Arctic vegetation and soils. Where slope is minimal, such disturbances are capable of expanding over large areas in as short a time as four years. The result is an artificial mosaic of patches of highly variable quality and quantity, compromising feeding and nesting habitats for vertebrate herbivores (Forbes, 2001). Other potential impacts include wildlife disturbance, especially of nesting seabirds (Marquez and Eagles, 2007), increased risk of introduction of invasive plants, and large local increases in human waste that, if not disposed of properly, could result in habitat degradation.

Top of Page

Ecosystem composition

Composition refers to the species and species groups that make up the Arctic Ecozone+.

Sources of knowledge about past distribution and abundance of Arctic species include Aboriginal Traditional Knowledge and records and narratives written by early explorers, who often recorded information from their Dene and Inuit guides along with their own observations and survey results. American and Canadian governments began sponsoring biological surveys in the late 19th century. These surveys brought to the attention of the world the iconic wildlife species of the Canadian Arctic: polar bears (Ursus maritimus), barren-ground grizzly bears (Ursus arctos), vast herds of migratory barren-ground caribou (Rangifer tarandus groenlandicus), small, nearly white Peary caribou (Rangifer tarandus pearyi) of the Northern Arctic, Arctic fox (Vulpes lagopus), Arctic hares (Lepus arcticus), lemmings (Arvicolinae) and muskox (Ovibos moschatus); marine mammals such ringed (Pusa hispida) and bearded (Erignathus barbatus) seals, bowhead (Balaena mysticetus) and beluga (Delphinapterus leucas) whales and narwhals (Monodon monoceros); walruses (Odobenus rosmarus); resident birds such as the snowy owl (Bubo scandiacus) and gyrfalcon (Falco rusticolis), and the Arctic breeding grounds of migratory birds such as the now presumably extinct Eskimo curlew (Numenius borealis).

Top of Page

Species diversity

Species diversity is relatively low in the Arctic Ecozone+, declining on a gradient from south to north and west to east. The northwestern region of the ecozone+ is part of the Beringia region and its diversity is enhanced by endemic or globally rare species that survived the last glaciations (Cannings et al., 2013).

Diversity of vascular plants in Canadian tundra ecosystems declines with decreasing temperature along the latitudinal gradient (Figure 65) (Rannie, 1986). While the number of bryophyte and lichen species also declines with latitude, the decrease in richness is not as great as for vascular species. Vertebrate species richness also declines from south to north, as shown in Figure 66.

Figure 65. Relationship between vascular plant species richness and mean July temperature in the Canadian Arctic.
Source: Rannie (1986)
Long description for Figure 65

This scatterplot graph shows the relationship between vascular plant species richness and mean July temperature in the Canadian Arctic. The graph shows that the relationship is linear and positive in Canadian tundra ecosystems, increasing from about 50 species at 30C to just over 200 species at 100C.

Top of Page

Figure 66. Terrestrial vertebrate species richness in Canada.
Source: GeoInsight Corporation (1999)
Long description for Figure 66

This map of Canada shows the spatial distribution of terrestrial vertebrate species richness.  In general, vertebrate species richness declines from south to north, with 1-34 species at the most northern point.. Species richness is highest (308+ species) in southern Alberta, Saskatchewan, and the southwest corner of Manitoba.

Top of Page

Arctic ecosystems have relatively simple food webs relative to more southern ecosystems (see section on Community and population dynamics on page 56). More complex ecosystems have multiple energy flow pathways and more species that occupy similar niches. The low species diversity and generally simple Arctic ecosystems may limit the ability of these ecosystems to resist perturbation in the first place, and to recover when damaged. However, the link between food web complexity and vulnerability to changes is not well-understood (see section on Community and population dynamics).

Nunavut, with a land area of 1.9 million km2, almost all of which is within the Arctic Ecozone+ , has 38 species of terrestrial mammals, 151 species of birds breeding in the territory, 20 species of freshwater and anadromous fishes, 8 amphibian and 1 reptile species, 47 butterfly species, and 626 species of vascular plants (Department of Environment, 2013a). By contrast, British Columbia, with half the land area, has 47 regularly occurring mammal species, 528 species of birds breeding in the province, 120 of freshwater and anadromous fishes, 22 amphibians and 22 reptile species, 275 of butterfly species, and 3,097 species of vascular plants (BC Conservation Data Centre, 2007). This comparison illustrates the relative simplicity, in terms of species richness, of Arctic ecosystems.

Top of Page

Trends in species of conservation concern

Species assessment

The biological status of species refers to their risk of extinction as determined through standard criteria and assessed at the global, national, or provincial/territorial level. Subspecies and populations can be assessed and included in the term “species”. Status at the global level is assessed by the International Union for the Conservation of Nature (IUCN). Species considered at risk in the world or of conservation concern for this overview are those assessed at a category of Near Threatened, Vulnerable, Endangered, or Critically Endangered.

The biological status of species at risk in Canada can be assessed as “Endangered” (facing imminent extirpation or extinction), “Threatened” (likely to become endangered if nothing is done to reverse threats), and “Special Concern” (at risk of becoming threatened or endangered). Species that are assessed by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) as being at risk, but are not yet legally protected under Canada’s Species at Risk Act (SARA), are undergoing consultation for listing (Table 7). Although the Arctic Ecozone+, when compared with other ecozones+, has the lowest number of species listed as at risk under SARA, several iconic or keystone species and species important to Inuit culture are under review.

Below are discussions of status and trends of selected species of conservation concern.

Table 7. Species with a Canadian distribution restricted to the Arctic Ecozone+ and/or Arctic marine ecozones+ that are listed under Canada's Species at Risk Act (SARA) as being of conservation concern or are assessed as at risk by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC).
Species with a Canadian distributionSpeciesSubspecies or PopulationCOSEWIC StatusSARA StatusSARA Schedule
PlantsHairy braya (Braya pilosa)Hairy braya (Braya pilosa)Endangered--
PlantsPorsild’s bryum (Haplodontium macrocarpum)Porsild’s bryum (Haplodontium macrocarpum)ThreatenedThreatenedSchedule 1
PlantsSpiked saxifrage (Micranthes spicata)Spiked saxifrage (Micranthes spicata)Threatened--
FishesAtlantic cod (Gadus morhua)Arctic Lakes populationSpecial ConcernNo StatusNo Schedule
FishesAtlantic wolffish (Anarhichas lupus)Atlantic wolffish (Anarhichas lupus)Special ConcernSpecial ConcernSchedule 1
FishesBering wolffish (Anarhichas orientalis)Bering wolffish (Anarhichas orientalis)Data DeficientSpecial ConcernSchedule 3
FishesBlackline prickleback (Acantholumpenus mackayi)Blackline prickleback (Acantholumpenus mackayi)Data DeficientSpecial ConcernSchedule 3
FishesBering cisco (Coregonus laurettae)Bering cisco (Coregonus laurettae)Special ConcernNo StatusNo Schedule
FishesBull trout (Salvelinus confluentus)Western Arctic populationsSpecial Concern  
FishesDolly varden (Salvelinus malma malma)Western Arctic populationSpecial ConcernNo StatusNo Schedule
FishesFourhorn sculpin (Myoxocephalus quadricornis)Freshwater formData DeficientSpecial ConcernSchedule 3
FishesNorthern wolffish (Anarhichas denticulatus)Northern wolffish (Anarhichas denticulatus)ThreatenedThreatenedSchedule 1
FishesRoundnose grenadier (Coryphaenoides rupestris)Roundnose grenadier (Coryphaenoides rupestris)EndangeredNo StatusNo Schedule
FishesSpotted wolffish (Anarhichas minor)Spotted wolffish (Anarhichas minor)ThreatenedThreatenedSchedule 1
FishesThorny skate (Amblyraja radiata)Thorny skate (Amblyraja radiata)Special ConcernNo StatusNo Schedule
BirdsBuff-breasted sandpiper (Tryngites subruficollis)Buff-breasted sandpiper (Tryngites subruficollis)Special ConcernNo StatusNo Schedule
BirdsEskimo curlew (Numenius borealis)Eskimo curlew (Numenius borealis)EndangeredEndangeredSchedule 1
BirdsHarlequin duck (Histrionicus histrionicus)Eastern populationSpecial ConcernSpecial ConcernSchedule 1
BirdsIvory gull (Pagophila eburnea)Ivory gull (Pagophila eburnea)EndangeredEndangeredSchedule 1
BirdsRed knot (Calidris canutus)rufa subspecies (C. canutus rufa)EndangeredEndangeredSchedule 1
BirdsRed knot (Calidris canutus)roselaari type (C. canutus roselaari type)ThreatenedThreatenedSchedule 1
BirdsRed knot (Calidris canutus)islandica subspecies (C. canutus islandica)Special ConcernSpecial ConcernSchedule 1
BirdsRoss’s gull (Rhodostethia rosea)Ross’s gull (Rhodostethia rosea)ThreatenedThreatenedSchedule 1
BirdsShort-eared owl (Asio flammeus)                             Short-eared owl (Asio flammeus)      Special ConcernSpecial ConcernSchedule 1
BirdsTundra peregrine falcon (Falco peregrines tundrius)Tundra peregrine falcon (Falco peregrines tundrius)Non-activeSpecial ConcernSchedule 3
Land MammalsBarren-ground caribou (Rangifer tarandus groenlandicus)Dolphin and Union populationSpecial ConcernSpecial ConcernSchedule 1
Land MammalsGrizzly bear (Ursus arctos)Western populationSpecial ConcernNo StatusNo Schedule
Land MammalsPeary caribou (Rangifer tarandus pearyi)Peary caribou (Rangifer tarandus pearyi)EndangeredEndangeredSchedule 1
Land MammalsPolar bear (Ursus maritimus)Polar bear (Ursus maritimus)Special ConcernSpecial ConcernSchedule 1
Land MammalsWolverine (Gulo gulo)Western populationSpecial ConcernNo StatusNo Schedule
Marine MammalsAtlantic walrus (Odobenus rosmarus rosmarus)Atlantic walrus (Odobenus rosmarus rosmarus)Special ConcernNo StatusNo Schedule
Marine MammalsBeluga whale (Delphinapterus leucas)Southeast Baffin Island-Cumberland Sound populationNon-activeEndangeredSchedule 2
Marine MammalsBeluga whale (Delphinapterus leucas)Eastern Hudson Bay populationEndangeredNo StatusNo Schedule
Marine MammalsBeluga whale (Delphinapterus leucas)Ungava Bay populationEndangeredNo StatusNo Schedule
Marine MammalsBeluga whale (Delphinapterus leucas)Cumberland Sound populationThreatenedNo StatusNo Schedule
Marine MammalsBeluga whale (Delphinapterus leucas)Eastern High Arctic–Baffin Bay populationSpecial ConcernNo StatusNo Schedule
Marine MammalsBeluga whale (Delphinapterus leucas)Western Hudson Bay populationSpecial ConcernNo StatusNo Schedule
Marine MammalsBowhead whale (Balaena mysticetus)Bering–Chukchi–Beaufort populationSpecial ConcernSpecial ConcernSchedule 1
Marine MammalsBowhead whale (Balaena mysticetus)Eastern Arctic populationNon-activeEndangeredSchedule 2
Marine MammalsBowhead whale (Balaena mysticetus)Eastern Canada–West Greenland populationSpecial ConcernNo StatusNo Schedule
Marine MammalsGrey whale (Eschrichtius robustus)Eastern North
Pacific population
Special ConcernSpecial ConcernSchedule 1
Marine MammalsHumpback whale (Megaptera novaengliae)Western North Atlantic populationNot at RiskSpecial ConcernSchedule 3
Marine MammalsKiller whale (Orcinus orca)Northwest Atlantic/ Eastern Arctic populationSpecial ConcernNo StatusNo Schedule
Marine MammalsNarwhal (Monodon monoceros)Narwhal (Monodon monoceros)Special ConcernNo StatusNo Schedule

Those species on SARA Schedules 2 and 3--or as yet on no schedule--are at some stage of assessment for consideration for listing on Schedule 1. Note that COSEWIC assessment and SARA listing are two separate processes. COSEWIC assessment as Endangered does not automatically mean SARA listing.

Source: SARA Registry, current to June 2013 (Government of Canada, 2013)

Top of Page

Peary caribou

This section is based on the ESTR technical thematic report Northern caribou population trends in Canada (Gunn et al., 2011c) and includes updates to that report and expanded discussion based on recent assessments and research results. See also the Main threats to caribou section (page 170). The classification of caribou used in this report follows the current Species at Risk Act (SARA) classification system. In 2011, COSEWIC adopted 12 designatable units for caribou in Canada that will be used in caribou assessments and subsequent listing decisions under SARA beginning in 2014 (COSEWIC, 2011).

Over the last 50 years the number of Peary caribou have declined from about 44,000 to about 11,000 to 12,000 caribou (Species at Risk Committee, 2012a) and two geographic populations appear to have disappeared. The rate of decline has varied over time and between the different geographic populations, with both reversals of some declines and absence of recovery for other populations. Survey intervals are irregular: only two of the six geographic populations (Banks and Bathurst Island complex) have been surveyed at regular intervals. Overall trends differ between the northern High Arctic islands and the larger, southern (mid-Arctic) islands and Boothia Peninsula. Although these larger mid-Arctic islands had a relatively high abundance of Peary caribou until the 1990s, declines occurred. Populations on the Boothia Peninsula and on Prince of Wales and Somerset islands had almost disappeared by the 1990s. There is no evidence for recovery over a 20-year period on Banks and northwest Victoria islands, despite severely restricted harvesting since the early 1990s (Species at Risk Committee, 2012a).

Geographic populations of Peary caribou are identified based on knowledge of movements, seasonal distribution, and genetics (Jenkins et al., 2011; Species at Risk Committee, 2012a). The Arctic islands are linked for much of the year by sea ice, and Peary caribou seasonally migrate across the ice. Sea ice coverage, however, is minimal in fall, which likely influences the rut distribution and thus the maintenance of island geographic populations. Jenkins et al. (2011) described seven geographic populations on the more northern islands: Melville and Prince Patrick, the Prime Minister group, the Bathurst Island group, Devon Island, Axel Heiberg, Ellesmere, and the Ellef group. Two of four geographic populations on the southern islands (Prince of Wales and Somerset; Boothia Peninsula) have essentially disappeared since the 1990s, leaving the two southern island populations of Banks and northwest Victoria islands.

On the High Arctic islands, weather is an overwhelming influence as periodic severe winters trigger large-scale mortality and reduction in productivity (Miller and Gunn, 2003; Harding, 2004). Although the signals of climate warming are strong in the High Arctic (Zhang et al., 2011), relating those trends in weather to changes in Peary caribou abundance is uncertain, partly because of high annual variability in climate and infrequent monitoring for most Peary caribou. The other reason is that harvest and predation also affect Peary caribou abundance. Muskox trends in abundance tend to differ from Peary caribou, although this is area-specific. Muskox increases relative to Peary caribou decreases have raised the question of competition. The role of intra- or inter-specific competition for forage is conjectural, as diet and habitat selection differ considerably between caribou and muskoxen (Gunn and Dragon, 2002). On Banks Island, however, there was overlap in some plants eaten by the two species, with, for example, willow being eaten by both Peary caribou and muskoxen (Larter and Nagy, 2004). This suggests that a competitive relationship could occur between the two species. Less emphasis has been placed on determining whether the increasing muskox abundance supported increased wolf numbers which, in turn, could increase predation rates on Peary caribou (Gunn and Dragon, 2002). Even less attention has been given to studying the relationship between caribou and muskoxen and their parasites. Hughes et al. (2009), however, discussed levels of intestinal nematode worms and warble flies in muskoxen and caribou for the range of the Dolphin and Union caribou.

There are limitations to describing the trends for Peary caribou due to the infrequency of surveys and the relatively brief period over which surveys have occurred. It is uncertain whether the documented high numbers of Peary caribou in the early 1960s followed by a decline and prolonged low numbers are: 1) part of regular fluctuations; 2) a period of relative stability within an unusually prolonged decline; or 3) atypically high peak numbers. Possibly Peary caribou regularly fluctuate in abundance, driven by a relationship between amounts of forage and caribou abundance (Tews et al., 2007a; Tews et al., 2007b). Alternatively, Peary caribou are in a “non-equilibrium grazing system” where sporadic, unpredictable weather events affect vital rates and population trends (Caughley and Gunn, 1993; Behinke, 2000).

Trends in abundance

Melville and Prince Patrick islands: The infrequent surveys documented steep declines between 1961 (the first range-wide aerial survey) and 1997. On Prince Patrick Island there was a 95% decline (from 1,797 to 84 year-plus old caribou). On Melville Island, Peary caribou declined by 92% (from 10,366 to 787 year-plus old caribou) (Tener, 1963; Gunn and Dragon, 2002). However, between 1997 and 2012, trends were reversed and a strong recovery was apparent for Melville, Prince Patrick, Byam Martin, Eglinton, and Emerald islands (Davison and Williams, 2012). The 2012 estimate was about 6,000 Peary caribou which, while not a complete recovery to the abundance recorded in 1961 (16,000), is the highest abundance recorded since then. See Figure 67 a and Figure 67 b. Lack of sea ice prevented the aerial survey of Mackenzie King, Brock, and Borden islands in 2012.

Prince of Wales and Somerset islands: Peary caribou seasonally crossed the sea ice between the islands in this group and some caribou also wintered on the Boothia Peninsula. Between 1974 and 1980, caribou numbers were stable in the range 4,000–6,000, one of the largest Peary caribou populations in the 1970s and 1980s (Gunn et al., 2006). There was a 15-year hiatus in surveys until 1995, when only a few caribou were found (Gunn and Dragon, 1998). In 2004, no caribou were seen during an aerial survey of the islands (McRae et al., 2010). See Figure 67 c.

Bathurst Island (and its satellite islands): Between 1961 and 1974, Peary caribou numbers declined by an order of magnitude (Miller, 1991a). Between 1974 and 1994, numbers recovered to the 1961 level (Miller, 1991a). An abrupt decline followed and, by 1997, fewer than 100 caribou remained (Gunn and Dragon, 2002). A survey in 2001 revealed the trend was for a recovering population (McRae et al., 2010). The Bathurst Island complex was resurveyed in May 2013, with preliminary estimates indicating that the herd has tripled in size since 2001 (Department of Environment, 2013a). The final population estimate based on this survey was not available at the time of report completion and is not included in the graph, Figure 67 d.

Banks Island: Peary caribou on Banks Island were one of the larger populations as they peaked at about 12,000 in the early 1970s (Gunn et al., 2000b) and remained relatively stable until 1982 (Nagy et al., 2009d). Numbers declined to about 1,000 caribou by 1992 (Nagy et al., 2009a) and an initial small recovery by 2001 was likely lost during an icing storm early winter 2003 (Nagy and Gunn, 2006). A 2010 survey led to an estimate of 1,097 ± 343 [95% confidence interval (CI)] non-calf caribou, which confirmed the persistence of low numbers on Banks Island (Davison et al., 2013). See Figure 67 e.

Northwest Victoria Island: Trends in Peary caribou on northwest Victoria Island are less clear than on Banks Island as surveys have been less frequent. Numbers were high, about 2,600 in 1987 (Gunn et al., 2000b), and declined during the late 1980s until, in 1993, only a few caribou were seen during an aerial survey (Gunn, 2005). The population then slowly recovered, based on estimates of 95 ± 60 (95% CI) in 1998 (Nagy et al., 2009b) and 204 ± 103 in 2001 (Nagy et al., 2009c). However, in 2005, the estimate was 66 ± 61 non-calf caribou, which suggested that some recovery was lost during two winters (2002/03 and 2003/04) with icing events (Nagy and Gunn, 2006). A subsequent survey in 2010 returned an estimate of 150 ± 104 non-calf caribou, which confirmed the persistence of low numbers (Davison et al., In Prep.). See Figure 67 f.

Boothia Peninsula: Peary caribou increased throughout the 1970s and early 1980s. A survey in 2006 in this region showed a decline from a 1985 estimate (Gunn and Dragon, 1998). Trends are difficult to distinguish as satellite telemetry has shown that both barren-ground caribou and Peary caribou calve and summer on the peninsula (Gunn et al., 2000a) and only three aerial surveys were flown between 1985 and 2006. During the most recent aerial survey in June 2006, only one caribou was identified as a Peary caribou and sightings of all caribou were few on the northern part of the peninsula which was used by Peary caribou for calving and summer in the mid-1980s and 1990s (Jenkins et al., 2011).

Eastern Queen Elizabeth Islands (Ellef Ringnes, Amund Ringnes, Devon, Ellesmere, Axel Heiberg Islands, Cornwall, King Christian, Graham): There is relatively little information available to assess trends as there have only been two extensive aerial surveys over a 50-year period. The islands were surveyed in 1961, although coverage was so low that the resulting figure of about 1,500 caribou was an approximation (Tener, 1963). Miller et al. (2005) re-analysed the 1961 data using updated areas for the islands and increased the estimate to 2,887± 642. The next extensive survey was between 2005 and 2008, when the Nunavut Department of Environment estimated 4,000 caribou 10 months and older based on aerial surveys (Jenkins et al., 2011):

  • Ellesmere Island (including Graham Island), surveyed partly in 2005 and partly in 2006: estimate of 1,021 caribou;
  • Axel Heiberg Islands, surveyed 2007: 2,291 (95% CI of 1,636 to 3,208) caribou
  • Amund Ringnes, Ellef Ringnes, King Christian, Cornwall, and Meighen Islands, surveyed 2007: total of 282 (95% CI of 157 to 505) caribou
  • Lougheed Island, surveyed 2007: 372 (95% CI of 205 to 672) caribou
  • Devon Island, surveyed 2008: 17 caribou counted in an extensive survey

The information does not support any overall trend. The reworked 1961 estimate and the 2005–2008 estimate overlap in their confidence limits, but, with only two data points separated by almost 50 years, there is not sufficient information to interpret the trend as stable as the extent and nature of local declines and recoveries are not known.

Figure 67. Peary Caribou population estimates.
Source: Gunn et al. (2011c); Species at Risk Committee (2012a); Jenkins et al. (2011). See also references in text.
Long description for Figure 67

This series of bar graphs show the following information:

Peary caribou population estimates. (Population Estimates)
YearMelville IslandPrince Patrick IslandPrince of Wales, Russell, and Somerset IslandsBathurst Island ComplexBanks IslandNorthwest Victoria Island
196112,7992,254-3,509Note a of Figure 67--
19722,551---12,098Note a of Figure 67-
19741,6796214,540Note a of Figure 67266--
1975--3,741Note a of Figure 67361--
1980--5,097Note a of Figure 67--4512
1993---2,787-*Note b of Figure 67
1994---3,155709*Note b of Figure 67
1995--*Note b of Figure 67---
1996---452Note a of Figure 67--
1997787Note a of Figure 6784Note a of Figure 67-74Note a of Figure 67--
1998--- 43695
1999--- --
2000--- --
2001---187Note a of Figure 671,142204

Notes of Figure 67

Note A of Figure 67

Bold values indicate population estimates not including calves.

Return to note a referrer of figure 67

Note * of Figure 67

Indicate surveys where only a few caribou were spotted.

Return to note b referrer of figure 67

Top of Page

Trends in distribution

Peary caribou only occur in Canada (except occasional sightings on the northwest Greenland coast) and are restricted to the High Arctic (Queen Elizabeth Islands (404 730 km2)) and the mid-Arctic islands (129 510 km2), as well as the northern extension of the mainland (Boothia Peninsula (26 000 km2)) (Miller, 1991b).

Figure 68. Global distribution of Peary caribou.
Source: adapted from Gunn et al. (2012), based on Miller (2013) and Miller and Barry (2002)
Long description for Figure 68

This map shows the global distribution of Peary caribou.  Peary caribou only occur in Canada (except occasional sightings on the northwest Greenland coast) and are restricted to the High Arctic (Queen Elizabeth Islands and the mid-Arctic islands), as well as the northern extension of the mainland Arctic (Boothia Peninsula).

Top of Page

Much of the eastern Queen Elizabeth Islands is mountainous with ice fields and glaciers. Areas above 750 m above ground level, including permanent snow and ice fields, account for 20–43% of Ellesmere, Axel Heiberg, and Devon islands. Caribou densities are much lower and so, although the eastern Queen Elizabeth Islands are 78% of High Arctic landmass, in 1961 they held only 10% of the Peary caribou (Miller et al., 2005). With the higher rate of decline on the western islands, the proportionate distribution has changed and, based on the most recent surveys, about 40% of High Arctic Peary caribou are on the eastern Queen Elizabeth Islands.

Knowledge of trends in distribution is largely based on sightings during systematic aerial surveys, which vary in timing from island to island. On a few islands (for example, northwest Victoria Island and Banks Island), tracking satellite-collared caribou has increased understanding of seasonal and annual distribution (Gunn and Fournier, 2000b; Poole et al., 2010; Gunn et al., 2012). Caribou shift their distribution during unfavourable winters, as, for example, has been recorded on Bathurst Island (F.L. Miller, unpublished data in Gunn et al., 2012) (Figure 69).

Figure 69. Home ranges of satellite-collared Peary caribou relative to favourable and unfavourable winters, Bathurst Island Complex, NU, 1993/94 and 1994 to 1997.

Three cow caribou were tracked for two years with favourable winters and two years with unfavourable winters. Each polygon encloses all tracked locations for one caribou for one year (the minimum home range for the individual for that year). The home ranges for the three females tracked were 5.5 times smaller during the favourable years than during the years with unusually deep snow.

Source: adapted from Gunn et al. (2012), based on Miller (2013) and Miller and Barry (2002)
Long description for Figure 69

This map shows the home ranges of satellite-collared Peary caribou relative to favourable and unfavourable winters on the Bathurst Island Complex in Nunavut for the years 1993 and 1994, and 1994 to 1997. The home ranges for the three females tracked were 5.5 times smaller during the favourable years than during the years with unusually deep snow.

Top of Page

As abundance declines, distribution becomes more restricted with, for example, changes in the use of the smaller islands. In the western Queen Elizabeth Islands in 1997, Peary caribou were not seen on three islands (Brock, Eglinton, and Emerald islands) during aerial surveys (Gunn and Dragon, 2002). However, Peary caribou were consistently seen on those islands during 1961, 1972–1974 and 1987–1988 aerial surveys, and were seen there again in 2012 (Davison and Williams, 2012). Any trends in distribution within the larger islands are difficult to determine due to the variability in timing of surveys (Species at Risk Committee, 2012a).

Gaps in information due to the sporadic surveys, as well as the scale of annual variation, impede describing trends in distribution. This is especially true for the eastern Queen Elizabeth Islands, as the frequency of surveys is so low (Jenkins et al., 2011). Overall, the distribution of Peary caribou was reduced by 15% between 1980 and 2006, taking into account the decline of two populations to possibly only a few individuals (on Prince of Wales and Somerset islands and on the Boothia Peninsula).

Top of Page

Dolphin and Union Caribou Population

This section is based partly on Northern caribou population trends in Canada (Gunn et al., 2011c) and includes updates to trend data presented in that report.

Status and trends

Historical information and Inuit hunter reports indicate that there may have been as many as 100,000 caribou on Victoria Island in the early 1800s (Manning, 1960). By the early 1920s, numbers declined and migration across Dolphin and Union Strait halted. The causes are possibly a combination of icing storms and the introduction of rifles. The recovery was slow and caribou were rare until the 1970s. By the 1990s, numbers were increasing (Figure 70 ). In October 1997 and 2007, surveys of caribou staging along the south coast of Victoria Island led to estimates of 27,948 ± 3,367 [standard error (SE)] and 21,753 ± 2,343 (SE) caribou, respectively (uncorrected to account for caribou assumed to be outside of the census zone) (Nishi and Gunn, 2004; Dumond, 2011, pers. comm.). Those two estimates, together with variable annual pregnancy rates, relatively low cow survival (over the period 1999-2006), and high harvest rates, suggest an increased likelihood of a decline (Poole et al., 2010; Dumond, 2011, pers. comm.).

Figure 70. Dolphin and Union Caribou population estimates.
Source: Gunn et al. (2011c). See also references in text.
Long description for Figure 70

This bar graph shows the following information:

Dolphin and Union Caribou Herd population estimates. (Value from summer visual survey)
YearPopulation estimates

Top of Page

Calving is dispersed over about half of northern and central Victoria Island and, based on sightings and satellite telemetry, the summer, fall, and winter ranges have increased in size since the early 1980s into the late 1990s. A trend to an increasing size of winter range is due to fall migrations across the newly formed sea ice to the mainland coastal areas, which is a resumption of migrations observed up until the 1920s. The caribou return across the sea ice to Victoria Island in April to May. The date of freeze-up is increasingly delayed: 8 to 10 days later than in 1982, a trend which may lead to changes in the fall migration across the sea ice (Poole et al., 2010).

Top of Page

Polar bear

Approximately 55 to 65% of the world’s 20,000 to 25,000 polar bears reside in Canada (COSEWIC, 2008). Figure 71 shows the global distribution and the status and trends of all polar bear subpopulations.

Figure 71. Circumpolar distribution of the polar bear, showing status and trends of subpopulations
Source: Environment Canada, 2013
Long description for Figure 71

This map shows the global distribution, population estimate, and the status and trends of all 19 polar bear management units in the circumpolar Arctic.  Population sizes and trends vary around the Arctic with some populations increasing, others decreasing, and some with insufficient data to determine trends. Data deficient units include Laptev Sea, Kara Sea, Barents Sea, East Greenland, Arctic Basin and Norwegian Bay. Units showing population declines are Western Hudson Bay, Baffin Bay, Kane Basin, and Chukchi Sea. The Southern Beaufort Sea unit shows a likely decline. Northern Beaufort Sea, Viscount Melville Sound, Lancaster Sound, Gulf of Boothia, Foxe Basin, and Southern Hudson Bay units are stable. M’Clinktock Channel unit shows a likely increase, and Davis Strait unit shows an increase.

Top of Page

Analysis of movement data from mark-recapture studies and tracking of adult female bears with satellite radio collars indicated that there are two subpopulations of polar bears in the Beaufort Sea: one that inhabits the west coast of Banks Island and Amundsen Gulf and a second that is resident along the mainland coast from about Baillie Islands in Canada to approximately Icy Cape in Alaska (Stirling, 2002). The central and Eastern Arctic has six population units: Viscount Melville Sound, Lancaster Sound, Norwegian Bay, Kane Basin, Baffin Bay and Davis Strait (Taylor et al., 2001). The Hudson Bay complex has three subpopulations: Western Hudson Bay, Foxe Basin, and Southern Hudson Bay. Although these subpopulations are sufficiently distinct for management purposes, there is enough movement among them for adequate gene flow (Obbard et al., 2010). Abundance estimate data for Canadian polar bear subpopulations are presented in Table 8.

Table 8. Polar bear abundance estimates for Canadian subpopulations
SubpopulationAbundance estimate95% confidence
Survey year
Baffin Bay2,0741542–26061997
Davis Strait2,1581798–25182007
Foxe Basin2,5802093–31802009/10
Gulf of Boothia1,592870–23142000
Kane Basin16494–2341997
Lancaster Sound2,5411759–33231998
M'Clintock Channel284166–4022000
Northern Beaufort Sea980670–12902006
Norwegian Bay203115–2911998
Southern Beaufort Sea1,5261210–18422006
Southern Hudson Bay969688–13652011/12
Viscount Melville16193–2291992
Western Hudson Bay1,000715–13982011

Source: Polar Bear Technical Committee (2013)

Top of Page

In addition to the science-based information, traditional and community knowledge forms part of decision-making about polar bear risk assessment and management. Aboriginal traditional knowledge (ATK) is considered alongside western science by wildlife management boards and government management agencies, especially in making decisions about harvest. Data brought to the management table from these different knowledge sources, reflecting different timeframes and scales as well as differences in experience, methodologies, and perspectives, are not always in agreement, though they are often complementary (e.g., see discussion in Peacock et al., 2011). As well, increasingly, shifts in distribution, timing, and habitat use make trends in overall abundance difficult to detect. Presentation of alternative information and perspectives on polar bear status and trends is beyond the scope of this primarily science-based report, but is covered elsewhere. For example, the recent Northwest Territories polar bear status assessment presents and synthesizes information from traditional and community knowledge and science studies (Species at Risk Committee, 2012b). Under the federal Species at Risk Act listing process, Environment Canada is obligated to consider ATK during assessments. Initiatives are underway to improve standardization of collection of ATK and to better integrate ATK and western science in assessments and management decisions.

Climate change and polar bears

Polar bears, adapted to hunting seals from the ice, cannot persist without seasonal sea ice (COSEWIC, 2008) and rapidly declining sea ice poses the most serious threat to polar bears (see section on Sea ice on page 36) (Peacock et al., 2011; Vongraven and Richardson, 2011; Reid et al., 2013). Polar bear subpopulations can be grouped according to ecoregions based on ice habitat (Vongraven et al., 2012). These ecoregions reflect differing types and levels of threat to polar bears from changes in ice timing and extent (Table 9). The divergent sea ice ecoregion has extensive annual ice that forms and then moves to the deep-water regions of the Arctic Ocean, while the convergent sea ice ecoregion has heavy multiyear ice. The convergent ice part of the range of the Norwegian Bay polar bear subpopulation is likely to remain the most viable for polar bear populations as seasonal ice shrinks. The Archipelago ecoregion has historically had a mixture of multiyear and seasonal ice filling the gaps between islands year-round, and polar bears in this ecoregion remain on ice all year. By contrast, in the seasonal ice ecoregion, bears are forced ashore and are deprived of food for periods when the ice melts in the summer.

Table 9. Relative imminent risk to polar bear subpopulations from climate change effects on ice habitat and known levels of toxic contaminants
EcoregionSubpopulationCurrent or imminent risk from climate changeKnown levels of toxic contaminants
DivergentSouthern Beaufort SeaHighLow
ConvergentNorthern Beaufort SeaMediumMedium
ArchipelagoGulf of BoothiaLowLow
ArchipelagoKane Basin?Low
ArchipelagoLancaster SoundMediumLow
ArchipelagoM’Clintock ChannelLowLow
ArchipelagoNorwegian BayLowLow
ArchipelagoViscount Melville SoundLow?
SeasonalBaffin BayHighLow
SeasonalDavis StraitHighLow
SeasonalFoxe BasinMediumLow
SeasonalSouthern Hudson BayHighLow
SeasonalWestern Hudson BayHighLow

Source: Vongraven et al. (2012)

Top of Page

Earlier sea-ice break-up around western Hudson Bay has led to poorer physical condition and poorer reproductive performance of polar bears (see next section on Declines in polar bear body condition). Sea-ice break-up has been linked with lower birth rates, lower survival in subadults and senescent bears, and lower body condition (Regehr et al., 2007). Satellite tracking has shown that polar bear movements have changed and productivity has d eclined in response to sea ice changes in both the Western and the Eastern Arctic (Stirling, 2002; Stirling et al., 2004), as has that of seals, their principal prey (Ferguson et al., 2005). Polar bear denning off Alaska has shifted landward and eastward in response to changing ice conditions (Fischbach et al., 2007). These and other observations led Derocher et al. (2004) to suggest that polar bears will likely not survive as a species, should the predicted scenarios for total disappearance of summer sea ice in the Arctic come true. A small area of the High Arctic may remain as a refuge for polar bears as their habitat shrinks (Durner et al., 2009; Vongraven et al., 2012; Eamer et al., 2013).

Declines in polar bear body condition

This section is extracted from the CAFF report Life linked to ice (Eamer et al., 2013), with an update for the Western Hudson Bay subpopulation as noted.

In areas where sea ice melts completely in the summer polar bears may be forced onto land. Earlier sea ice break-up in these areas reduces the amount of time bears have for hunting seals on the ice. In some areas where this is occurring, bears are becoming thinner, resulting in decreases in survival and reproduction.

Changes in sea ice over the past two decades have led to significant declines in physical condition of bears in the Western Hudson Bay (Stirling et al., 1999; Stirling and Parkinson, 2006), Southern Hudson Bay (Obbard et al., 2006), and Baffin Bay subpopulations (Rode et al., 2012). Regehr et al. (2007) showed that survival decreased in association with earlier sea ice break-up and that this contributed to a 22% decline in the size of the Western Hudson Bay polar bear subpopulation between 1987 and 2004. A recent aerial survey indicates that, while the population continues to decline, the rate of decline is less than anticipated based on the models used by Regehr et al. (2007). Mark-recapture studies are underway to obtain a new population estimate for the Western Hudson Bay subpopulation (R. Vallender, Environment Canada, pers. comm., 2013). Reduced survivorship in relation to sea ice conditions has also been demonstrated in the Southern Beaufort Sea polar bear population (Regehr et al., 2010).

For the western Hudson Bay population, the body condition of bears measured during the ice-free period declined from 1980 to 2007, as did the average weight of female polar bears in the fall (Figure 72). The female bears weighed were suspected to be pregnant.

For the Baffin Bay population, the decline in body condition since the early 1990s is associated with deteriorating ice conditions (Rode et al., 2012). Polar bears were in significantly worse condition in years with less summer sea ice cover, starting in the 1990s when ice in these regions began its sharp decline (Table 10).

Figure 72. Declining fall weights of female polar bears, western Hudson Bay, 1980 to 2007.

Body weights were estimated for females on their own, and thus likely to be pregnant.

Source: Eamer et al. (2013), based on Stirling and Derocher (2012)
Long description for Figure 72

This scatterplot graph shows the average weight of female polar bears in the fall for the years 1980 to 2007. Mean body mass declined during this period from just under 300 kg to about 230 kg.

Top of Page

Table 10. A) Trends in body condition for the Baffin Bay polar bear subpopulation and relationship with sea ice conditions, 1977 to 2010. (Overall trend in body condition)
Sampling PeriodMalesFemalesCubs
1978-1995 spring capturesno trenddeclineimprovement
1992-2010 spring capturesdeclinedeclineinsufficient data
1991-2006 fall capturesdeclinedeclinedecline
Table 10. B) Trends in body condition for the Baffin Bay polar bear subpopulation and relationship with sea ice conditions, 1977 to 2010. (Relationship between sea ice and body condition)
Sampling PeriodMalesFemalesCubs
1978-1995 spring capturesno trendno trendno trend
1992-2010 spring capturesbody condition better in years with more sea icebody condition better in years with more sea iceinsufficient data
1991-2006 fall capturesbody condition better in years with more sea icebody condition better in years with more sea icebody condition better in years with more sea ice

Top of Page

Other threats: harvest and contaminants

Polar bears are also very sensitive to overharvest (COSEWIC, 2008) and currently regulated harvest is the main factor limiting the size of some subpopulations (Species at Risk Public Registry, 2012). Other threats include disturbance and loss of habitat due to increased activity and opening of new areas for shipping, oil and gas development, and mining. Polar bears, as long-lived animals at the top of marine food chains, are vulnerable to the accumulation of contaminants, including mercury and persistent organic pollutants (Table 9). Mercury in polar bears can exceed threshold values for toxicological effects and there is evidence that levels are increasing (Dietz et al., 2013). In contrast, levels of some, but not all, organic contaminants in polar bear tissues have declined in recent years (Mckinney et al., 2011). Levels of contaminants, globally, in polar bears, are highest in the East Greenland subpopulation (Vongraven et al., 2012).

Top of Page

Grizzly bear

Grizzly bears have been extirpated across much of their former North American range, losing habitat in the southern areas and expanding in parts of the Arctic (Figure 73). Grizzly bears are a useful indicator species for revealing landscape changes in ecosystems. Body size, vital rates, and food habits correlate with habitat and seasonality of food supplies, which led Ferguson and McLoughlin (2000) to distinguish three groups of grizzlies: Pacific-coastal, interior, and barrenground. The barrenground bears typically require large home ranges: those of adult males are 1,154 to 8,171 km2 (Nagy et al., 1983; Clarkson and Liepins, 1994; McLoughlin et al., 1999), four times larger than the largest home ranges of 77 to 1,918 km2 for grizzly bears in other areas of Canada (Ross, 2002). In a study of 41 collared grizzly bears on the coastal plain and mountainous terrain of the Yukon North Slope, ranges of male grizzlies averaged 1,020 km2 with the most extensive range being over 3,000 km2 (Wildlife Management Advisory Council (North Slope), 2008b).

Figure 73. Approximate boundaries of current and historic (19th century) distribution of the grizzly bear in North America, showing current densities.

Worldwide, grizzly bear range has decreased by 50% since the mid-19th century

Source: COSEWIC (2012) and references therein
Long description for Figure 73

This map of North America shows the current densities and approximate boundaries of current and historic (19th century) distributions of the grizzly bear. Grizzly bears have been extirpated across much of their former North American range, losing habitat in the southern areas and expanding into parts of the Arctic.

Top of Page

The most recent estimates are 3,500 to 4,000 grizzly bears in the Northwest Territories and between 1,500 and 2,000 in Nunavut, which is almost a quarter of the 26,000 grizzly bears estimated for all of Canada (COSEWIC, 2012). Trends in abundance are difficult to gauge in the Arctic, partly because there is little historic information and survey methods and jurisdictional boundaries have changed, and only few areas have been repeatably sampled over time. Estimating population sizes of grizzly bears is inherently difficult and costly, due to the low densities and difficulty in spotting bears. Information on status and trends will become more available through regional monitoring studies using mark-recapture techniques, including those based on DNA identification of hair that has been sampled by snagging on barbed wire. Studies using this technique are underway in the central barrens as a result of a collaborative effort between the mining industry and the Government of the Northwest Territories (for example NWT Department of Environment and Natural Resources, 2011) and on the Yukon North Slope (Wildlife Management Advisory Council (North Slope), 2008b). A pilot project was carried out using hair snagging around Kugluktuk, Nunavut, from 2004 to 2006, followed by a larger study of grizzly bear populations in the West Kitikmeot area (Department of Environment, 2013a). Results from these studies were not available at the time of writing. Barrenground grizzly bear density estimates from earlier work are available for several areas of the Southern Arctic. The densities vary from 3.5 bears/1,000 km2 on the central barrenlands, in 1995–1999 (McLoughlin et al., 2003) to higher densities in the western NWT: the Anderson-Horton River area, for example, had 9.1 bears/1,000 km2 in 1994 (Clarkson and Liepins, 1994).

COSEWIC (2012) describes a trend of barren-ground grizzly bears expanding their range and becoming more common in the low to mid-Arctic tundra regions of northwest Canada. McLoughlin et al. (2003) reported the grizzly bear population in Nunavut was increasing in 1999 at an annual rate of 3%. There are increasingly reports of grizzly bears in areas where where they were seldom seen or not observed until recently on the Arctic islands, including Banks, Victoria, King William and north to Melville Island (COSEWIC, 2012).

The trend toward expanding their northern distribution has had unexpected consequences. In April 2006 a sport hunter shot a hybrid polar-grizzly bear on Banks Island (Roach, 2006). His guide, Roger Kuptana of Sachs Harbor, identified it as a hybrid; DNA tests showed it had a polar bear mother and a grizzly bear father. Although polar-grizzly bear hybrids are known from zoos and are fertile (Stirling, 1999), this was the first confirmation of a wild hybrid.

Habitat trends include a large and widespread decline in caribou, which make up a large part of the diet of grizzly bears in the central barrens (Mowat and Heard, 2006). In a study in Nunavut and the NWT (Gau et al., 2002), caribou (both calves and adults) was the most prevalent item in the diet, although in early summer, when caribou were elsewhere, horsetails, sedges, and cotton grass dominated the diet, with berries also being prevalent in late summer. Some bears accompany the pre-calving migration of the caribou (Gau et al., 2002) to take advantage of the high densities of caribou on calving grounds. While caribou abundance has changed, most herds return to the same calving grounds each year, providing a predictable resource for the bears. Although the role of muskoxen in the diet of barren-ground grizzly bears has not been studied, grizzly bear predation is known to occur on muskoxen, especially on calves (Case and Stevenson, 1991; Gunn and Fournier, 2000a).

Barren-ground grizzly bears are vulnerable to the effects of industrial activities through possible displacement, but also through nuisance bear kills, and, possibly, through disturbance during denning (Harding and Nagy, 1980; Nagy et al., 1983; Clarkson and Liepins, 1994; McLoughlin et al., 1999; Edwards, 2006). Barren-ground grizzly bears have low productivity, probably as a consequence of the relatively short foraging season. In the Low Arctic tundra, average duration of den occupancy is 185 days (6.2 months) for males and 199 days (6.6 months) for females (McLoughlin et al., 2002). In the Western Arctic there may be an emerging trend toward bears entering into hibernation later in the year and emerging from hibernation earlier, based on observations of Inuvialuit hunters. This change is perhaps due to the longer growing seasons experienced in the Arctic in recent years (Wildlife Management Advisory Council (North Slope) and Aklavik Hunters and Trappers Committee, 2008; COSEWIC, 2012).

Top of Page


Wolverine abundance on the Arctic tundra is an indicator of both the availability of caribou for human harvest and the level of human activities. Wolverines are the largest and mostly solitary carnivore active in the winter on the tundra, as bears hibernate in winter and wolves mainly hunt in packs. Reproductive rates are relatively low and strongly influenced by food availability in winter. Typically, wolverine breed at two to three years of age and have litters averaging three kits (COSEWIC, 2003).

Wolverines scavenge carcasses from wolf and grizzly bear kills. In winter on the tundra, wolverines mainly eat caribou, most likely scavenged, as well as ptarmigan (Lagopus spp.) and small mammals. Wolverines also rely on food caches. In summer, their diet includes a wide range of small mammals, birds including geese and goose eggs, as well as berries (Mulders, 2000). Wolverines’ dependence on caribou when in the tundra zone during the winter may be an important factor in habitat selection, as well explaining their large home ranges.

When caribou numbers decline, their fall and winter ranges contract and they spend more time on tundra rather than in the boreal forests. When caribou abundance is reduced, therefore, tundra-living wolverines likely still have access, at least initially, to caribou. However, if caribou abundance continues to decline, especially for herds that winter on the tundra (for example, the Dolphin and Union herd), wolverine abundance may also change. On the Bathurst caribou herd’s summer-fall range, where caribou abundance declined 93% between 1986 and 2009 (Boulanger et al., 2011a), wolverine densities have also declined. For example, at Daring Lake, wolverine declined from an average density of 8 wolverine per 1,000 km2 to 4 wolverine per 1,000 km2 over a period of 7 years (2005–2011) (Boulanger et al., 2011b; Boulanger and Mulders, 2013).

Wolverine, with low reproductive potential, large ranges, and relatively low winter survival, are susceptible to reduction in numbers from trapping. Sustainable harvest relies on there being sufficient productive females surviving to recolonize areas following removal of animals through trapping (Slough, 2013). Monitoring, both of the harvest and of the distribution and density of wolverines, is key to ensuring that refugia are sufficient to sustain harvested populations, especially when other impacts, such as habitat loss due to resource extraction development, or decline of caribou populations, are present.

Wolverine population studies and programs to collect information from trappers are ongoing in the NWT (Slough, 2013), the Yukon North Slope (Wildllife Management Advisory Council (North Slope), 2012), and Nunavut (Department of Environment, 2013a). For example, in Nunavut, a harvest monitoring program, initiated in Kitikmeot in the 1980s, was expanded to other regions in 2009. The study covers geographic distribution, age and sex of the harvested animals as well as feeding habits. A population monitoring study that involves local hunters in the collection of snagged hair samples to identify individual wolverine using DNA is planned for the Baker Lake area. This will establish baseline information: “natural” wolverine density in areas with limited or no harvest pressure (Department of Environment, 2013a).

Trends in abundance

Wolverine densities are at moderate levels on the western tundra in the NWT and Nunavut, and low on the Arctic islands and in eastern Nunavut: at the time of the 2003 COSEWIC assessment, the population estimates were 3,500 to 4,000 wolverine for the Northwest Territories and 2,000 to 2,500 for Nunavut (COSEWIC, 2003). The trends were considered stable for the NWT and Nunavut, but sensitive to changes in harvesting, as wolverine fur is highly regarded (COSEWIC, 2003). The estimate for numbers of adult wolverine resident in the Northwest Territories has since been revised to 5,100 (Slough, 2013).

Wolverine in northern Quebec and Labrador have been reduced to unconfirmed sightings since 1978 and the 1950s, respectively (COSEWIC, 2003). The declines are attributed to trapping and hunting and the extreme reduction in the caribou herds early in the 20th century, as well as human encroachment, reduced wolf numbers and poison baiting (COSEWIC, 2003). Despite the increase in caribou populations up until the 1990s and restrictions on hunting and trapping of wolverine, populations of wolverine in these locations have not recovered (COSEWIC, 2003).

Further information on wolverine, based both on science studies and ATK, will soon be available. A new COSEWIC assessment of wolverine is due to be released early in 2014. The NWT’s Species at Risk Committee is conducting an assessment of the western population, also to be released in 2014.

Top of Page

Trends in species of special interest

Species and groups of species of special interest for ecological, cultural, and economic reasons are discussed in this section. See the ESTR report Ecosystem status and trends report: Arctic marine ecozones (Niemi et al., 2010) for discussion of anadromous fish species of special interest (dolly varden and Arctic charr).

Top of Page

Migratory tundra caribou

This section is mainly excerpted and summarized from the ESTR national thematic report Northern caribou population trends in Canada (Gunn et al., 2011c) and includes updates as noted. Caribou groupings are based on that report while acknowledging that reviews of historical data and new information provide varying interpretations of geographical units relevent to caribou. Note that COSEWIC has published a report defining “designatable units” for caribou that is now in effect for the COSEWIC assessment process (COSEWIC, 2011).

Caribou, with their central role in tundra and taiga ecology and their inter-connection with the culture of many Aboriginal people, have parallels with the role of salmon on Canada’s Pacific West Coast. Ecosystem aspects of caribou are discussed in previous sections and the importance of caribou to Aboriginal people is discussed in the section on Ecosystem goods and services (page 167).

Northern caribou include migratory caribou of three subspecies plus Peary caribou (Banfield, 1961; Rothfels and Russell, 2005). The three subspecies included in migratory tundra caribou are: 1) barren-ground caribou (Rangifer tarandus groenlandicus), ranging east of the Mackenzie River; 2) Grant’s caribou (R. t. granti), ranging west of the Mackenzie River,; and 3) the two large herds in Ungava of woodland caribou plus two smaller herds in the Hudson Plains Ecozone+ (R. t. caribou).Peary caribou (R. t. pearyi) range on the High and southern (mid-Arctic) Arctic islands and the Boothia Peninsula (Figure 74 and Figure 75).

Figure 74. Northern caribou subspecies and groupings.
Source: Gunn et al. (2011c)
Long description for Figure 74

This flow chart shows the subspecies included in Northern caribou. Northern caribou include Peary caribou (R. t. pearyi) plus three subspecies of migratory caribou. The three subspecies included in migratory tundra caribou are: 1) barren-ground caribou (R. t. groenlandicus); 2) Grant’s caribou (R. t. granti); and 3) woodland caribou (R. t. caribou).

Top of Page

Figure 75. Current distribution and status designations of Rangifer in North America.

The map depicts COSEWIC status designation. See Environment Canada (2011) for an update of the current distribution of boreal caribou.

Source: adapted from Hummel and Ray (2008). Map reprinted with permission from Dundurn Press Ltd. © 2008
Long description for Figure 75

This map shows the current distribution and status designations of Rangifer in North America as well as the historical extent of caribou distribution.  The Gaspésie, Peary, and Selkirk populations are shown as endangered and the Queen Charlotte Islands population is extinct. The Boreal populations are shown as threatened, and the Dolphin Union and Northern Mountain populations as special concern. Alaska, Migratory Tundra, and Newfoundland populations are shown as either not at risk/not assessed.

Top of Page

Trends in abundance

Aboriginal elders recall periods of abundance and scarcity. Other indicators of past caribou abundance and distribution include traditional place names (Legat et al., 2002). Highs and lows in historic abundance since the 1800s have been reconstructed from the frequency of hoof scars on spruce roots, at least for the Bathurst and George River herds (Payette et al., 2004a; Zalatan et al., 2006). Current ranges and recent trends are presented in Figure 76 , based on information summarized in Gunn et al. (2011c). Note that the Beverly and Ahiak herds are shown here as they have been historically defined, consistent with Gunn et al. (2011c). Other authors using re-evaluation of satellite-collared caribou on the north-eastern mainland have reached different interpretations leading to uncertainty about the trends, distribution, and structure of subpopulations in that area (Nagy et al., 2011; Gunn et al., 2013b). Detailed discussion of this topic is beyond the scope of this report.

Figure 76. Ranges and recent trends of northern caribou populations in Canada.

The time spans used to assess the recent trends vary, depending on survey data available. This map has been updated from the version published in Canadian Biodiversity: Ecosystem Status and Trends 2010 (Federal, Provincial and Territorial Governments of Canada, 2010). Further information on herd trends is available in the ESTR technical thematic report on caribou (Gunn et al., 2011c).

Source: Gunn et al. (2011c). Bluenose East Herd information updated with 2013 census information  (Environment and Natural Resources, 2014)
Long description for Figure 76

.This map shows the ranges and recent trends of northern caribou populations in Canada. The various populations trends are described as increasing, decreasing, stable at mid to high population, stable at low population, unknown/under study, and preliminary data. Population trends vary across the Arctic. Decreasing populations are located south of and along the southern boundary of the Arctic Ecozone+, on Baffin Island, and in the far north (Peary caribou). The increasing populations are the Teshekpuk, Central Arctic, and Porcupine populations in northern Yukon and Alaska. Cape Bathurst and Bluenose-West populations are shown as stable at low population, and Cape Churchill and Southampton Island populations are stable at mid to high population. Dolphin and Union, Wager Bay, Lorillard, Qamanirjuaq and South Baffin Island populations are shown as unknown/under study, and Ahiak and North Baffin Island populations show preliminary data only.

Top of Page

On the mainland, numbers were low from the 1950s to the 1970s, when the major herds began to increase (Kelsall, 1968; Gunn et al., 2011c). The increases continued into the 1980s. All eight major mainland caribou herds from the Western Arctic east to Hudson Bay declined following their peak abundance in the mid-1980s to mid-1990s (the exact timing depends on the herd). Two herds considered to be still in decline are the Leaf River and George River herds.

By 2012, the Cape Bathurst and Bluenose-West herds had stabilized at extremely low numbers following a period of sharp declines. A 2012 census for the Bathurst Herd revealed a slight decrease in the number of breeding females and a slight increase in the number of younger caribou--essentially the herd is stable at a very reduced abundance, despite the harvest being reduced from over 3,000 to 300 (B. Croft, pers. comm., 2012). At such low numbers it is difficult to detect whether “stability” is a slow decline, a slow recovery, or no trend. After a calving ground photographic census in 2008, which was the first census since 1994, the trend for the Qamanirjuaq Herd was determined to be a statistically insignificant decline.

The George River Herd declined following the mid-1980s, based on the census results for 2010. The neighboring Leaf River Herd, which increased from the mid-1980s at least until the most recent census (2001), is now considered to be declining based on information on demographic rates.

The status of the Ahiak and several herds on the northeast mainland (Wager Bay, Lorillard, Melville Peninsula, and other smaller herds on Boothia Peninsula and Simpson Peninsula), Baffin Island, and the smaller islands in Hudson Bay are currently unknown. The exception is the Southampton Island Herd whose abundance is tracked during aerial surveys at relatively regular intervals. By 2007, the herd had declined to half the peak size estimated in 1997 (30,000 caribou) (Gunn et al., 2011c) and the decline continued to about 7,900 caribou by 2011 (Greer, 2013). The 2013 survey resulted in an estimate of 7,000 caribou, with a higher proportion of calves considered to indicate that the population is stabilizing (based on interview with M. Campbell reported in Greer, 2013). On Baffin Island, Inuit observations and science-based studies indicate that caribou numbers are at a low in the cycle of abundance (Baffinland Iron Mines Corporation, 2012; Department of Environment, 2013b).

The trends in abundance are based on one indicator--the number of caribou in the herd, estimated either through calving-ground or-post calving counts (Gunn and Russell, 2008). In a few herds, such as the Bathurst and George River herds, the trends in total numbers are supported by measured trends in demographic indicators such as adult or calf survival. In other herds, especially the Beverly Herd, monitoring of herd size was infrequent and supporting data on demographic rates were not collected.

The rates of increase and decrease of individual herds vary greatly as can be seen when the rates of change for herds are plotted for periods when they were increasing (after 1970) and periods when they were decreasing (generally after the 1990s) (Figure 77). The herds with the greatest rates of increase were the Southampton and Bathurst, while the Bluenose-West and Porcupine herds showed the lowest rates of increase among herds for which there are sufficient data. During the decline phase, the Cape Bathurst Herd had the greatest rate of decline, although, with only a few breeding females on the Beverly traditional calving grounds in recent surveys, the rate of decline of the Beverly Herd may have been greater. Data are insufficient for the Beverly Herd to calculate this rate.

Figure 77. The exponential rate of increase and decline of major tundra dwelling caribou herds in Canada.

The chart shows the annual rate of change during increase and decline phases, based on conversion of the population estimates to natural logarithms. The years used vary among herds depending on when herds were increasing and decreasing and when population estimates were made. For the Porcupine Herd, where a change in trend direction was detected in the 2010 survey, the rate of increase is the average of 0.033 (1972–1989) and 0.035 (2001–2010).

Source: Gunn et al. (2011c)
Long description for Figure 77

This graphic represents a bar graph showing the following information:

The exponential rate of increase and decline of major tundra dwelling caribou herds in Canada.
HerdExponential rate of declineExponential rate of incres
Cape Bathurst-0.170.09
George River-0.140.1
Leaf River00.09

Top of Page

The current declining trends for some caribou herds, as well as the recent declining trends with current indications of stabilization or recovery for other herds, are likely a reflection of natural cycles in caribou abundance accentuated by the cumulative effects of increasing human presence on the caribou ranges (Gunn et al., 2013c). More conjectural is to what degree climate warming and attendant broad-scale habitat changes are factors in the natural cycles.

The causes of declines are complex, with the roles of the various contributing factors changing as the declines continue. Caribou are similar to other northern herbivorous mammals (voles, lemmings, and hares) in that their abundance is cyclic (Morneau and Payette, 2000; Gunn, 2003b; Zalatan et al., 2006) and, overall, the cycles are likely driven by climate interacting with forage availability, predation, and pathogens. Weather tends to have a decadal pattern, influenced by major patterns, such as the Arctic Oscillation, switching between negative and positive phases (Bonsal and Shabbar, 2011). Winter temperatures and snowfall patterns interact with forage growth and availability. Winter conditions and forage availability influence caribou condition, which determines birth rates and calf survival (Couturier et al., 2009a; Couturier et al., 2009b). Trends in annual calf survival and fecundity also play a role in changing herd abundance.

Weather also interacts with parasites such as warble flies, whose activity depends on summer weather. Weather affects the transmission of internal parasites, which in turn influences forage intake as caribou try to reduce their exposure to the parasites (Van der Wal et al., 2000). Predation and harvest by humans have a pivotal role in declines as even small annual reductions in adult female survival strongly influence population trends (Gaillard et al., 1998). Without mandatory harvest reporting, however, it is not possible to assess the impact of harvesting on the caribou populations.

Combining population estimate data on migratory herd numbers (including the Dolphin and Union population) since 1970 and scaling herd size relative to maximum estimates for each herd indicates that, on average, caribou numbers have increased from lows around 1975 to a peak around 1995, followed by a decline with some indication of a recent levelling off or reversal of the decline (Figure 78). The timing and magnitude of the changes vary.

Figure 78. The relative size of tundra dwelling wild Rangifer herds (Canada).

The line represents the six-year running average. Other symbols represent individual herds. Relative population size is calculated as the population estimate for the year as a proportion of the maximum recorded estimate. Note that the maximum recorded estimate is not necessarily the peak population over this timeframe, as surveys usually did not cover the entire period and were not conducted every year.

Source: Gunn et al. (2011c)
Long description for Figure 78

This scatterplot and line graph shows the following information:

The relative size of tundra dwelling wild Rangifer herds (Canada). (Relative size of herds)
YearPorcu-pineCape BathurstBluenose-WestBluenose-EastBathurstDolphin and UnionBeverlyQamanir-juaqSouth-ampton IslandGeorge RiverLeaf River6 year running average
19830.76----- 0.46--0.160.42
1984----0.81-0.96 -0.83-0.54
1993------0.32 -1.00-0.69
2000-0.580.681.00------ 0.76

Top of Page

The cumulative effects of increasing human presence on caribou ranges (number of people as well as non-renewable resource exploration and extraction and infrastructure development) are largely unknown. However, tools are being developed to examine how responses of the individual caribou can be scaled up to measure population-level effects (Gunn et al., 2011b). Some recently constructed mine projects monitored effects on caribou and reported changes in caribou distribution and time spent foraging (Gartner Lee Limited, 2002). In response to large open-pit mines on the tundra summer range of the Bathurst Herd, caribou distribution was reduced in a 10 to 15 km zone of influence around the mines (Boulanger et al., 2004). Changes in the atmospheric transport of contaminants on individual caribou body burdens are monitored for some herds (Gamberg, 2009) and the results evaluated in relation to potential impacts on human health. These evaluations conclude that nutritional benefits of consuming caribou far outweigh any risks from the low levels of contaminants (Van Oostdam et al., 2005; Donaldson et al., 2010).

See Gunn et al. (2011c) for herd-specific assessments. These include summaries of status and trends and notes on each herd’s ecology.

Trends in distribution

Herd distributions will change through time; shifts in distribution, however, are not well documented and are uncertain. Information from aerial surveys and satellite-collared individuals generally has not been analyzed to describe trends in distribution. Migratory tundra caribou characteristically shift their winter distribution among years and winter ranges often overlap between neighbouring herds (Schmelzer and Otto, 2003; Bergerud et al., 2008). Additionally, as herd abundance rises and falls, distribution--especially winter distribution--can shift (Bergerud et al., 2008). Maps of historical distribution (Banfield, 1961) and winter distribution since the 1970s, at least for the Beverly, Qamanirjuaq, and Bathurst herds (Gunn et al., 2001; BQCMB, 2004), hint at a contraction in the southern boundary of the winter distribution in northern Manitoba, Saskatchewan, and Alberta. During the 1996 to 2010 decline of the Bathurst Herd, the winter distribution of the satellite-collared cows showed a trend towards wintering further north of the 60th parallel (Gunn et al., 2011b).

Top of Page


The muskox (Ovibos moschatus) is important as a symbol of the Arctic tundra and its long association with people, extending back to Paleolithic times (Lent, 1988). Its ecological role stems from its function as a large-bodied social herbivore, well-adapted for the pulse of summer productivity and the long winters that characterize tundra ecology. In many tundra areas, muskoxen are the only large-bodied prey available to wolves and humans during winter, as muskoxen are rarely long-distant migrants.

Muskoxen are distributed through northeast Canada and Greenland and have been introduced or re-introduced to Alaska, western Greenland, Scandinavia and Russia (Figure 79 ). Muskoxen reintroduced to Alaska have spread, in small numbers, to the Yukon North Slope in the 1980s (Wildlife Management Advisory Council North Slope, 2008). By the early 1900s, muskox abundance in Canada had collapsed on the Arctic mainland and on some Arctic islands. Since then, numbers have built up, through natural recovery and range extension, aided by 30 to 50 years of almost no harvesting. Combining the most recent estimates for the islands and mainland, Canada has about 114,300 muskoxen, which is about three-quarters of the world’s muskoxen (for references, see below). Most muskoxen occur on the Arctic islands (85%), and 77% of Northwest Territories and Nunavut muskoxen occur on the large islands (Banks and Victoria), even though these two islands form only 40% of the landmass of the Arctic islands’ muskox range.

Circumpolar muskox status and trends

The world distribution of muskoxen, including introduced and re-introduced populations, is shown in Figure 79. Muskoxen have spread widely from their release sites, perhaps in search of new range as populations expanded (Reynolds, 1998; Gunn and Adamczewski, 2003). Alaska muskox numbers have stabilized and, in 2013, numbered over 4,200 in five regions (Harper, P., 2011; Gunn et al., 2013a). In Greenland, status and trends in muskoxen are unknown. The most recent survey, in northeast Greenland about 20 years ago, estimated 9,500 to 12,000 muskoxen (Boertmann et al., 1992). More is known about the introduced population at Kangerlussuaq, where the initial 27 muskoxen introduced in the 1960s increased rapidly. Rough estimates are for a current population of 10,000 to 25,000. Other populations in West Greenland have also shown good growth, and quota-based harvesting has been implemented. Based on the above, the total numbers of muskoxen in Greenland may be from 20,000 to 40,000 (Greenland Institute of Natural Resources, 2012). In Russia, re-introductions started in the mid-1970s in several locations. The total muskox population in Russia was estimated at over 10,000 in 2013, with the great majority (8,700) being on the Taimyr Peninsula (Gunn et al., 2013a).

Figure 79. Muskox world distribution.

Not all introduction sites represent current populations: for example, 17 muskoxen were introduced to Svalbard in 1929 and the small population that became established was extinct by 1982 (Long, 2003). The small areas marked in Nunavut are locations where populations were successfully introduced (Chubbs and Brazil, 2007). Muskoxen have since extended their range, but the current range is not sufficiently defined to map.

Source: Updated from Wikimedia Commons, based on Gunn and Forchhammer (2008) and Large Herbivore Network (2009)
Long description for Figure 79

This map of the circumpolar Arctic shows the global distribution of muskox and distinguishes between endemic and introduced populations.  The map shows muskox are endemically distributed through northeast Canada and northern Greenland and have been introduced or re-introduced to Alaska, western Greenland, Scandinavia, and Russia.

Top of Page

Since the Pleistocene, muskoxen have varied little genetically and in appearance. Skull and dental characteristics from contemporary and Pleistocene muskoxen are similar (Harington, 1970) and a comparison of ancient and modern DNA also suggests muskoxen have changed little over thousands of years and their variation has not been sufficient to recognize subspeciation (Tener, 1965; Groves, 1995). This is not to say that mainland and Arctic island muskoxen do not differ, but that the differences were not enough to warrant sub-specific designation. Muskoxen have notably low genetic variability, though mainland muskoxen are slightly more genetically variable than island muskoxen, based on microsatellite DNA analysis (Van Coeverden de Groot, 2001).

Information on changes in the number or distribution of distinct muskox populations is not available, as Canadian muskoxen are managed for harvesting based on management units (rather than natural populations). Often, for the Arctic Islands, the management units are islands or groups of islands. On the mainland, the units are based on hunting patterns and on changing information about distribution, and the managment unit boundaries are modified at intervals. Nunavut’s management units were recently revised as part of the development of new muskox management plans (Dumond, 2006; Kivalliq Wildlife Board, 2010; Government of Nunavut, 2012). The NWT has seven muskox management units (Environment and Natural Resources, 2012a). The boundaries of aerial survey areas do not always coincide with management unit boundaries, or they may cover several units, which complicates analysis of historical trends (Fournier and Gunn, 1998; Dumond, 2006).

Trends in abundance

The record of European exploration and settlement explains why muskoxen on the mainland were reduced to isolated remnants by the early 1900s, as traders encouraged commercial hunting for sales of hides in European markets (Barr, 1991a). The muskox hides were marketed to replace the loss of bison robes as the bison numbers were collapsing, also in part due to unregulated commercial harvesting (Barr, 1991a). From 1862 to 1916, the Hudson’s Bay Company purchased more than 15,000 muskox hides (Tener, 1965). The company continued to purchase hides as the number brought in fell from more than 1,000/year in 1891 to only a few score in 1908, reflecting the declining population. In 1916, the last year before the Canadian government provided legislative protection, just one muskox skin was sold (Figure 80).

Figure 80. The rise and fall of the muskox fur trade in Canada: total numbers of muskox hides purchased annually by the Hudson’s Bay Company, 1862 to 1916.
Source: Tener (1957)
Long description for Figure 80

This bar graph shows the following information:

The rise and fall of the muskox fur trade in Canada: total numbers of muskox hides purchased annually by the Hudson's Bay Company, 1862 to 1916.
YearMuskox hides
1892no data
1893no data
1894no data
1895no data
1896no data

Top of Page

On the Arctic islands, European explorers hired Inuit to provide meat for their expeditions, and Hone (1934) calculated that, between 1852 and 1916, British and Canadian expeditions to Melville Island killed over 600 muskoxen for meat and to capture calves for zoos. Over 1,000 muskoxen were killed on Ellesmere Island by Norwegian and American explorers between 1880 and 1917 (Barr, 1991b). However, how those harvests affected long-term trends is essentially unknown. On Banks and western Victoria islands, muskoxen had essentially disappeared by the late 1800s. This drastic decline was attributed to ice storms (Gunn, 1990).

Harvesting of muskoxen was suspended from 1924 to 1969, when a small quota was established for southern Ellesmere Island. The Thelon Game Sanctuary (later to become the Thelon Wildlife Sanctuary) was established in 1927, in large measure to protect muskoxen (Taylor, 2006). During the 1970s and 1980s, conservative quotas for harvesting became more widespread as hunters’ reports and surveys revealed increasing muskox numbers (Barr, 1991b). Recovery appeared to be relatively slow over the decades, partly as a consequence of lack of surveys to track the recovery, and partly because the muskoxen were dispersing and recolonizing their previously occupied ranges (Barr, 1991b). Muskoxen have a potential doubling rate of three years, although that is relatively rare. They are similar to other large-bodied herbivores, having high adult survival, low fecundity, late onset of breeding, a single birth every one to three years, and females having a lifespan often exceeding 15 years (Gunn and Adamczewski, 2003). Muskox numbers reached an estimated 134,000 in 2001 (summarized in Dumond, 2006) but, since then, have declined, mostly because of the decline in muskoxen on Banks Island. Although the mid-Arctic islands of Banks and Victoria are only 9% of the landmass of the Arctic tundra, they are home to 72% of Canada’s muskoxen. Only 14% of muskoxen are on the mainland. Estimates of Canadian muskox abundance from 1967 to 2012 are shown in Figure 81 . The breakdown of the most recent estimates by location across the Arctic Ecozone+ is shown in Table 11.

Figure 81. Estimated muskox abundance in Canada, 1967–2012.
Source: Dumond (2006) and references therein; references in Table 11
Long description for Figure 81

This bar graph shows the following information:

Estimated muskox abundance in Canada, 1967–2012.

Top of Page

Table 11. Muskox population estimates by region
RegionAreaMuskox numbersError estimate
(where available)
Survey yearSource
NunavutBathurst Island Complex82-2001Jenkins et al. (2011)
NunavutCornwallis18-2002Jenkins et al. (2011)
NunavutDevon Island513302–864 (95% CI)2008Jenkins et al. (2011)
NunavutPrince of Wales2,0861,582–2,746 (95% CI)2004Jenkins et al. (2011)
NunavutSomerset1,910962–3,792 (95% CI)2004Jenkins et al. (2011)
NunavutNorth Ellesmere Island8,1156,632–9,930 (95% CI)2006Jenkins et al. (2011)
NunavutSouth Ellesmere Island456312–670 (95% CI)2005Jenkins et al. (2011)
NunavutAxel Heiberg Island4,2373,371–5,325 (95% CI)2007Jenkins et al. (2011)
NunavutAmund Ringnes, Ellef Ringnes, King Christen, Cornwall, Meighen, and Lougheed islands21-2007Jenkins et al. (2011)
NunavutKing William Island300-2007White (2002)
NunavutWest of Bathurst Inlet (MX 19)Note * of Table 112,141±586 SE2005Dumond (2007a)
Nunavut(MX 14, western part)434±168 SE2005Dumond (2007a)
NunavutWest of Coppermine R.132±71 SE2007Dumond (2007b)
NunavutBoothia Peninsula
High density area
348±62 SE2006Dumond (2007c)
NunavutMedium density area645±119 SE2006Dumond (2007c)
NunavutLow density area104±72 SE2006Dumond (2007c)
NunavutEast Queen Maud Gulf2,621379 SE2000Campbell and Setterington (2001)
NunavutNortheast Kivalliq165-2000Campbell and Setterington (2001)
NunavutNorth of Baker Lake4,736554 SE2010Campbell and Setterington (2001)
NunavutSoutheast Victoria Island (NU)25,000-30,000-1992-1999Dumond (2006)
NunavutThelon Sanctury (NU and NT)1,095±281 SE1994Gunn et al. (2009)
Northwest TerritoriesNorthwest Victoria Island (NT)11,4421,637 (95% CI)2010Davison et al. (2013)
Northwest TerritoriesBanks Island36,6764,031 (95% CI)2010Davison et al. (2013)
Northwest TerritoriesMelville Island3,033852 (95% CI)2012Davison and Williams (2012)
Northwest TerritoriesPrince Patrick Island507320 (95% CI)2012Davison and Williams (2012)
Northwest TerritoriesEglinton Island213211 (95% CI)2012Davison and Williams (2012)
Northwest TerritoriesSouth Paulatuk1,215±525 SE2002Community of Paulatuk et al. (2008)
Northwest TerritoriesNorth Great Bear1,457±919 SE1997Veitch (2013)
Northwest TerritoriesAlymer Lake161±39 SE1991Shank and Graf (2013)
Northwest TerritoriesArtillery Lake1,606±277 SE1998Bradley et al. (2001)
Northwest TerritoriesBeaverhill Lake532±149 SE2000Gunn et al. (Gunn et al., 2009)
YukonYukon North Slope101-2011Wildlife Management Advisory Council (North Slope) (2012)
YukonQuebec (Nunavik)1,400-2003Chubbs (2007)

Surveys are conducted in the spring and are for one-year-plus-old muskoxen. In general, when there is no range or error estimate provided, the number respresents a mimimum count (i.e., the actual number seen in the survey).

Notes of Table 11

Note * of Table 11

MX indicates Nunavut muskox management units

Return to note * referrer of table 11

Top of Page

Trends in regional populations

Arctic mainland

By the 1930s, remnants of muskox populations (with a total of perhaps about 500 muskoxen) were scattered across the Arctic mainland, with a cluster of herds north of Great Bear Lake, south of Bathurst Inlet, in the Thelon Game Sanctuary, and south of Boothia Peninsula (Anderson, 1934; Barr, 1991b). As hunters reported sightings of muskoxen, aerials surveys were undertaken for several mainland regions. North of Great Bear Lake, muskoxen recolonized, and their numbers increased during the 1990s (Fournier and Gunn, 1998; Dumond, 2006).

While the muskoxen continued to spread west toward the Mackenzie River, their numbers declined behind the “colonizing front}. For example, north of Great Bear Lake and west of Coronation Gulf, muskoxen increased from scattered remnants in the 1900s to 425 in 1967 (see review in Dumond (2007b)). In a small portion of this area, the Rae-Richardson watersheds, there were 869 ± 300 (SE) in 1980, 1295 ± 279 (SE) in 1983, and 1,805 ± 289 (SE) in 1988 (Fournier and Gunn, 1998; Dumond, 2007b). However, between 1989 and 1994, the muskox numbers declined, possibly due to a parasitic lungworm (Umingmakstrongylus pallikuukensis) which may have increased muskox vulnerability to grizzly bear predation (see section on Wildlife diseases and parasites on page 62). Between 1994 and 2007, numbers appeared to stabilize. The 1994 estimate of 540 ± 139 (SE) is similar to a 2007 estimate of 509 adult muskoxen (Dumond, 2007b).

East of the Coppermine River to Bathurst Inlet, a 2005 survey found 2,141 ± 586 (SE) muskoxen in the western part of the study area, indicating an increase in muskox abundance since a 1991 survey. However, the abundance had declined to 434 ± 168 (SE) in the eastern part of the study area, compared to a 1986 survey (Dumond, 2007a).

Few muskoxen were reported in the Queen Maud Gulf and Adelaide Peninsula area until the 1960s. By August 1982, however, numbers were estimated at 8,494 ± 2,673 (SE) (Gunn et al., 1984). The trend between 1982 and 1996 was a 50% decline in numbers, as the 1996 estimate was 4255 ± 680 (SE) (Fournier and Gunn, 1998). It is unclear to what extent the decline is a distributional shift, as hunters in Kivalliq and northeast Kitikmeot areas report increasing numbers and expanding distribution through the 1990s (Campbell and Setterington, 2001). South of the Queen Maud Gulf area, in the Kivalliq Region, muskoxen were surveyed in 1986, 1991, 1999, and 2010. Preliminary analyses of the 2010 findings indicate both population and range expansion. The 2010 estimate was 4736 ± 554 (SE) muskoxen with a colonizing edge moving east toward Hudson Bay coast and a decline in densities in the survey area (Ford et al., 2012).

Further north and east of the Queen Maud Gulf, Gunn and Dragon (1998) estimated 554 muskoxen in 1995 on the Boothia Peninsula where they were scarce or absent through the 1980s (Gunn and Dragon, 1998). This number had approximately doubled to an estimated 1,097 by 2006 (Dumond, 2006).

South and west of the Thelon Game Sanctuary, muskoxen were re-colonizing their former ranges toward the treeline. The estimated number of adult muskoxen in the sanctuary in 1994 was 1,095 ± 281 (SE) (Gunn et al., 2009). Hunters in Lutsel K’e also reported increasing numbers of muskoxen east of Artillery Lake and surveys in 1989 and 1998 revealed that muskox numbers had doubled and that muskoxen had spread into the taiga to the west and southwest (Bradley et al., 2001). A survey documented the increased spread of muskoxen south toward the treeline when, in 2000, 1,320 ± 183 (SE) muskoxen were estimated (Gunn et al., 2009).

Muskoxen disappeared from the Yukon North Slope between 1858 and 1865 and were re-established in the Arctic National Wildlife Refuge in Alaska in 1969–1970 (Reynolds, 1998). From there a few animals moved into the Yukon in 1972 (L. Harding, pers. obs.) but it was not until the mid-1980s that repeated sightings of cow muskoxen were reported; mixed groups with calves were first sighted in 1987 (Johnson et al., 2005). The range continued to expand east to the Mackenzie River and south into the Taiga Cordillera Ecozone+. Because the North Slope muskoxen move between Alaska and the Yukon, population estimates include animals on the Alaska and Yukon North Slope and areas to the south. The North Slope population was estimated in 2006 to be 400 animals, a decline from 700 animals estimated in the mid-1990s, based on extensive aerial surveys. A total of 291 muskoxen were spotted in the survey conducted in 2011--of these, 101 were in the Yukon (Wildlife Management Advisory Council (North Slope), 2012). Reports of sightings by Aboriginal people in the region indicate a decline since the 1990s (Arctic Borderlands Ecological Knowledge Co-op, 2006).

Muskoxen from Ellesmere Island were initially introduced to a farm in 1967 near Kuujjuaq and then released between 1973 and 1983 at three locations in Nunavik (Lehenaff and Crete, 1989a; Chubbs and Brazil, 2007). The muskoxen rapidly increased to 1,400 by 2003 and have been sighted in Labrador (Chubbs and Brazil, 2007).

Arctic islands

The most recent muskox numbers available for the Arctic Islands indicate a total of 94,130 (non-calves): these estimates span the period 1992 to 2012. This total includes 13,300 muskoxen on the eastern Queen Elizabeth Islands, based on aerial surveys between 2004 and 2008 (Jenkins et al., 2011). No trend for these islands is discernible as the previous estimate was in 1961 (Tener, 1963).

On the western Queen Elizabeth Islands, the more frequent surveys in the early 1970s, mid-1980s, and 1990s indicate that there have been two sharp declines and one or two periods of recovery. The total number of estimated muskoxen in 2012 on the western Queen Elizabeth Islands was 3,750, an increase of 48% over 15 years (Davison and Williams, 2012).

The trends for the western Queen Elizabeth Islands can be related to a pattern of increases and sharp decreases during winters with above-average snow depths and icing (while acknowledging that the weather data are sparse). For the islands in the Western Arctic with sufficient number of estimates to describe trends, there is very little regional consistency. This can be seen for Prince Patrick, Melville and Bathurst islands, which are stretched over 600 km across a progressively decreasing Beaufort maritime influence (Maxwell, 1981). Muskox numbers on Prince Patrick Island increased from none to 507 ± 320 (95% confident interval) between 1961 and 2012, while, during the same period, muskox numbers on Melville Island peaked in 1986 and declined by the mid-1990s before recovering to 3,033 ± 852 (95% confident interval) (Davison and Williams, 2012). Based on aerial surveys, muskoxen on Bathurst Island have been through two declines and one recovery since 1961 (Figure 82 ). In 1961, muskox numbers were relatively high: Tener (1963) estimated 1,136 muskoxen. When the population was next estimated in 1973, it had declined 59% and, by August 1974, the continued decline had reduced the estimated number of muskoxen to 164 ± 70 (SE) after a severe winter (Miller et al., 1977). Muskox numbers recovered and increased between 1974 and 1993 to an estimated 1,200 (Ferguson, 1987; Miller, 1987; Miller, 1998). A series of three winters, 1994–1997, with rain on snow and unusually deep snow reduced the muskox abundance and, in summer 1996, Miller (1998) estimated that there were 425 ± 136 (SE) muskoxen alive. By 1997, only an estimated 124 ± 45 (SE) one-plus-year-old muskoxen were alive, which indicated a 90% decline since 1994 (Gunn and Dragon, 2002). During an aerial survey in 2001, only 94 muskoxen were observed and no population estimate was derived (Jenkins et al., 2011).

Figure 82. Muskox trends, Bathurst Island, 1961 to 2001.

Error bars are ± standard error. See text for more information on trends.

Source: Tener (1963); Miller et al. (1977); Miller (1998); Gunn and Dragon (2002); Jenkins et al. (2011)
Long description for Figure 82

This bar graph shows the following information:

Muskox trends, Bathurst Island, 1961 to 2001. (Muskox numbers)

Top of Page

The muskox trends on the larger, more southern mid‐Arctic islands are quite different from those on the High Arctic islands (Queen Elizabeth Islands). A relatively well‐documented trend is that of Banks Island, where the population grew from hundreds in the 1960s, based on Inuvialuit accounts to a peak around 2001, followed by a decline to the most recent survey in 2010 (Table 11, Figure 83) (Gunn et al., 1991b; Inuvialuit Game Council, 2002; Lyberth, 2003; Davison et al., 2013).

Figure 83. Muskox trends, Banks Island, 1982 to 2010.

Error bars are 95% confidence intervals.

Source: Davison et al. (2013)
Long description for Figure 83

This bar graph shows the trends in muskox populations on Banks Island between 1982 and 2010. The graph shows that the population grew from approximately 10,000–12,000 in 1982, to a peak of over 60,000 around 2001, followed by a decline to under 40,000 in the most recent survey in 2010.

Top of Page

Similar to the trends on Banks Island, muskox numbers increased on Victoria Island during the 1970s to 1990s: the Nunavut part of Victoria Island had an estimated 25,000 to 30,000 muskoxen in 1992–1999 (Dumond, 2006; Gunn and Patterson, 2012) and the Northwest Territories portion had an estimated 11,442 in 2010 (Davison and Williams, 2013). In 2004 surveys of the large islands east of Victoria Island, 2,086 muskoxen were estimated on Prince of Wales, and 1,910 muskoxen were estimated on Somerset Island (Jenkins et al., 2011). The muskox populations on those two islands have been increasing since at least the 1970s (Nagy et al., 1996; Fournier and Gunn, 1998; Jenkins et al., 2011; Gunn and Patterson, 2012; Davison and Williams, 2013).

Trends in distribution

The Canadian population is almost entirely within the Arctic Ecozone+ except for in the Thelon Wildlife Sanctuary and, recently, where muskoxen have spread along the treeline from northeast of Great Slave Lake toward the Saskatchewan border. They are distributed on the tundra in areas with either shallow snow (20 to 40 cm) or sufficient relief for the wind to maintain areas with shallow snow. Summer habitats include low-lying river valleys or coastal plains with sedge meadows, riparian willows, and gravel bars. Muskoxen are widely distributed and occur in the Low, Middle, and High Arctic tundra ecoregions (an area of 2,293,372 km2).

There have been no overall analyses yet of the trends in distribution. Since the early 1900s, muskoxen have recolonized most of the mainland tundra except northeastern areas, including the Melville Peninsula. Likewise, muskoxen occur on most Arctic islands except the northeast islands (Baffin and the islands in Hudson Bay). Muskoxen disappeared from Baffin Island during the fifteenth century, with only ocassional records since then, such as a herd of eight muskoxen observed south of Clyde River in 1968 (Barr, 1991b). The recolonization of historical ranges has been natural, except for the Yukon North Slope and the successful introduction to northern Quebec. Muskoxen apparently did not naturally colonize Quebec after the glacial retreat (Lehenaff and Crete, 1989b). The rate of recolonization is slow--perhaps less than 10 km/yr (Fournier and Gunn, 1998).

Muskoxen occasionally appear on and disappear from smaller and medium-sized islands. Tener (1963) did not find muskoxen on Prince Patrick, Eglinton, and Mackenzie King islands in 1961, although they were recorded there during the aerial surveys in 1972–1974, 1986, 1997, and 2012 (Miller et al., 1977; Gunn and Dragon, 2002; Jenkins et al., 2011; Davison and Williams, 2012). Within the islands, relatively little analysis has been undertaken for trends in distribution. Even so, information does suggest some changes. Between 1972 and 1980, exceptionally high densities of muskoxen were found on Bailey Point on southwest Melville Island (Thomas et al., 2013). The densities were from 0.6 to 1.1 muskoxen/km2 over the entire peninsula, reaching 2.6/km2 below 100 m, which was comparable to other “hotspots” of muskox abundance, including the Thomsen River, Banks Island. Bailey Point was considered to have a particularly favourable microclimate to serve as a refugium for muskoxen. The high densities persisted at least to 1987 (Miller, 1988), but in 1997, when numbers had declined by about 50% since 1987, almost no muskoxen were seen at Bailey Point, a situation which persisted in 2005 and 2012 (Gunn and Dragon, 2002; Davison and Williams, 2012).

Trends in harvest

Muskoxen are harvested under a quota system introduced in 1969. Quotas have increased and, by 2011, in Nunavut had reached a Total Allowable Harvest of 2,303 muskoxen (Giroux et al., 2012b). This includes harvest for domestic use, commercial harvest, and sport hunting by non-residents (Wildlife Research Section, 2011). In the Northwest Territories, the total harvest quota is 1,112 in the Inuvialuit Settlement Region, which is mainly two quotas that cover Banks and northwest Victoria islands (NWT Environment and Natural Resources, 2011). These two quotas were assigned to influence the rate of increase in the 1990s and to encourage commercial harvesting. Annual use of the quotas varies, especially in the commercial harvesting for meat and qiviut (wool) sales.

Figure 84. Annual Inuvialuit total muskox harvest, 2006 to 2011.
Source: NWT Environment and Natural Resources (2011)
Long description for Figure 84

This bar graph shows the following information:

Annual Inuvialuit total muskox harvest, 2006-2011
NW Victoria244208270211217

Top of Page


Threats that cause or contribute to population declines may lie in the unexpected, which may  be related to shortcomings in data collection or predictive models, or unanticipated environmental changes accumulating in ways not encompassed by traditional knowledge. The extent of effective management (including habitat and landscape management planning) is the deciding factor for anticipating and dealing with threats. The status of muskox management planning varies between Nunavut and Northwest Territories. Within Nunavut, regional management plans with a focus on harvest management have recently been completed (Dumond, 2006; Kivalliq Wildlife Board, 2010; Government of Nunavut, 2012) which include adaptive management to adjust harvest levels if popuation levels change. The Inuvialuit Settlement Area has community conservation plans and harvest reporting (Community of Paulatuk et al., 2008; NWT Environment and Natural Resources, 2011); governance, however, is more complex elsewhere in the Northwest Territories. The harvest, except on Banks and northwest Victoria islands, is relatively low and an overall approach to muskox management is absent.


Predation is a natural driving force in muskox evolutionary ecology. However, almost nothing is known about how muskoxen trade off the risks of predation (and parasitism) relative to foraging. Wolf predation on muskoxen is likely common, with packs or single wolves observed killing adult as well as younger muskoxen (Gray, 1987; Mech and Adams, 1999). On northern Ellesmere Island, and northern and eastern Greenland, muskox occurrence in wolf scats ranges from 65–98%, with lemmings and Arctic hares as the next most frequent item (Marquard-Petersen, 1998). On Banks Island, where Peary caribou are less numerous, muskox remains were found in 103 out of 115 stomachs and 34 out of 38 scat samples (1992–2001), with lemmings being the next most frequent item (Larter, 2013). Trends in wolf predation on muskoxen are unknown.

Although unmeasured, an increase in grizzly bear sightings on muskox ranges is being reported (Ford et al., 2012). Newborn calves that cannot keep up with a herd are easily killed by pursuing bears (Clarkson and Liepins, 1994). Deep, crusted snow increased the vulnerability of muskoxen in at least one instance of a grizzly bear killing a bull (Case and Stevenson, 1991).

Predation in some areas, such as west of Kugluktuk, may be accentuated by infection with a lungworm (see section on Wildlife diseases and parasites on page 62 and below). Muskox abundance increased and peaked at 1,800 in 1987, but then declined by about 50% in 1994 and remained stable in 2007 (Dumond, 2007b). The muskoxen were infected with the lungworm Umingmakstronylus pallikuukensis, which may increase muskox vulnerability to predation, especially from grizzly bears, as the lungworm forms cysts in the lungs, causing difficulties breathing.

Health and trends in health

It is difficult generally to detect trends in health, as knowledge of muskox parasites has been increasing since the 1990s. However, it likely that the spread of at least one parasitic lungworm may be associated with a warmer climate. In the 1990s the lungworm Umingmakstrongylus pallikuukensis was described (as a previously unknown species) with a range restricted to the western Canadian Arctic mainland, extending west of Kugluktuk (Hoberg et al., 1995). Climate models predicted range expansion of the parasite and it has now been found on southernVictoria Island (Kutz et al., 2009) (see section on Wildlife diseases and parasites on page 62).  

Muskoxen may have a low resistance to parasites even when in good body condition (Alendal and Helle, 1983; Korsholm and Olesen, 1993). This characteristic may be related to their low genetic variability. Muskoxen in some areas, notably Banks Island, are subject to outbreaks of the bacterial disease Yersinia pseudotuberculosis,which may have an environmental trigger such as unusually warm weather. Exposure to Yersinia is high (Larter and Nagy, 1999) and periodic outbreaks include deaths. A previously unknown disease in muskoxen that led to deaths on Banks Island in 2012 was erysipelas, a bacterial disease that is common in livestock (M. Branigan. pers. comm., 2012).

Weather-related deaths

Most weather-related deaths are indirect, as they are caused by malnutrition. The general projections for a warmer climate include more incidents of rain on snow, which are associated with muskox deaths, at least on Banks Island (Rennert et al., 2009; Nagy and Gunn, 2009). Periods of rain when calves are young can also increase mortality (P. Hale, pers. comm., 2013).

Industrial disturbances

The trend is for increasing industrial exploration and development, including construction of new roads. Predicting how muskoxen and their habitat will be influenced is uncertain as relatively little is known about cumulative effects on muskoxen. Earlier studies focused on behavioural responses of muskoxen to seismic and aircraft disturbance (e.g., Miller and Gunn, 1980). Studies in Alaska related to the National Petroleum Reserve provide more information on potential disturbance and mitigation measures associated with development (Bureau of Land Management, 2012).

Top of Page

Large carnivores

Wolves are the major large carnivore in Arctic tundra systems, with grizzly bears being important in much of the Southern Arctic. Medium-sized predators include Arctic and red foxes and wolverines, and avian predators include snowy owls and jaegers. Grizzly bears are discussed above in the section on Trends in species of conservation concern (page 119), and wolves and wolverines are discussed below. See also the section on Community and population dynamics (page 56) and the Main threats to caribou (page 170) for discussion of predator-prey dynamics.

With changes in land-use patterns and the expansion of human populations in southern Canada, carnivores have lost much of their former North American ranges, making the northern regions of the continent increasingly important for species conservation (Figure 85).

Figure 85. Reduction in ranges for large carnivores in North America.
Source: Hummel and Ray (2008)
Long description for Figure 85

This figure is comprised of three maps of the United States and Canada showing the reduction in ranges for the grey wolf, grizzly bear and wolverine.  The maps show that these carnivores have lost much of their former North American ranges, particularly in the south. However, within the Arctic Ecozone+, their ranges are largely unchanged.

Top of Page


A well-known species and top predator, the continued presence of wolves is an important indicator of Arctic ecosystem integrity. Wolf populations are not monitored regularly. Moreover, since some follow caribou and some do not, interpreting abundance and distribution data would be problematic for some areas. Wolves in highly productive environments can reproduce rapidly, making them resilient to various types of disturbance, including hunting.

Gray wolves, Canis lupus, in the Arctic Ecozone+ were formerly considered to include up to eight subspecies (Mech, 1974). Although subspecies status is currently in doubt, genetic analyses suggest that wolves in the Northern Arctic (sometimes called “Arctic” wolves) are separable from Southern Arctic (“tundra”) wolves (Carmichael et al., 2007), and that these, in turn, are separable from boreal forest (“gray”) wolves (Musiani et al., 2007). Aside from their genetic structures, the three groups have different prey specialization patterns (Carmichael et al., 2001; Musiani et al., 2007).

Wolf populations of the Arctic Archipelago form a metapopulation, some populations of which have endured severe bottlenecks that probably occurred during catastrophic declines in their principal prey, Peary caribou and muskoxen. Arctic hares are also important prey species. From 2000 to 2006, Arctic wolves, Arctic hares, and muskoxen were counted in a study area on Ellesmere Island. Wolf numbers were strongly correlated with hare numbers, but not with muskox numbers (Mech, 2007).

Wolves on Banks, Ellesmere, and Devon islands show genetic signatures of recent population declines (Carmichael et al., 2008). Island wolf populations in general have significantly less genetic variability than mainland populations, suggesting that there is movement among the islands and that populations recover after a decline, mainly through recolonization from other islands (Carmichael et al., 2008). The wolves’ genetic diversity appears to have been augmented, however, by periodic migration from mainland populations, occurring primarily through two corridors: Baffin Island and Victoria Island. This gene flow could be compromised by the loss of sea ice due to climatic warming and increased human activity (Carmichael et al., 2008). In particular, year-round marine traffic would keep an open water route through certain marine passages, blocking movement of wolves in the winter.

Tundra wolves in the Southern Arctic make heavy use of eskers, which are also a source of gravel for industrial uses such as mine site and road construction, a resource often in short supply in Arctic infrastructure construction (McLoughlin et al., 2004). Some tundra wolves are caribou specialists and follow migrating caribou for long distances (Frame et al., 2004).

Trends in wolves

There is some information on trends in wolf abundance, based on community and traditional knowledge and on observations of wolves during systematic aerial surveys to count caribou and muskoxen. There is no direct information on predation rates.

On Banks Island, wolves were poisoned in the 1950s and recovered slowly, increasing during the 1980s and 1990s (Figure 86 ) and often spotted in areas with high muskoxen density.

Figure 86. Wolves sighted during aerial surveys of Peary caribou and muskoxen, Banks Island, NWT.

Group size includes adults and pups. Values for 1994 and 1998 are averages of two surveys.

Source: data from Species at Risk Committee (2012a)
Long description for Figure 86

This bar graph shows the following information:

Wolves sighted during aerial surveys of Peary caribou and muskoxen, Banks Island, NWT.
YearGroups observedAverage group size
1994Note * of Figure 8612.52.7
1998Note * of Figure 86123.1

Note of Figure 86

Note 1 of Figure 86

Values for 1994 and 1998 are averages of two surveys.

Return to note * referrer of figure 86

Top of Page

For the Bathurst Herd, wolf sightings between 1987 and 2008 during late winter aerial surveys to estimate caribou calf survival suggest no consistent trend in either wolf sightings or mean pack size (Williams and Fournier, 1996; Gunn, 2013) (B. Croft, pers. comm., 2010). During that time, the numbers of caribou declined and the number of wolves seen at their dens, as well as the number of occupied dens, declined (D. Cluff, unpublished data). The rate of sightings in the vicinity of caribou, however, suggests that predation rates were maintained.

Top of Page


Arctic-nesting birds winter in many parts of the world where they may be vulnerable to stressors including loss of food supplies and habitat, pollution, disturbance, and overharvesting during winter and during migration. In the Arctic, they are vulnerable to changes in their habitat and food supplies and, in some cases, to overharvest. Data are lacking or sparse for many species, and it is often difficult to determine the causes of trends. Many Arctic-nesting shorebird and landbird species are known to be declining, as are some sea ducks. Other Arctic bird groups, such as geese and swans, have mainly stable populations or have increased over the past few decades. Trends in Arctic seabirds (not including sea ducks) are discussed in the ESTR report Ecosystem status and trends report: Arctic marine ecozones (Niemi et al., 2010). An overview of seabird trends for all of Canada is presented below (text box and Figure 87 ). For more information on the status of bird populations in Canada see The State of Canada’s Birds 2012 (NABCI, 2012).

Trends of seabirds in Canada

Seabirds nesting in the Arctic are covered in more detail in the ESTR report on Arctic marine ecosystems (Niemi et al., 2010). This text box, based on the ESTR review of status and trends of seabirds in Canada (Gaston et al., 2009a), is extracted from the report Canadian biodiversity: ecosystem status and trends 2010 (Federal, Provincial and Territorial Governments of Canada, 2010).

Figure 87. Trends in the status of breeding seabird populations in Canada, 1980s to 2000s.

Note: only populations with significant breeding populations, long-term datasets, and those unaffected by terrestrial human activities are included.

Source: adapted from Gaston et al. (2009a)
Long description for Figure 87

This graph shows the following information:

Trends in the status of breeding seabird populations in Canada, 1980s to 2000s. (Number of populations in each category)

Top of Page

Worldwide, the status of seabirds is deteriorating faster than any other bird group (BirdLife International, 2010). In Canada, trends are regional in nature and result from a variety of factors, including climate change, fishing by-catch, resource extraction, transportation, and pollution (Coe and Rogers, 1997; Melvin and Parrish, 2001; Schreiber and Burger, 2002; Stenseth et al., 2004; ACIA, 2005; Gaston et al., 2009a). A trend to an earlier breeding date has been found in several populations (Bertram et al., 2001; Gjerdrum et al., 2003; Hipfner and Greenwood, 2008), as have changes in diet and condition (Parsons et al., 2008).

With the exception of ivory gulls (Pagophila eburnea), which are declining rapidly, change in Arctic seabird populations is slow and possibly the result of events on wintering grounds in the Northwest Atlantic (Gaston, 2002; Gaston et al., 2003). Changes in seabird diet and growth have been found to be related to reduction of Hudson Bay sea ice. This may have negative consequences for populations in the long term (Gaston et al., 2003). Conversely, in the High Arctic, less sea ice may benefit the birds (Gaston et al., 2005; Gaston et al., 2009b).

Top of Page


Waterfowl (ducks, geese and swans) are among the most important wildlife species for Inuit for food and ceremonial uses and large concentrations of migrating and nesting waterfowl are a prominent feature of the ecozone+. Status and trends of waterfowl are comparatively well known as federal agencies in the USA and Canada monitor populations annually in order to set hunting limits and guide conservation programs. There is limited monitoring on Arctic nesting grounds, but for many species regular or periodic monitoring is conducted on wintering grounds or at points where the birds congregate during migration. This section describes trends for selected Arctic tundra nesting waterfowl species with substantive parts of their global populations nesting in the Canadian Arctic. In addition, the ecozone+ provides important habitat for waterfowl species, such as Canada geese (Branta canadensis) and scoters (Melanitta spp.) that also nest in more southerly ecozones+.



King eiders (Somateria spectabilis) nesting in the Canadian Arctic overwinter both in the eastern and western part of the continent. There is growing evidence that the western population of King Eiders has declined over past few decades. The total population of king eiders nesting in the western and central Canadian Arctic was estimated in 1960 to be 900,000; in the early 1990s the population had dropped to between 200,000 and 260,000 (Canadian Wildlife Service Waterfowl Committee, 2008). Counts are conducted during spring and fall migration at Point Barrow, Alaska, about every 10 years (Peacock et al., 2013; Quakenbush et al., 2013). The most recent estimate indicates that numbers remained stable between 1996 and 2003 (Figure 88). Surveys on nesting grounds in western Victoria Island conducted in the early 1990s were repeated in 2004 and 2005 (Raven and Dickson, 2006), showing a 56% decline in abundance (Figure 89).

Figure 88. King eiders migration counts at Point Barrow, Alaska, 1953 to 2003.

The western population of king eiders migrate past Point Barrow, Alaska, each spring and fall to nesting areas in Alaska and western and central Canadian Arctic. Error bars are 95% confidence intervals.

Sources: Suydam et al. (2000) and Quakenbush et al. (2013)
Long description for Figure 88

This line graph shows the king eiders migration counts at Point Barrow, Alaska, from 1953 to 2003. The most recent estimate indicates that numbers were stable from 1953 to 1970 at just under 800,000, increased slightly from 1970 to 1976 , declined to about 300,000 between 1976 and 1994, and fluctuate minor amounts between  1994 and 2003.

Top of Page

Figure 89. King eider nesting populations, western Victoria Island, 1990s and 2000s.

Error bars are 95% confidence intervals.

Source: Raven and Dickson (2006)
Long description for Figure 89

This bar graph shows the following information:

King eider nesting populations, western Victoria Island, 1990s and 2000s.
YearNW Victoria IslandSW Victoria Island

Top of Page

Data on wintering grounds in Greenland indicate that the eastern population of king eiders is also declining, though this decline may be related to a shift in distribution related to human disturbance (Canadian Wildlife Service Waterfowl Committee, 2008). Surveys in 2010 confirmed that large numbers of king eiders were wintering at the northern tip of Labrador and southern tip of Baffin Island (Canadian Wildlife Service Waterfowl Committee, 2012). King eiders nesting in the Rasmussen Lowlands, Nunavut, declined 86% between surveys conducted in 1974–1975 and repeated in 1994–1995 (Gratto-Trevor et al., 1998).

Three subspecies of common eiders (S. mollissima)nest in the Canadian Arctic: Pacific (nesting in the western and central Arctic), northern (nesting in the Eastern Arctic), and Hudson Bay common eider.

Numbers of Pacific common eiders in migration counts at Point Barrow, Alaska, declined by 50% between 1976 and 1996, then increased again by 2002 (Figure 90). Areas of particular importance for nesting are the Dolphin and Union Strait, Coronation Gulf, and Queen Maud Gulf. The breeding population for the central Arctic was estimated at about 37,000 in surveys conducted during 1995–1998 (Canadian Wildlife Service Waterfowl Committee, 2012). Additional surveys, conducted at nesting areas in Bathurst Inlet to establish a baseline for detection of trends, indicate a decline of about 50% since 1995 (Canadian Wildlife Service Waterfowl Committee, 2012).

Figure 90. Pacific common eider migration counts at Point Barrow, Alaska, 1976 to 2003.

95% confidence intervals are shown where available.

Source: Suydam et al. (2000) and Quakenbush et al. (2013)
Long description for Figure 90

This line graph shows the following information:

Pacific common eider migration counts at Point Barrow, Alaska, 1976 to 2003.

Top of Page

Figure 91. Pacific common eider counts on nesting grounds in Bathurst Inlet area, 1995, 2007, and 2008.
Source: Raven and Dickson (2009)
Long description for Figure 91

This bar graph shows the following information:

Pacific common eider counts on nesting grounds in Bathurst Inlet area, 1995, 2007, and 2008.

Top of Page

Northern common eiders winter along the southern coasts of Newfoundland and Labrador, Quebec, and Greenland. Surveys in Greenland, where most overwintering occurs, indicate severe declines since the 1970s (Canadian Wildlife Service Waterfowl Committee, 2012). No clear trends in nesting populations were shown in surveys in Ungava Bay (historical data and surveys in 2000), while surveys along the Labrador coast showed increases between the 1980s and early 2000s. Surveys have been conducted on the wintering range in eastern Canada every third year since 2003. Results to 2009 indicate a stable population of about 200,000 (Canadian Wildlife Service Waterfowl Committee, 2012).

Hudson Bay eiders, wintering entirely in Arctic waters near Belcher Islands and off the western coast of Quebec, are subject to periodic weather-related population crashes (Canadian Wildlife Service Waterfowl Committee, 2008). Belcher Islands breeding eiders declined 70% between 1985–1988 and 1997 (Erickson and Meegan, 2007), probably as a result of a large die-off during the winter of 1991–1992 that occurred when areas of open water froze (Robertson and Gilchrist, 1998). There have been no significant winter kill events since then, and the population appears to be recovering (Canadian Wildlife Service Waterfowl Committee, 2012).

Threats: harvest and disease

Declines in some eider populations may be related to harvest. An estimated 115,000 eiders were harvested in 2001 in Chukotka, but the proportion of these that nest in Canada is not known (Canadian Wildlife Service Waterfowl Committee, 2012). The commercial and subsistence harvest of northern common eiders in southwest Greenland was estimated at over 100,000 birds, a large proportion of which are Canadian Arctic nesters (Canadian Wildlife Service Waterfowl Committee, 2008). Demographic modelling indicated that this was not sustainable, and more restrictive harvest regulations were put in place in Greenland from 2002 to 2004 (Canadian Wildlife Service Waterfowl Committee, 2012). Subsistence harvest in Canada is relatively low, but better information on harvest is needed, especially for the northern common eider.

Avian cholera may also be an emerging issue for northern common eiders--the first Arctic outbreak was recorded in 2004 in northern Quebec, and it has recurred there and in the vicinity of Southampton Island (Canadian Wildlife Service Waterfowl Committee, 2012). A small island off Southampton Island, the largest breeding colony of northern common eiders in the Canadian Eastern Arctic, has been studied since 1996. Colony size was stable at around 5,000 pairs between 2001 and 2005, then nearly doubled to an estimated 9,800 pairs in 2006. During an outbreak of avian cholera mortality, colony size dropped to 4,700 pairs in 2007 (Buttler, 2009). Further studies have demonstrated that outbreaks of avian cholera in common eider colonies at a frequency of more than one outbreak per decade could drive colonies to extinction (Descamps et al., 2012).

Tundra swans

The western population of tundra swans (Cygnus columbianus) breeds along the Alaska coast and winters in the western United States, while the eastern population breeds from Alaska to the northeast shore of Hudson Bay and Baffin Island, wintering in the eastern United States.

Estimates of the eastern population of tundra swans acquired through midwinter surveys on the U.S. wintering ground (Figure 92) show fluctuations around a mean of about 90,000 to 100,000 birds.

Figure 92. Tundra swan (eastern population) abundance, based on January surveys on wintering grounds, 1982–2012.
Source: U.S. Fish and Wildlife Service (2012)
Long description for Figure 92

This line graph shows the following information:

Tundra swan (eastern population) abundance, based on January surveys on wintering grounds, 1982–2012. (Estimated number, thousands)
YearTundra swan (eastern population) abundance
(Estimated number, thousands)

Top of Page

Snow geese

Snow geese breed only in the Arctic, including along the Hudson Bay coast. Northernmost eastern breeding populations, greater snow geese (Chen caerulescens atlanticus), breed in the Canadian Eastern Arctic and northern Greenland and winter along the U.S. Atlantic coast, staging in southern Quebec during migration. More southerly and westerly breeding populations, lesser snow geese (C. c. caerulescens), nest in colonies in coastal and inland areas across the Canadian Arctic and winter in southern Canada and all North American flyways.

Greater snow geese increased dramatically from a few thousand birds in the 1930s to over 500,000 in the early 1990s, counted on the main staging grounds in southern Quebec (Canadian Wildlife Service Waterfowl Committee, 2012). The increase is related to changes in agricultural practices in the U.S. wintering grounds leading to a change of feeding habits from marshes to cultivated fields (Canadian Wildlife Service, 2003). The spring 2012 population estimate was about one million geese (Canadian Wildlife Service Waterfowl Committee, 2012). The population has remained fairly stable since 1999 when special conservation measures were implemented (Reed and Calvert, 2007). The dramatic increase in this population has had significant impacts on staging and breeding areas in Canada through overgrazing (Environment Canada, 2007a).

Lesser snow geese also show population growth, with a levelling off of the population, as estimated through the midwinter counts, after the late 1990s and a recent increase (Figure 93). This trend is also evident in inventories on major nesting grounds, with the exception of West Hudson Bay (Figure 93). Inventories provide only a sampling, and banding and harvest studies indicate that the population of lesser snow geese may be much higher than has been estimated--likely exceeding 15 million in 2010 (Canadian Wildlife Service Waterfowl Committee, 2012).

Figure 93. (A) Nesting lesser snow geese estimated through photo-inventories of major breeding colonies 1970s to 2000s, and (B) midwinter abundance indices, 1970 to 2012.

Midwinter counts are primarily of lesser snow geese that breed in the central and Eastern Arctic.

Sources: Canadian Wildlife Service Waterfowl Committee (2012) (breeding colony data) and U.S. Fish and Wildlife Service (2012) (midwinter population counts)
Long description for Figure 93

This figure is comprised of a bar graph (A) and a line graph (B) showing the following information:

Nesting lesser snow geese estimated through photo-inventories of major breeding colonies 1970s to 2000s, and midwinter abundance indices, 1970 to 2012.
Major breeding colonies1970s1980s1990s2000s
Baffin Island446,600455,0001,735,0001,620,000
Southampton Island155,800233,000721,000653,000
West Hudson Bay317,000436,300212,000261,000
Queen Maud Gulf56,000317,100740,600-
Western Canadian Arctic169,600207,500486,100580,000
Nesting lesser snow geese estimated through photo-inventories of major breeding colonies 1970s to 2000s, and midwinter abundance indices, 1970 to 2012.
YearMidwinter population: abundance index

Top of Page


This section is extracted from the ESTR technical thematic report prepared for this ecozone+, Shorebird trends Arctic Ecozone+ (Gratto-Trevor et al., 2011) and has not been comprehensively updated.

The Arctic Ecozone+ is of great importance globally for shorebird production. Sixty percent of North American shorebirds breed in the Arctic. The Canadian Arctic alone provides 75% of the North American breeding range for 15 of the 49 species of shorebirds that are common to North America (Donaldson et al., 2000).

Globally, 44% of estimated population trends for Arctic-breeding shorebirds are declining (Figure 94) making the problem more widespread than was originally thought (Morrison et al., 2001). Overall, the Arctic breeders as a group are declining 1.9% per year (Bart et al., 2007).

Figure 94. Summary of population trends for Arctic-breeding shorebirds, 2003.

Globally, population trends have been estimated for 52% of Arctic-breeding shorebirds (100 biogeographical populations of 37 species). Of these, 12% are increasing, 42% are stable, 44% are decreasing and 2% are possibly extinct.

Source: Delany and Scott (2006)
Long description for Figure 94

This bar graph shows the following information:

Summary of population trends for Arctic-breeding shorebirds, 2003.
Population trend% of population for which trends are known
Possibly extinct2

Top of Page

An analysis of fall migration count data was undertaken to determine if the declining numbers of birds recorded on migration counts could be explained by changes in migration routes or timing or by changes in detection rates (Bart et al., 2007). The authors concluded that migration counts most likely reflected a true reduction in population size. They found no evidence of major shifts in the number of birds migrating along specific routes and no major changes in variables related to detection. Annual rates of change were calculated over the period 1974 to 1998 in this study--results are shown in Figure 95 for Arctic-breeding shorebirds with sufficient survey counts in fall migration surveys conducted in the Canadian-United States North Atlantic or United States Midwest regions.

Figure 95. Estimated trends in Arctic-breeding shorebird fall migration counts, 1974–1998.

NA = North Atlantic migration survey; MW = Midwestern migration survey.

Source: data from Bart et al. (2007)
Long description for Figure 95

This bar graph shows the following information:

Estimated trends in Arctic-breeding shorebird fall migration counts, 1974–1998.
TrendsAnnual percent changeP
Black-bellied plover (North Atlantic migration survey)-5<0.01
American golden-plover (North Atlantic migration survey)-7.2<0.01
Sempalmated plover (North Atlantic migration survey)0.4-
Whimbrel (North Atlantic migration survey)-3.3-
Hudsonian godwit (North Atlantic migration survey)-3.50.01<p<0.05
Ruddy turnstone (North Atlantic migration survey)-12.3-
Red knot (North Atlantic migration survey)-3.3-
Sanderling (North Atlantic migration survey)-3.3-
Semipalmated sandpiper (North Atlantic migration survey)-40.05<p<0.1
White-rumped sandpiper (Midwestern migration survey)-5.2-
Baird's sandpiper (Midwestern migration survey)-1-
Pectoral sandpiper (North Atlantic migration survey)-4.5<0.01
Dunlin (North Atlantic migration survey)-2.6-
Red-necked phalarope (Midwestern migration survey)-7.60.01<p<0.05

Top of Page

Two major shorebird trend reviews by the U.S. Shorebird Conservation Plan Committee (in 2001 and 2004) and Canadian Wildlife Service Shorebird Committee (in 2001) assessed 18 species of Arctic-breeding shorebirds with very similar results (Table 12). Eight species were listed in both assessments as having significant population declines (Brown et al., 2001; Morrison et al., 2001; U.S. Shorebird Conservation Plan, 2004).

Table 12. Population trend assessments for Arctic-breeding shorebirds.
SpeciesTrend SummaryU.S. Shorebird Conservation Plan (2004)CWS Shorebird Committee
Black-bellied plover (Pluvialis squatarola)Significant declining population trendSignificant declineSignificant decline
American golden-plover (Pluvialis dominica)Significant declining population trendSignificant declineSignificant decline
Semipalmated plover (Charadrius semipalmatus)Conflicting informationNot enough informationSignificant decline
Eskimo curlew (Numenius borealis)Significant declining population trendSignificant declineLikely extinct
Whimbrel (Numenius phaeopus)Conflicting informationSignificant declineMixed trends
Hudsonian godwit (Limosa haemastica)Probable or declining population trend, not statistically significantNot enough informationDecline
Ruddy turnstone (Arenaria interpres)Significant declining population trendDeclineSignificant decline
Red knot (Calidrus canutus)Significant declining population trendSignificant declineSignificant decline
Sanderling (Calidrus alba)Significant declining population trendSignificant declineSignificant decline
Semipalmated sandpiper (Calidris pusilla)Significant declining population trendSignificant declineSignificant decline
White-rumped sandpiper (Calidris fuscicollis)Not enough information to conclusively determine population trend (mixed trends)Not enough informationMixed trends
Baird’s sandpiper (Calidris bairdii)Conflicting informationNot enough informationDecline
Pectoral sandpiper (Calidris melanotos)Not enough information to conclusively determine population trend (mixed trends)Not enough informationMixed trend
Purple sandpiper (Calidris maritima)Conflicting informationStableSignificant decline
Dunlin (Calidris alpina)Significant declining population trendSignificant declineSignificant decline
Buff-breasted sandpiper (Tryngites subruficollis)Probable or declining population trend, not statistically significantDeclineDecline
Red-necked phalarope (Phaloropus lobatus)Significant declining population trendDeclineSignificant decline
Red phalaropen (Phalaropus fulicarius)Significant declining population trendSignificant declineSignificant decline

What is of most concern is that over the past 30 years many species trends have changed from slightly declining to significantly declining, indicating that the decline is persistent and ongoing (Morrison et al., 2001; Delany and Scott, 2006). The declines are observed in species with a range of migration, habitat, and breeding strategies and needs. Preliminary investigations by Thomas et al. (2006a) and Bart et al. (2007) found no common factors among declining species.

In the U.S. Shorebird Conservation Plan (Brown et al., 2001), population trend information was combined with five other variables (relative abundance, threats during breeding season, threats during non-breeding season, breeding distribution, and non-breeding distribution) to create a conservation prioritization scheme. The scheme, adopted in the Canadian Shorebird Conservation Plan (Donaldson et al., 2000), is useful because species with stable or slightly downward-trending populations with threats on their wintering grounds and very specific breeding ground habitat requirements may be more at risk than species with significant population declines. The highest priority species were those designated ‘highly imperiled’. Using this prioritization scheme, the only Arctic species listed in 2001 as ‘highly imperiled’, Eskimo curlew (Numenius borealis), is believed to be extinct (Environment Canada, 2007b).

In 2004, species were re-evaluated (U.S. Shorebird Conservation Plan, 2004) and the status of several species was upgraded (Table 13).

Table 13. Conservation status of tundra-nesting shorebirds as classified in the U.S. Shorebird Conservation Plan.
Highly imperiled (first priority)Species of high concern (second priority)
  • Eskimo curlew (Numenius borealis) (believed to be extinct)
  • Note a of Table 13*Buff-breasted sandpiper (Tryngites subruficollis) (globally)
  • Note a of Table 13*Red knot (Calidrus canutus) (Canadian Arctic-Atlantic Coast population)
  • American golden-plover (Pluvialis dominica) (globally)
  • Whimbrel (Numenius phaeopus) (North American populations)
  • Hudsonian godwit (Limosa haemastica) (globally)
  • Ruddy turnstone (Arenaria interpres) (North American populations)
  • Red knot (Calidrus canutus) (populations other than the Canadian Arctic-Atlantic Coast population)
  • Sanderling (Calidrus alba) (North American populations)
  • Note a of Table 13Dunlin (Calidris alpina) (Alaska-East Asian and Alaska-Pacific Coast populations)

Source: U.S. Shorebird Conservation Plan (2004)

Notes of Table 13

Note a of Table 13

Upgraded species are denoted with an asterisk (*)

Return to note a referrer of table 13

Top of Page

Local studies have recorded population declines over a range of periods. Analysis of the Atlantic coastal migration stop-overs from 1972 to 1983 (Howe et al., 1989) found significant declines for black-bellied plover (Pluvialis squatarola) (decreasing by 5.4% per year), whimbrel (Numenius phaeopus) (–8.3% per year) and sanderling (Calidrus alba) (–13.7 per year). Breeding populations of red phalarope (Phalaropus fulicarius), black-bellied plover, and American golden-plover (Pluvialis dominica) decreased significantly, by 76, 87, and 79% respectively, in the Rasmussen Lowlands (Central Arctic) over a 20-year period (Gratto-Trevor et al., 1998). Given the long time interval between studies, natural fluctuation as a result of a series of poor breeding seasons rather than a persistent and continuous population decline could explain the differences between the two study periods, but it may represent a true decline in these species (Gratto-Trevor et al., 1998).

A study in the Foxe Basin (Prince Charles and Air Force islands) found significant population declines for white-rumped sandpiper (Calidris fuscicollis) (–61%) and red phalarope (–43%) over an eight-year time span (1989–1997) (Johnston and Pepper, 2009). For red phalarope the decline was even more pronounced at East Bay, Southampton Island, where there was a 93% decline over six years (1999–2005) (Pirie et al., 2012). All shorebird species (n=5) at East Bay declined by more than 90% over the same interval. In 2007 there was a small rebound to about 33% of the original 1999 values. This coincided with a high lemming (and therefore low predation) year (Pirie et al., 2012).

Near Churchill, Manitoba, a comparison of six qualitative bird abundance studies between 1930 and the 1990s found that semipalmated sandpiper (Calidris pusilla), stilt sandpiper (Calidris himantopus) and red-necked phalarope (Phalaropus lobatus) experienced a ‘great decrease’, and dunlin (Calidris alpina) a ‘decrease’ (Jehl and Lin, 2001). Huge declines were also noted at La Perouse Bay, Manitoba (40 km east of Churchill), for semipalmated sandpiper and red-necked phalarope (Gratto-Trevor, 1994).

One of the current major limitations to determining population trends for Arctic-breeding shorebird species is the lack of reliable population estimates. In many cases intensive surveys of shorebirds on the Arctic breeding grounds have led to increases to the world population estimate for a given species (Johnston et al., 2000; Latour et al., 2005; Johnston and Pepper, 2009). This does not reflect an increase in world population size but instead is an indication that initial population estimates were probably low (Brouwer et al., 2003; Morrison et al., 2006). The large-scale Program for Regional and International Shorebird Monitoring (PRISM), which has an Arctic component, is partway through a multi-year survey program that will produce continental population estimates for 19 species of shorebirds that breed in the North American Arctic. Once the first pass of surveys is complete, a second set is planned to assess species-specific as well as North American Arctic-wide population trends (Skagen et al., 2003; Bart and Earnst, 2004; Bart et al., 2005; Bart and Johnston, 2012).

Proposed causes of shorebird population declines include: loss of migration stop-over sites, loss of wintering habitat, and life history characteristics (migratory behaviour, life history, and biogeography) which may predispose shorebirds to population decline. Future population decline is expected to be accelerated by habitat changes on the Arctic breeding grounds.

Since many shorebirds are long-distance migrants that tend to gather in very large numbers at relatively few sites, loss of one or two major stop-over sites could have a huge effect on shorebird populations. Declining food availability at existing stop-over sites can also have a large impact on populations because birds may not be able to take in enough fuel to move to the next stop-over site, or may not be able to acquire the body stores essential for survival and successful reproduction (Senner and Howe, 1984; Donaldson et al., 2000; Morrison et al., 2001; Baker et al., 2004; Morrison et al., 2004; Morrison et al., 2007). Analysis of population trends of North American shorebirds found species that followed continental migration routes (as opposed to coastal or oceanic migration routes) were at higher risk of population decline because of ecosystem loss and alteration (Thomas et al., 2006a; Bart et al., 2007). Continental migrants use small, ephemeral ponds and wetlands that are scattered over a large area. These ponds and wetlands are difficult to delineate for conservation initiatives making it harder to preserve them as compared to larger stop-over sites (Thomas et al., 2006a). Little is known about Arctic stop-over sites because of their remoteness. Observations along a 200 km stretch of coast line in the Kivalliq Region (northwestern Hudson Bay) during the 2008 spring migration found hundreds of High Arctic nesting migrants feeding on insects in the wrack lines on their journey north to the breeding grounds (Johnston and Rausch, unpublished data). The importance of sites such as these to migration and subsequent breeding success is not known.

Loss or degradation of habitat on the non-breeding grounds from human activities such as oil pollution (Harrington and Morrison, 1980), mechanical dredging or fishing (Piersma et al., 2001), conversion of native grasslands and wetlands to agriculture (Isacch and Martinez, 2003; Shepherd et al., 2003), and tourism and development on marine beaches (Blanco et al., 2006) may be a cause of population decline (Thomas et al., 2006a). Complicating the assessment of the importance of wintering habitat is that little is known about food resources on the wintering grounds (Morrison et al., 2004). Threats on the wintering grounds, however, have been found to have a weak influence on the likelihood of a species being in population decline (Thomas et al., 2006a).

The intrinsic biology of shorebird species may make them more susceptible to population decline. Migratory behaviour (such as distance and routes) is suspected to be the most influential intrinsic factor, with more continental migrants in population decline than coastal or oceanic migrants (Thomas et al., 2006a). Phylogenetic characteristics such as body and clutch size, lifespan, and relatedness were found to be unimportant to population decline, but limited clutch sizes means that recovery following a decline is likely to be slow (Myers et al., 1987). Sexual selection may have an influence on declining populations since most socially polygamous species have declining populations while socially monogamous species have stable or increasing population trends--but the data are not conclusive. There are no clear intrinsic factors held in common by shorebird species with declining population trends and extrinsic factors are more likely to be the primary cause of decline (Thomas et al., 2006a; Thomas et al., 2006b; Bart et al., 2007).

Habitat changes in the Arctic caused by climate change are expected to have an exacerbating effect on the declining population trends of Arctic-breeding shorebirds (Bart et al., 2007). Arctic-breeding shorebirds are adapted to the annually variable weather conditions of the Arctic during the breeding season. However, their conservative life-history strategy (low reproduction and long lifespan) makes it difficult for them to adapt to climate change. This puts Arctic-breeding shorebirds more at risk of population decline than other groups (Donaldson et al., 2000; Meltofte et al., 2007). Effects of climate change on breeding habitat include: drying of tundra ponds (Walsh et al., 2005; Smol and Douglas, 2007a), shrub encroachment (Callaghan et al., 2005a), and asynchrony of insect-chick hatch (Tulp and Schekkerman, 2006).

The synchrony of shorebird chick hatch with the peak of insect emergence is not as critical as hatch occurring when there is sufficient food supply. The availability of the food supply is strongly influenced by weather and a sufficient supply is only available for 40% of the insect season (Tulp and Schekkerman, 2008). The peak date of insect emergence fell between 8 July and 23 July for 75% of the 33-year study period. These earliest and latest peak emergence dates were recorded in consecutive years, showing that the date of peak emergence is not advancing linearly with time. Overall, however, the date of peak insect emergence as well as the range of dates with sufficient food available for the normal growth of chicks is getting earlier in the season (Tulp and Schekkerman, 2008).

Since Arctic shorebirds time nest initiation to occur as soon as the snow melts, the advancement in the timing of insect emergence may not be as critical for the survival of chicks hatched from the earliest nests. It could be a serious problem for chicks from late nests, or from re-nests (clutches laid late to replace an earlier nest that was unsuccessful) because they will hatch too late in the season to obtain sufficient food resources (Meltofte et al., 2007). Studies indicate that, while shorebirds exhibit considerable flexibility in nest initiation, the accommodation in timing may not completely match the change in date of snowmelt, especially in years with exceptionally early springs (Smith et al., 2010; McKinnon et al., 2012; Grabowski et al., 2013). Further analysis is needed to determine if snowmelt is advancing at the same rate as the timing of insect emergence, which would permit birds to nest earlier. It is not known whether shorebirds will be able to adjust their migration strategies to arrive on the breeding grounds sooner in response to an earlier snow-free season. Species which make the final jump to the breeding ground from latitudes closer to the Arctic may be more successful than species that use internal length-of-day cues to initiate migration from very distant wintering grounds (Tulp and Schekkerman, 2008).

Top of Page


This report has been extracted from the ESTR thematic technical report Landbird trends in Canada, 1968-2006 (Downes et al., 2011) and has not been updated with new information.

The Arctic Ecozone+ is relatively pristine and there are few immediate threats to landbirds from human activity, although species are affected by climate change, contaminants, and other wide-ranging factors. All birds listed in Table 14 overwinter in more populated areas of Canada and the United States where development pressures are more intense in both their wintering ranges and along their migration routes. Canada has a high stewardship responsibility for these species because large portions of their Western Hemisphere breeding populations are concentrated in the Arctic Ecozone+.

There are relatively few landbird species in the Arctic Ecozone+ and few data on their population trends. The lack of information on population status and trends has been highlighted as the most pressing conservation need in relation to landbirds for this region (Rich et al., 2004). Because of the remoteness and lack of roads, the Breeding Bird Survey (BBS) has not been carried out in the Arctic and there are few other surveys of breeding birds. However, many birds that breed in the Arctic spend their winters in the United States and more southerly parts of Canada, where their populations can be monitored by the Christmas Bird Count (CBC). The CBC, now over 100 years old, monitors the status and trends of winter bird populations through an all-day, annual census conducted by groups of volunteers throughout North America. Data from the CBC complement the BBS by providing results for some species that cannot be monitored on their breeding grounds. Results presented below are preliminary findings based on CBC data from Canada and the United States combined (Butcher and Niven, 2007).

CBC trends (Table 14 and Figure 96) suggest that several species such as Harris’s sparrow (Zonotrichia querula) and snowy owl, have been undergoing long-term, statistically significant declines since the 1960s. Other species, such as the rough-legged hawk (Buteo lagopus) and Lapland longspur (Calcarius lapponicus), have shown relatively stable overall population trends.

Table 14. Trends in annual abundance of selected Arctic Ecozone+ landbirds, 1966 to 2006, based on Christmas Bird Count results for North America
SpeciesPopulation Trend (%/yr)CBC Abundance Index

CBC Abundance Index

CBC Abundance Index

CBC Abundance Index

CBC Abundance Index

Hoary redpoll (Acanthis hornemanni)–4.97%significant trends (P<0.05)
American tree sparrow (Spizella arborea)–2.16%significant trends (P<0.05)62.856.342.434.4-45%
Harris's sparrow (Zonotrichia querula)–2.13%significant trends (P<0.05)
Snowy owlNote a of Table 14 (Bubo scandiacus)–2.12%significant trends (P<0.05)
American pipit (Anthus rubescens)–0.97%significant trends (P<0.05)
Snow bunting (Plectrophenax nivalis)–0.93%-15.814.411.69.3-41%
Rough-legged hawk (Buteo lagopus)–0.06%-
Lapland longspur (Calcarius lapponicus)0.40%-
Common redpoll (Acanthis flammea)0.60%-1917.818.117.9-6%

"Change" is the percent change in the average index abundance between the first decade for which there are results (1970s) and the 2000s decade (2000–2006).

Source: Downes et al. (2011)

Notes of Table 14

Note a of Table 14

This reflects sightings of snowy owls in southern latitudes. Note, however, that many snowy owls remain the in the Arctic or northern taiga throughout the winter, far from human settlements.

Return to note a referrer of table 14

Top of Page

Figure 96. Annual indices of population change for the rough-legged hawk, snowy owl, Harris’s sparrow, and snow bunting, 1966 to 2005, based on Christmas Bird Count results for North America.

The rough-legged hawk and snow bunting show no significant trends, though the latter may be in decline; the snowy owl and Harris’s sparrow have declined significantly (P<0.05). See the note on snowy owl for Table 14.

Source: Downes et al. (2011)
Long description for Figure 96

These four line graphs show the following information:

A) Rough-legged Hawk
YearChristmas Bird Count
B) Harris's Sparrow
YearChristmas Bird Count
C) Snowy Owl
YearChristmas Bird Count
D) Snow Bunting
YearChristmas Bird Count

Harris’s sparrow, a species with its entire breeding range in Canada, is classified by Partners in Flight as a Continental Watch List species (Rich et al., 2004). The species has apparently experienced a long-term decline over the last 40 years (Figure 96b). Because of its isolated breeding range, direct influence of human activity on the breeding range is unlikely to be a factor in its decline. Harris’s sparrows, however, are susceptible to predation, especially by merlins (Falco columbarius), whose populations are increasing. The influence of factors such as climate change is unknown (Niven et al., 2004; Norment and Shackleton, 2008).

Population indices for the snow bunting (Plectrophenax nivalis) vary annually but this species has apparently experienced a large decline in its population over the long term (Audubon Society, 2007) (Figure 96d). The Arctic has a very high stewardship responsibility for snow buntings (Rich et al., 2004), which breed throughout the Arctic Cordillera and Northern Arctic and the northern portions of the Southern Arctic. Causes of the apparent snow bunting decline are not known and further demographic studies are needed. Potential factors affecting populations on the breeding grounds include availability of nest sites (rock talus or cavities, driftwood cavities, or sometimes buildings), predation on nests and incubating females, and availability of food (Lyon and Montgomerie, 2011). Reduction in snow bunting populations may be related to earlier thawing of the tundra, which allows more woody plants to grow, converting the open foraging habitats preferred by snow bunting to more shrub-dominated communities (Audubon Society, 2007).

Top of Page

Insects and pathogens

Warmer temperatures will benefit free-living bacteria and parasites whose survival and development is limited by temperature and there are indications that this is occurring in some Arctic muskox and caribou populations (see section on Wildlife diseases and parasites on page 62). Arthropods such as ticks that transmit disease agents may also benefit from climate change, and the diseases they spread may consequently become more prevalent or widespread. Other abundant flying insects in the Arctic, especially oestrids (warble flies) and mosquitoes, are a significant factor in caribou ecology and their abundance is closely related to climate conditions. Inuit from the Queen Elizabeth Islands report that mosquitoes and black flies have increased along with warmer weather in certain areas (e.g., Nunavut Tusaavut Inc., 1997).

Top of Page

Major range shifts of species native to Canada

Arctic residents have reported changes in animal behaviour and distribution. Indigenous people of the Arctic especially are familiar with the long-term trends, variability, and extent of species distribution, and habitat use and behaviour. As they travel and work in their lands, they observe changes that cannot be easily detected through other monitoring and research. These observations have been documented through several studies using interviews, questionnaires, and group discussions, but there is not a systematic methodology or repository for this information. See Table 15 for examples of reported observations of range shifts.

Table 15. Examples of documented observations by Inuit regarding range shifts of species native to Canada.
LocationObservations that may indicate species range shiftsReference
NunatsiavutSee new insects and strange kinds of birds: 2001 small yellow and red birds came in large groups that ate
anything including seal skin and seal fat. Hummingbirds seen (new species).
Nickels et al. (2005) Nunatsiavut workshop
NunavikDeer were seen in the summer of 2002 for the first time around the community of Kangiqsujuaq.Nickels et al. (2005) Nunavik workshop
Chesterfield InletInsects, birds and even grizzly bear were reported to be appearing farther north than usual.Nunavut Research Institute (2004)
KitikmeotNew birds seen for the first time such as the robin and an unidentified yellow songbird. More abundant and new species of shrubs and lichens.Thorpe (2000)
Baker LakeAt least ten new kinds of insects were reported to be seen in the area, all winged insects, some recognized from the treeline area.Fox (2004)
Banks IslandNew bird species observed include robins and barn swallows. Shorter winters, longer summers and more water were thought to have caused an increase in the number of insects and led to the arrival of new species of beetles and sand flies. Changes in behaviour include overwintering of birds that normally migrate.Ashford and Castleden (2001)
Beaufort coastRavens and eagles more numerous further north including Tuktoyaktuk.Gordon et al. (2008)
Western Inuvialuit Settlement RegionCougars are new to the treeline areas.Nickels et al. (2005) Inuvialuit workshop

Top of Page

Root et al. (2003) reviewed 143 journal articles related to changes in species distribution (globally) and found that 80% of species show changes that can be explained by physiological constraints. These included changes in density at a particular location, changes in range either poleward or up in elevation, changes in timing of migration, changes in phenology (plant growth and flowering), change in the timing of egg laying, changes in morphology (body size), and shifting frequencies of genetic markers. Many of the species with documented range changes were birds and butterflies, which can more easily adjust their distribution than other taxa (for example, small mammals) that may be constrained by geological features such as rivers (Root et al., 2003).

Climate-related changes in tundra ecosystems--including reduction of lichens, increase in shrubs, and increased canopy height (see section on Changes in tundra plant communities on page 98)--underlie observed and anticipated changes in distribution of flora and fauna in the Arctic Ecozone+ (Gilg et al., 2012; Reid et al., 2013). More rapid changes may be caused by human introductions-- either inadvertently or as deliberate moves to establish populations--for example, the introduction and re-introduction of muskoxen to some regions (see section on Muskoxen on page 129). Land-use change and disturbance from industrial activity and other human activities may also lead to changes in animal distribution.

Range extensions due to climate change could show up as previously vagrant/accidental species that appear more and more frequently and finally become established. Such extensions could involve boreal species moving into the Southern Arctic, or species from that zone moving into the Northern Arctic. Detecting such changes in status requires significant effort in well-designed monitoring programs.

Mammals that have experienced relatively recent range shifts include grizzly bears, which have moved northward in some areas (see the section on Trends in species of conservation concern, on page 104) and moose, which have also expanded their range northward (see text box below). Red foxes have expanded their distribution in parts of the Canadian Arctic and not in others. The current, best-supported hypothesis is that red fox expansion in the 20th century in Arctic Canada was promoted by human food supplementation, and that such food supplementation could lead to similar effects elsewhere if not checked (see the discussion on this topic under Major human stressors on ecosystem structure on page 100).

Expansion of moose northward

Moose (Alces alces) are found primarily in shrub habitats south of the Arctic Ecozone+ but they have been increasing in recent decades in much of the Arctic Ecozone+. Expansion of moose ranges can have an impact on populations of predators and of other ungulates. Moose can provide alternate prey for predators, and alternate food sources in Arctic subsistence economies, especially when caribou are scarce. Wolf populations in caribou winter ranges (in the taiga ecozones+) can increase in response to higher levels of other prey such as moose. When caribou return to the winter range they are preyed on more heavily by the increased number of wolves (e.g., Basille et al., 2013).

Evidence of changes in moose distribution includes:

  • Expansion of moose range northward into suitable habitat has been documented for the Northwest Territories and Nunavut, with more frequent sightings of moose in shrub-rich tundra regions, especially since the 1970s. In the NWT, moose have been sighted as far north as near Bathurst Inlet, Coronation Gulf, and the east side of Victoria Island (NWT Department of Environment and Natural Resources, 2012).
  • Although not historically resident in Nunatsiavut, moose have increased in numbers and range over the past four decades. Moose hunting commenced there in 1977. The northward range expansion of moose reached the treeline in Nunatsiavut in the 1990s (Chubbs and Schaefer, 1997). Inuit report that moose have moved north of Nain and to the Voisey’s Bay area, not far south of the Arctic Ecozone+ (Nickels et al., 2005 Nunatsiavut workshop).
  • Moose from the British–Richardson Mountains of the Yukon (in the Taiga Cordillera Ecozone+) probably began frequenting the Yukon North Slope coastal plain within the last 100 years (Wildlife Management Advisory Council (North Slope), 2008a), utilizing mainly riparian shrub zones. Surveys in the 1980s showed that most of the moose summering along the coastal plain migrated about 100 km south during the winter--but a survey in 2000 (Smits, 2000) as well as observations by Inuvialuit (Arctic Borderlands Ecological Knowledge Co-op, 2003) indicate that increasing numbers are wintering in the Arctic Ecozone+ section of their range . Moose in the northern part of their range in the Yukon are increasing: counts in the Richardson Mountains/coastal plains surveys increased by 67% from 1989 to 2000 (Smits, 2000).

Top of Page

Human stressors on ecosystem composition

Climate change

Climate change is currently a human-induced stressor for some species, such as polar bears, and is thought to contribute to changes in others, including shorebirds and caribou. Implications of climate change are dealt with throughout this report.

Top of Page

Human population increase

Despite the low density of human population in the Arctic Ecozone+, human settlements and associated infrastructure affect ecosystem composition through habitat change, disturbance, and harvest. Increasing human populations place stress especially on large mammals. For example, disturbance to female polar bears during denning, or prior to denning if it alters their choice of dens, may affect growth of cubs (Lunn et al., 2004). Disturbance at den sites may reduce productivity and alter the distribution of tundra wolves (Walton et al., 2001).

The reverse is also true: decreasing intensity of human use may result in increasing wildlife populations. For example, muskox populations in the Queen Maud Gulf mainland and islands increased following withdrawal of Inuit families to more distant settlements in the 1950s. Barren-ground caribou, lesser snow goose, and Ross's goose (Chen rossii) populations have also rapidly increased in the area since the late 1960s (Gunn et al., 1984).

Top of Page

Contaminants and pollution

Sources of Arctic wildlife contamination include long-range transport of persistent organic pollutants (POPs) and mercury, mines, DEW line sites, transportation-related emissions, and at least one military base in addition to the DEW line sites. An emerging issue is potential future contamination related to the anticipated increase in marine traffic and oil and gas activity in the Arctic Ocean as the ice melts.

A local source of pollution, with the potential for transfer of contaminants to wildlife, is liquid and solid waste disposal in Arctic communities. For example, Nunavut dumps are seldom contained and waste is often burned in an uncontrolled manner (ARKTIS Solutions, 2011). Wastewater from small communities is stored in open lagoons that may lack liners. In summer some communities decant the liquids from the lagoons onto the surrounding ground, into lakes, and into the sea (Jamieson and Krkosek, 2013).

Contaminants in wildlife

The concentration of toxic contaminants in wildlife has been a concern in Canada since the 1970s and monitoring has been ongoing since then for some species and some compounds. In the late 1980s concerns about elevated levels of contaminants in wildlife species that were important to the traditional diets of northern Aboriginal peoples led to increased monitoring and research in northern Canada. Early results found a wide variety of substances, many of which had no Arctic or Canadian sources, but which were, nevertheless, reaching unexpectedly high levels in Arctic ecosystems primarily through long range atmospheric transport (INAC, 2008).

Air currents transport several types of contaminants, for example, PCBs and mercury, to the Arctic from southern, more populated and industrial parts of the world. When the airborne pollutants arrive in the Arctic, cold temperatures cause the air, and the pollutants, to fall lower in the atmosphere and to be deposited on the land and into aquatic environments. The contaminants are transported through soil and water, incorporated into biota, and often concentrated as they move up the food chain. Many of these substances become concentrated in particular tissue types, especially fat, or in particular organs such as livers. In general, the highest contaminant levels tend to build up in marine fish-eating animals or their predators, especially those that are long-lived such as whales and polar bears.

Concentrations of all contaminants in wildlife vary depending on the individual, the species, and the location. Trends for some animal populations are shown in Table 16. Three classes of contaminants are of concern in the Arctic Ecozone+ (Stow, 2008):

  • Concentrations of persistent organic pollutants (POPs), such as the pesticide dichlorodiphenyltrichloroethane (DDT), polychlorinated biphenyls (PCBs) and toxaphene, have generally declined (Riget et al., 2010). In long-lived species such as whales, decline of contaminant levels has been slower.
  • Brominated flame retardants [e.g., polybrominated diphenyl ethers (PBDEs) and fluorinated surfactants, perfluorooctane sulfonate (PFOS)] are examples of toxic contaminants that have been increasing since the mid-1980s in most locations and some species [e.g., ringed seals (Pusa hispida); see the ESTR report Ecosystem status and trends report: Arctic marine ecozones (Niemi et al., 2010)].
  • Increases in mercury have been observed since the mid-1970s in some Arctic marine mammals, seabirds, and fish around the circumpolar Arctic, with the highest proportion of increases being in Canada and Greenland (Riget et al., 2011). There were far fewer records for terrestrial mammals, and few are of adequate length to determine trends (Riget et al., 2011). While some of the mercury found in wildlife is from natural sources, much of the mercury found in marine and aquatic systems is from industrial sources.
Table 16. Trends in mercury and persistent organic pollutants from Northern Contaminants Program datasets.
TrendMeasurement taken from…ContaminantTime FrameReference
Significant increasing trendCaribou (Bluenose East Herd)Mercury1994–2002 increase, decline 2002-2006Gamberg (2008)
Significant increasing trendCaribou (Porcupine Caribou Herd)Mercury1993 to 2007 (slight increasing trend)Gamberg (2008)
Significant increasing trendCaribou (Bathurst Herd)Mercury1992–2006Gamberg (2008)
No significant trend but indications of possible change – more data needed to draw conclusionsCharr muscle (Amituk Lake, Resolute)Mercury1989–2007Muir (2008)
Significant decreasing trendCharr muscle (Amituk Lake, Resolute)PCBs, DDTs, toxaphene1989–2007Muir (2008)
No significant trend but indications of possible change – more data needed to draw conclusionsCharr muscle (Amituk Lake, Resolute)Mercury1993–2007Muir (2008)
No significant trendCharr muscle (Amituk Lake, Resolute)PCBs1993–2007Muir (2008)
Significant decreasing trendCharr muscle (Amituk Lake, Resolute)PCBs, DDTs, toxaphene1993–2007Muir (2008)
Significant decreasing trendCharr muscle (Lake Hazen, Quttinirpaaq National Park)PCBs, DDTs, toxaphene1990–2007Muir (2008)
No significant trend but indications of possible change – more data needed to draw conclusionsCharr muscle (Lake Hazen, Quttinirpaaq National Park)Mercury1990 - 2007Muir (2008)
No significant trendCharr muscle (Amituk Lake, Resolute)Mercury1993–2007Muir (2008)
No significant trendCharr muscle (Amituk Lake, Resolute)PCBs1993–2007Muir (2008)
Significant decreasing trendPolar bear (Western Hudson Bay)DDTs1991–2008Letcher (2008)
No significant trendPolar bear (Western Hudson Bay)PCBs1991–2008Letcher (2008)

Top of Page

Based on a review of research on ecological effects of POPs and metals, Fisk et al. (2005) found that concentrations in Arctic terrestrial wildlife, fish, and seabirds are generally below effects thresholds, with a few possible exceptions. These were PCBs in burbot (Lota lota) in some Yukon lakes (in the Boreal Cordillera Ecozone+), Greenland shark (Somniosus microcephalus), glaucous and great black-backed gulls (Larus hyperboreus and L. marinus), and dioxin-like chemicals in seabird eggs. PCB and DDT concentrations in several Arctic marine mammal species exceed effects thresholds, although there was no evidence of stress in these populations. Some polar bears and beluga whales have high enough levels of contaminants that they may experience sublethal effects such as immune suppression. Papers reviewed found weak relationships between cadmium, mercury, and selenium burdens and health biomarkers in common eiders in Nunavut, although these metals were probably not influencing the health of these birds. The authors concluded that there is little evidence that contaminants are having widespread effects on the health of Canadian Arctic organisms, with the possible exception of polar bears. However, they recommended that further research and better understanding of contaminant exposure in Arctic biota is needed considering factors such as tissue levels that exceed effects thresholds, exposure to “new” organic contaminants of concern (such as brominated flame retardants), contaminated regions, and climate change (Fisk et al., 2005).

Top of Page

Sources of local and regional pollution

Current and past activities in the Arctic result in potential for both contamination of wildlife, affecting their safety as human food, and effects on the animals themselves. Overall, the Arctic has relatively few sources of local pollution, with little industry and a very low population density. Sources of pollutants include mining activity, especially abandoned mines, military activities, especially DEW line sites (see text box The Distant Early Warning (DEW) Line in the introductory section on page 12), and oil and gas activities.

An example of localized contamination from a still-active radar station is Saglek Bay, Nunatsiavut (Labrador), which had been the site of a military radar station since the late 1950s. PCB contamination in soils and leaching into the marine environment was discovered in 1986 and cleanup was initiated in 1997–1999. Sediments, fish, and seabirds (black guillemots, Cepphus grylle) have accumulated high levels of PCB concentrations due to marine sediment contamination (Kuzyk et al., 2003; Kuzyk et al., 2005).

In 2005, two sites in the Arctic Ecozone+ had been designated as “contaminated” by the federal contaminated sites program: Kittigazuit Military Site and Atkinson Point Military Site (INAC, 2005), both within the Inuvialuit Settlement Region.

Top of Page