Skip booklet index and go to page content

Woodland Caribou (Rangifer Tarandus Caribou)

Limiting Factors and Threats

General

Some known threats to forest-dwelling caribou, by COSEWIC population, are summarized in Table 7. They are an interrelated mixture of causes and effects and do not include weather or climatic change. Major limiting factors are discussed separately but with the caveat that all of them interact. Treating factors individually is a reductionist approach, which is antithetical to ecology. Part of the problem is hypothesis generation, which often partitions the ecology of caribou into factors for statistical analysis yet simplifies ecology. Many factors are involved and they interact. For example, weather affects condition of caribou, which in turn affects calf mortality and vulnerability to predation. Forestry operations increase access for wolves and hunters, cause fragmentation of caribou range, change predator-prey relationships, and influence the local climate. Major challenges this century are to learn more about how limiting factors interact, how to assess their cumulative effects, and how to mitigate effects.


Habitat Loss, Degradation, and Fragmentation

Caribou populations cannot exist without habitat of adequate quantity and quality. Loss, degradation, and fragmentation of habitat is caused by the cumulative effects of many factors both natural and of human origin. While predators cause most deaths of forest-dwelling caribou, and predation is of great concern (Table 7), it is a proximate factor that is influenced significantly by the effect of human developments. Access and disturbance, fragmentation (isolation), and low caribou numbers are of high concern (Table 7) and all are increasingly a result of development and human activities rather than natural causes. We first discuss some theoretical ecological considerations before listing major factors that reduce the amount and quality of caribou habitat.

Hunting and predation are thought to depress most caribou populations to densities well below the vegetative carrying capacity of their ranges (e.g., Bergerud 1974, 2000). However, periodic restricted availability of forage because of weather is a component of long-term carrying capacity. Therefore, a distinction must be made between absolute and relative forage availability. With the addition of other ungulates, predators, and habitat disturbance, it is more realistic to discuss ecological carrying capacity for caribou. It may be zero in marginal habitats and especially those considerably modified by human activities. Examples are ranges of local population in the South Purcell Mountains of B.C. and northern Idaho, each with less than 35 caribou. Adequate quantity of range can be assessed by considering minimum viable population size and ecological carrying capacity, including critical habitat components and predators. Peak caribou densities may be a crude index of ecological carrying capacity but environmental conditions should be specified and examined critically.

Adequate quality means that nutritious food must be available seasonally as well as calving and post-calving areas and other ‘security habitat.’ Certain important plant species may be in short supply or over-used even though total forage is superabundant. The effects of inadequate or degraded habitat, whether natural or human caused, may be subtle with lag effects (e.g., Messier et al. 1988). A slight decrease in reproduction or increase in mortality can result in population decreases. Erupting populations on new or renewed range are characterized by high pregnancy rates, whereas reproductive pauses (Dauphiné 1976, Cameron 1994) are common in populations that have intensively used range for decades or centuries. Such populations are believed to be below carrying capacity but the concept is largely theoretical and difficult to assess in the real world. It changes with weather, vegetation succession, disease, and other disturbances. It is not an absolute ceiling and is better described as “spongy” or buffered. Long before carrying capacity is reached or overreached there would be a range of density effects such as reduced fertility and higher mortality of calves. Also, density-independent factors such as ice on vegetation could cause density-dependent responses.

Successional changes after disturbance are poorly understood in spite of their importance. One reason is the large variation in succession related to cover type, soil conditions, disturbance characteristics, slope, aspect, elevation, and climate change. Succession after fire to adequate cover of terrestrial lichens preferred by caribou is protracted in Jasper National Park (Thomas and Armbruster 1996a), intermediate in the taiga (Thomas et al. 1998), and relatively rapid in central Saskatchewan (Thomas and Armbruster 1996b). 

A major paradigm shift in the last decade in forest operations is to emulate, as much as practicable, natural disturbance patterns as a means of preserving biodiversity. However, fire return intervals are highly variable and mean intervals between fires depend on what time period is selected. Therefore, there is no agreement on the fire cycle before humans attempted to control nature. More importantly, forest succession after fire differs significantly from that after logging. In particular, succession of lichens after logging depends on surface disturbance, surface treatment, and restocking methods. Prescriptions to generate good winter range for caribou must be tailored to several local site and stand conditions. Ecosystem management for diversity of all species (Seip 1998, Euler 1998) is preferable to small block designs but there must be special provisions for caribou.

Availability of habitat is of high or moderate concern for 32%, 74%, and 48%, for local populations in the NMP, SMP, and BP, respectively (Table 7). Fires can remove suitable habitat for 25 to 100 years or longer depending on fire intensity, geography, and type of forage normally consumed by caribou. Temporary loss of range from fire was a concern for 57%, 45%, and 76% of local populations in the COSEWIC populations listed above.

In Yukon, the spread of agriculture, forestry, and mining have affected caribou (Farnell et al. 1998). Caribou numbers have decreased significantly in southern B.C. from loss, alteration, and fragmentation of important habitat as a result of wildfire and human activities (Simpson et al. 1996, 1997). Management recommendations for the terrestrial feeding ecotype of caribou include maintaining some old forests, even-aged stands, and large harvest units and ‘leave’ areas (areas left for cutting in a future years) (Seip 1998). The mountain/arboreal ecotype of caribou requires a higher proportion of mature and old forests, uneven-aged stands, small cut blocks, and mature forest connectivity (Seip 1998).

Age of forest was the greatest determinant of habitat suitability in the Purcell Mountains of southern B.C. (Apps and Kinley 1998). Low elevations were first subject to developments such as forestry, roads, agriculture, homes, powerlines, pipelines, dams, and recreation. The result was loss and alteration of habitat used by caribou in late winter and spring, changes in predator-prey relationships (Seip 1992), and increased access. Increases in moose, which are more productive than caribou, resulted in greater predation on caribou (Seip and Cichowski 1996). Increased access resulted in increased recreation and more hunting of caribou. Now logging is proceeding at higher elevations and into range used by caribou in winter (Stevenson 1991).

Difficulties in managing caribou range in forest management areas in west central Alberta were chronicled by Hervieux et al. (1996). The Northeast Region Standing Committee on Woodland Caribou was established in Alberta in 1991 to provide an interactive base for those interested in conservation and development. The Alberta Woodland Caribou Conservation Strategy Development Committee was established in 1993 to ensure the survival of threatened caribou in Alberta (Alberta Woodland Caribou Conservation Strategy Development Committee 1996). Still, improved guidelines will have to be instituted in forests to conserve some local populations (Stevenson 1991, Hervieux et al. 1996, Rock 1992, Dzus 2001).

Government and industry generated six general principles for caribou conservation but plans for implementation are deficient. There are many overlapping activities in Alberta's caribou range including timber harvesting, oil and gas development, coal mining, and increasing road access (Hervieux et al. 1996). The cumulative effects of these disturbances are not understood and have received little attention (Edmonds 1998). Some effects can be mitigated but forest rotation cycles of only 60-100 years and only 15-20 years between cuts can remove most lichen-producing cover types unless there are special prescriptions. Smith et al. (2000) recommended that adequate usable and core winter range be retained for the current population along with minimal fragmentation.

In Saskatchewan, forestry operations, other human developments, and fire have fragmented large areas of forested land in the commercial forest. More deer, elk, and wolves are now found on caribou range. Suggested changes include zonal management of caribou and moose, altered moose management, a standard access policy, and involvement of Aboriginal people (Rock 1992). Additional strategies include encouragement of wolf harvest by trappers, management for low moose and deer densities, old-growth forest reserves, protection of lichens with winter harvesting, large-block timber harvest designs, and protection of calving areas and travel corridors (Godwin and Thorpe 2000). Many local populations are associated with fen complexes where timber values are low but adjacent commercial forest must be managed for caribou.

Woodland caribou disappeared from southern parts of their former range in Manitoba as agriculture expanded (Johnson 1993). Concerns include improved access to uncontrolled hunting, forestry, hydroelectric developments, and possible transmission of the meningeal worm from white-tailed deer to caribou (V. Crighton in Edmonds 1991, Rebizant et al. 2000).

In Ontario, recommendations for caribou conservation included large block forest cutting designs (Darby and Duquette 1986). Identification of distinct summer and winter range of one local population led to a suggestion that those ranges be managed for caribou (Cumming and Beange 1987) and unaltered by development (Cumming 1992). Loss of mature coniferous forests was recognized as a serious threat to caribou, especially where numbers of moose, deer, and wolves attain relatively high densities after logging (Darby et al. 1989). The southern limit of “continuous” caribou range closely approximates the northern limit of forestry operations in Ontario (Armstrong 1998). In light of caribou declines in the 1980s, Ontario reviewed woodland caribou status and ecology (Darby et al. 1989, Racey et al. 1991, Racey and Armstrong 1996, Armstrong 1998). Favourable winters resulted in a large increase in the deer population since lows in the 1970s. Therefore, weather not only affects caribou directly but also indirectly through changes in densities of other species. Forest management must consider eight “feature species” including woodland caribou. In the northwest, prescriptions to conserve caribou in commercial forests include large block mosaics and protection of winter range, calving habitat, and travel corridors (Racey et al. 1991; Racey and Armstrong 1996, 2000; Armstrong 1998). Racey and Armstrong (1996) listed 12 points in a caribou management strategy for northwestern Ontario.

Clear-cutting mature forests on summer and winter range affects the distribution of caribou (Chubbs et al. 1993, Smith et al. 2000). Overgrazing of forage leading to population declines could become a problem in Newfoundland and an objective is to prevent declines by managing hunting (Mahoney and Schaefer 1996).

The Gaspésie population decline was attributed to habitat loss (agriculture and logging), predation, and hunting (RENEW 1993). Woodland caribou were extirpated in several states of the USA south of 49o with the exception of a few caribou in the Selkirk Mountains in Idaho (Zager et al. 1996).

The maximum habitat change tolerable by forest-dwelling caribou will depend on minimum viable population size, the area and quality of the habitat mosaic that is sustained, the ability of caribou to accommodate human activities, management of predators, and little or no hunting. Reindeer in Scandinavia, in the virtual absence of predators and hunting, can persist in highly developed areas. However, wild reindeer in Norway avoid developments (Nellemann et al. 2001, Vistnes et al. 2001).


Disturbance

Disturbance can mean habitat disturbance or individual caribou being alarmed by some stimulus. Habitat disturbances related to developments were discussed in the previous section. The distinction between natural and human-related disturbance becomes blurred in the case of climatic warming, fire protection, disease and parasites in plantations, and salvage logging of burned areas. Natural disturbances include wildfire, insects and diseases that kill trees, tornadoes and windstorms, extreme weather, predators, and parasitic insects. All those factors can cause caribou to move. Here the emphasis is on the direct effects of human activities on individual caribou rather than indirect effects mediated through habitat changes.

Several examples were noted in previous sections of the effects of human activities on caribou and reindeer (Trottier 1988b, Rock 1992, Chubbs et al. 1993, Bradshaw et al. 1997, 1998, Dyer 1999, James 1999, James and Stuart-Smith 2000, Smith et al. 2000, Dyer et al. 2001, Nellemann et al. 2001, Vistnes et al. 2001). There are three well-documented cases of caribou being displaced by logging activities in Ontario (Darby and Duquette 1986, Darby et al. 1989). Construction of a hydroelectric development in Newfoundland displaced some caribou and disrupted the timing of migration (Mahoney and Schaefer 2001). Forest-dwelling caribou are sensitive to disturbance but the cause of their withdrawal from human activity is not clear. They naturally are fearful of unusual activity within their range and avoidance often has survival value. Caribou may associate linear and other developments with predators and hunters.

In Alberta, the A la Pêche caribou population lost 16% to 21% of its members in 1991-92 from collisions with vehicles on Highway 40 north of Hinton (Brown and Hobson 1998). The following year it was 14% to 18%. Broadcast salt attracted caribou to the highway. Mortality was reduced when caribou were scared off the right-of-way by volunteers on snowmobiles. Since 1997, that population has remained in the mountains (Dzus 2001), perhaps as a response to disturbance along the highway. Hauling logs by truck through winter habitat in Ontario displaced some caribou and caused others to move out of the area (Cumming and Hyer 1998).

Studies were conducted in Labrador on the short-term impacts of low-level jet aircraft flights on caribou populations (George River and Red Wine Mountain). Low-level flying did not significantly affect daily activity of caribou or travel distance, though comparison with ‘control’ caribou suggested potential effects (Harrington and Veitch 1991). Searches by helicopter for radio-collared caribou in a side valley of Jasper National Park caused two caribou to leave the area immediately and others to depart over the next few hours (Thomas and Armbruster 1996a).

There is an urgent need for more studies to assess the effects of human activities on individual caribou and on populations. How long does it take caribou to accommodate to various stressors? No accommodation to linear developments was detected within 2 years in Alberta (Dzus 2001). Experimental studies are needed in which adequate numbers of marked caribou (collars or paint) are subjected to specific activities and their reproduction and mortality compared with that of undisturbed marked caribou. Empirical observations and experience will be invaluable, including those from Aboriginal peoples.


Predation

Predators influence caribou distribution and limit densities (Bergerud 1974, 1978, 1980, 1983, 1996, 2000; Edmonds 1988, Seip 1992, Brown et al. 1994, Crête et al. 1994, Boertje et al. 1996, Seip and Cichowski 1996, Stuart-Smith et al. 1997, Bergerud and Elliott 1998, Rettie and Messier 1998, James 1999, Schaefer et al. 1999, James and Stuart-Smith 2000). In a classic study, recruitment of caribou increased 113% and adult mortality decreased 60% when wolf numbers were reduced 80% on the range of the Finlayson population in Yukon (Farnell and McDonald 1986). Removal of 60-90% of wolves over three winters increased recruitment of the Horseranch caribou population to 16.7% from 5.5% (Bergerud and Elliott 1998).

Wolves may reduce or even eliminate caribou populations in areas where habitat has been significantly altered. Increased abundance of other large prey species is one factor. In Yukon, wolves limited densities of moose to 7-12/100 km2, with each wolf consuming an average of 2.4 moose per 100 days (Hayes and Harestad 2000). Caribou prosper in areas where other ungulates and wolves are absent or rare. Alpine habitat enables caribou to reduce contact with wolves. In spring and summer, wolves spend much of their time at low elevation near den sites and numerous large prey (Bergerud 1978, Edmonds and Smith 1991, Seip 1992, Brown et al. 1994, Farnell et al. 1996, Simpson et al. 1997).

On caribou range in central Saskatchewan, moose densities are low and wolves also prey on deer, elk, and caribou. Black bears (Ursus americanus) may be a significant predator of caribou calves. More predators combined with increased access to caribou by wolves, coyotes, and humans can combine to cause declines. Increases in just one prey species, moose, was thought to have caused declines in caribou populations in B.C. (Seip 1992) and possibly in southern Labrador (Schaefer et al. 1999). In contrast, caribou and introduced moose populations have grown and expanded in Newfoundland, where wolves no longer occur.

Adverse climate combined with hunting in the late 19th century seemed to drive down numbers of ungulates and predators. Moose expanded out from refugia in western and eastern Canada. People in western Canada still remember when they saw their first moose and white-tailed deer. Wolves and coyotes were poisoned in the 1950s and early 1960s (Cringan 1957, Bergerud 1978, Edmonds and Bloomfield 1984, Edmonds 1988, Rock 1992, Bergerud and Elliott 1998). Some relatively large populations of caribou were recorded in the late 1960s and early 1970s. Large legal harvests of caribou at that time in B.C., Alberta, and Saskatchewan (Bergerud 1978, Edmonds and Bloomfield 1984, Rock 1992) combined with recovering wolf populations and adverse weather probably caused caribou population decreases in the 1970s. The caribou population highs in the 1960s probably were atypical and should not be considered management objectives.

Public and scientific attitudes to predator control changed markedly during the 20th century. Attitudes to hunting and predators combined with firearm regulations mean that the ability of wildlife agencies to manage densities of ungulates and predators is declining. An interesting perspective is that temporary lethal control of wolves may be counterproductive as it could generate a rebound effect (Valkenburg et al. 1996, Bergerud and Elliott 1998).

Predation on forest-dwelling caribou by wolves is essentially incidental to the wolf-prey system because low-density caribou populations cannot sustain wolves. Each adult wolf requires about 29 adult caribou annually (Hayes and Russell 1998). A “capital” of about 200 adult caribou is needed to sustain each wolf feeding entirely on caribou, assuming wolves account for all of a 15% average annual mortality of adult caribou. Therefore, a pack of five wolves requires a population of 1000 caribou for sustainability of both species, assuming no other prey. In reality, there are other sources of food for wolves and other forms of caribou mortality.

To sustain densities of 2, 4, and 8 wolves per 1000 km2 would require sustained populations of 387, 773, and 1547 caribou in that area to feed each wolf if caribou were their only prey and wolves accounted for all caribou mortality. However, most densities of forest-dwelling caribou are in the range of 10-200 caribou per 1000 km2 (Table 6). Thus, caribou can only form a small part of the diet of wolves in the forest and other prey such as moose, deer, and beaver (Castor canadensis) must form the bulk of their diet. Wolves are likely to concentrate their predation on species with high productivity and biomass. Caribou tend to use areas where moose and deer are absent or rare (Cumming and Hyer 1998), especially in summer when calves are vulnerable to predation. Examples are alpine and subalpine areas, islands, peatlands, and shrub-poor pine forests. Lakeshores provide forage and caribou can escape wolves by swimming.

That predation of caribou by wolves is incidental is supported by density considerations. Wolf densities in the southern boreal forest of Ontario were in the range of 4-8/1000 km2 and half of that in northern parts of the forest (Darby et al. 1989). Bergerud (1988) stated that wolf densities higher than 6.5/1000 km2 caused caribou to decline whereas moose could persist at densities of 8 wolves per 1000 km2 (Bergerud and Elliott 1998). In the southern Yukon, wolf densities typically were 8-10/1000 km2 (Hayes et al. 1991). They were 9-10/1000 km2 in the area of the Wolf Lake caribou population when it grew at an average annual rate of 10.5% (Farnell et al. 1996). Wolf control for 7 years in Alaska resulted in caribou population growth for 14 years -- from 2 200 to 10 690 -- followed by a decline to 3 660 because of predation and deep snow (Boertje et al. 1996). The moose populations grew over 19 years from 2 500 to 13 800. The wolf population decreased from 239 to 80-157 during control, recovered in 4 years to 195 and peaked at 267 in the third year of 4 years of weather adverse to moose. Caribou numbers in the Delta population increased when wolf densities were 11-12 per 1000 km2 (Boertje et al. 1996). These few examples reveal that predator-prey relationships are extremely complex and generalizations are rife with exceptions.

Predator-prey relationships are not only complex but also dynamic. They involve the distribution and relative densities of multiple prey from moose to snowshoe hare (Lepus americanus) and multiple predators from grizzly bears (Ursus arctos) to coyotes. Growth of one caribou population while the other declined was attributed to differential wolf predation on moose at different densities, hare densities, skewed sex ratios, and hunting (Farnell et al. 1996).

There is speculation that caribou actively maintain low densities (e.g., Bergerud 1992) but more likely it is a passive phenomenon linked to predation rate. Populations will grow at their innate biological capacity unless limited by environmental conditions. High densities of caribou in Newfoundland, where wolves are absent, contradict the hypothesis that caribou self-regulate to low densities. Low densities are more likely a consequence of predation and other limitations rather than an adaptation by caribou.

In Newfoundland, lynx predation limited caribou until lynx numbers declined after fur prices increased and snowmobiles provided access (Bergerud 1971, 1974, 1980, 2000). In the Corner Brook Lake area, there was significant predation on caribou by black bears (Snow and Mahoney 1996). If the wolf was to re-enter Newfoundland after 7-8 decades, it could have a devastating effect on caribou populations because of a dense moose population and naive, sedentary caribou. Observations of calf-hiding behaviour by woodland caribou in east central Newfoundland are the first reported (Chubbs et al. 1993).

Between 1987 and 1993, the mortality rate of calves in summer in the Gaspésie population approached 90%. Predation by black bears and coyotes, present in Gaspésie only since the early 1980s, was responsible for the high rate of mortality (RENEW 1993, 1994). The likely cause of death for 11 of 16 radio-collared calves monitored in 1989 and 1990 was predation by coyotes (7 deaths), black bears (3 deaths), and golden eagle (Aquilachrysaetos) (1 death) (Crête and Desrosiers 1995). Between 1990 and 1992, the Gaspésie Caribou Recovery Team removed 70 coyotes and 37 black bears from the Park (Crête and Desrosiers 1995). By 1992, calf survival improved. Human access in the Mont McGerrigle area was restricted to lessen the potential of caribou fleeing to forest cover, where calves were more vulnerable to coyotes (RENEW 1994).

In Idaho and southern B.C., predation by mountain lions is an important limiting factor. That species could become more numerous in other parts of the range of forest-dwelling caribou if other ungulates increase in number and the climate warms. There are few data on the relative vulnerability of ungulates to various predators (Dale et al. 1995, Thomas 1995). Caribou in some locations of the Cordilleran Mountains are subject to predation by grizzly bears, black bears, wolves, coyotes, mountain lions, wolverine (Gulo gulo), lynx, and golden eagles. The additive effects of mortality from so many predator species must severely limit caribou populations. 

Because predation rate can be measured, its influence can be oversold unless one is vigilant to separate proximate from ultimate factors and consider ecological relationships and interaction of limiting factors. For example, predation may become a problem because of habitat fragmentation and alteration, creating abundant food and niches for other ungulate species. Managing ungulates at high populations for hunting increases predator numbers. Studies in Alberta confirm empirical observations that wolves tend to travel roads and other linear corridors through caribou habitat. The same applies to coyotes (Canis latrans). Peatlands afford caribou relative security from wolves, which tend to frequent dry ground (Bradshaw et al. 1995, James 1999, James and Stuart-Smith 2000). Rate of travel by wolves in winter was 2.8 times faster in linear corridors than in the forest (James 1999). However, wolves travel rapidly through peatlands in spring when an icy crust forms on the snow surface.


Weather

Weather (short term) and climate (long term) are known to be the most important limiting factors for caribou in the High Arctic (Miller 1990, Gunn et al. 2000). Adverse weather can suspend reproduction for up to 3 years (Thomas 1982) and cause mortality of calves (Miller 1974) and adults (Adamczewski et al. 1993, Miller 1990). Weather is also a significant factor in the Arctic and subarctic (Edmonds and Smith 1991, Boertje et al. 1996, Adams and Dale 1998, Finstad and Prichard 2000). A large and growing literature indicates that weather affects all aspects of the ecology of caribou. However, the effects of unfavourable weather on caribou often are indirect and subtle – a small decline in pregnancy rate and survival of calves (Adams et al. 1995, Boertje et al. 1996) and adults through increased vulnerability to wolf predation (Adams et al. 1995, Valkenburg et al. 1996). Weather and demographic variables typically are not measured with adequate sample sizes over sufficient time to establish a significant correlation. Establishing cause and effect is even more demanding. Only extreme weather that results in large changes in demographics is recorded. For example, decreases in numbers of woodland caribou in Saskatchewan were associated with deep snow in 1971-72 and 1973-74, combined with unsustainable hunting and predation (Rock 1992). Caribou can forage through up to a metre of snow or more (Brown and Theberge 1990) but at a cost to energy reserves.

The effect of severe winters may be reduced survival among calves (Miller 1974) but such relationships are difficult to detect because of confounding factors such as predation. Studies of reindeer in Alaska clearly indicate effects of weather because large samples are available and there is some control of confounding factors. Weather affected forage quality, which influenced growth and age of first conceptions in reindeer (R. t. tarandus) (Finstad and Prichard 2000). Shallow snow followed by warm weather in May and June and cool temperatures in July were conditions that resulted in calves becoming pregnant. There may be lag effects that dictate that several seasons of weather data and caribou demographics must be evaluated. 

Calf survival in caribou and reindeer is related to weight at birth (Boertje et al. 1996). Birth weight, in turn, is related to reproductive history the previous year and availability of food in winter (e.g., Adams and Dale 1998) and particularly in the last trimester of pregnancy. Gestation is delayed in undernourished females (Cameron et al. 1993, Bergerud 2000). Adaptation to a compressed period of breeding and births (Dauphiné 1976) implies that survival is highest at the peak of calving in most years. Selection for an average optimal time for calving involves multiple factors with energetics and predation being primary. Therefore, weather can have subtle yet significant biological effects that often go undetected.

Forest-dwelling caribou have adapted to a wide range of climate – from areas of high precipitation in mountains of southern B.C. to relatively dry conditions on the central plains of western Canada. Vegetation composition varies primarily with climate and caribou have adapted their winter feeding behavior accordingly. Small populations on the southern periphery of the range are vulnerable to climatic warming and greater weather variability. Detrimental effects could include a greater temporary range loss from fire, freezing rain, thaw periods in winter, deep snow over terrestrial lichens, loss of snowbanks in summer, hot weather in early summer, and changes to food supplies. They were in marginal habitat before climatic change and industrial activity.

Some populations that were protected from development, such as one in south Jasper National Park, declined from the 1960s to the early 1990s (Stelfox et al. 1978, Brown et al. 1994). However, the relative roles of weather, predation, development, and their interactions could not be partitioned (Brown et al. 1994, Thomas and Armbruster 1996a). Snow conditions affected predation by forcing caribou to winter in valley bottoms among other ungulates and wolves (Brown et al. 1994). An example of interaction of variables was increased predation of caribou when spring migration was delayed because of deep snow (Edmonds and Smith 1991). Adverse weather for 4 years, combined with high wolf densities, caused the Delta population in Alaska to decline from 10 690 to 3 660 (Boertje et al. 1996).

Another example of linked variables is the effect of weather on parasites. Weather affects the abundance of Elaphostrongylus rangiferi, which caused a near 3-fold decline in the Avalon population in Newfoundland (Ball et al. 2001). Weather affects all other variables or factors that limit caribou populations. A concern related to climate warming is drying of peatlands, which would increase their probability of burning, would affect food supplies of caribou, and would increase access by predators. Drying of peatlands was noted in the 1980s and early 1990s in Saskatchewan (T. Rock pers. comm. 1995). Dry conditions culminated in many large fires in eastern Alberta and Saskatchewan in 1995 and 2002. Draining peatlands to improve tree growth and extraction of peat are concerns. Mushroom gathering in lichen-rich pine forests is potentially detrimental to caribou.


Hunting

Hunting is implicated in many declines in caribou (Kelsall 1968, Bergerud 1974, 1978). Nevertheless, obtaining accurate harvest data continues to be a major information deficiency. Hunting is generally considered additive to other limiting factors and therefore any reduction in hunting mortality is beneficial to a declining caribou population. Caribou populations in large undisturbed areas, where predators are not managed, can withstand only 2-3% annual mortality from hunting (Yukon Renewable Resources 1996). The huntable proportion is zero in marginal habitats modified by multiple predators and human developments. Hunting mature males will not harm a population provided that an adequate proportion survives to breed females.

Caribou declines in B.C. were investigated in the 1970s (Bergerud 1978) and hunting subsequently was reduced or curtailed in some populations. There are limited entry and open seasons for bulls with at least five tines on one antler in the north central and west central metapopulations. Limited hunting of “mountain” caribou in the Kootenay region lasted until 1996. In 2001, no recreational hunting was permitted in the southern metapopulation.

Recreational hunting of forest-dwelling woodland caribou was closed in Ontario, Alberta, Saskatchewan, and Manitoba in 1929, 1981, 1987, and 1992, respectively (Table 9). Hunting permits are issued for the Cape Churchill population of forest-tundra caribou and, in 1997-98, 178 caribou were killed (Elliott 1998). The Pen Island population of forest tundra caribou also is hunted. In the 1980s, an estimated 600 to 700 caribou were harvested annually under Treaty rights in all of Ontario (Darby et al. 1989). No current figures are available. Uncontrolled hunting complicates assessment and management of woodland caribou in Ontario (Harris 1999) and all other jurisdictions.

Forest-dwelling (sedentary) caribou are hunted in northern parts of their general distribution in Quebec. There are limited entry hunts for residents with (zone 23S) and without (zone 22A) outfitters or outfitted hunts for residents and non-residents (zone 22B). Limits are two caribou per hunter (FAPAQ 2002). In winter, areas occupied by sedentary caribou are invaded by the Leaf River and George River populations so the proportionate kill of each ecotype is unknown. Caribou hunting is not permitted in any of Quebec's wildlife reserves (reserves fauniques), hunting reserves, or parks. In 1981, the Grands-Jardins Conservation Park, an area of 310 km2, was created to conserve an important part of caribou habitat in the Charlevoix region (Banville 1998). Caribou hunting was banned in the Gaspésie in 1937 (Boileau 1996). 

Growth of the Avalon population from 125 in 1956 to 2500-3000 in 1979 was attributed to reduction of poaching (Bergerud 1980). Hunting of caribou in Newfoundland is promoted as part of the sports and tourism industry and to manage population growth. Licensed guides must accompany non-resident hunters.

Unregulated hunting was a high or moderate concern for 70% of local populations in the NMP, 30% in the SMP, and 42% in the BP exclusive of Ontario and Quebec (Table 7). In Yukon, the average annual harvest by licensed hunters declined from over 300 in the 1980s to 271 in the 1990s. Harvest has been restricted to bulls since 1984 and six populations were closed to hunting. Hunting by First Nations is suspected to equal that of licensed hunting (Farnell et al. 1998).

In the NWT, harvest of the South Nahanni and Redstone populations may not be sustainable (Adamczewski and Veitch 1998). In 1996, Mackenzie Mountain outfitters reported a legal kill of 172 bulls. There is no closed season or limit for holders of a General Hunting Licence in the NWT. Resident and non-resident hunters are permitted one caribou in specific hunting areas south of 68o N (Gray and Panegyuk 1989).


Parasites

Insects are potentially a major limiting factor for caribou. Effects include parasite and disease transmission, harassment, loss of blood, and immune system reactions. Important insects include warble flies (Oedemagena spp.), nose bot flies (Cephenomya trompe), mosquitoes (Aedes spp.), black flies (Simulium spp.), horseflies (Tabanus spp.), and deer flies (Chrysops spp.). Kelsall (1975) noted that the highest average counts of warble fly larvae in infected caribou were in the mountains of Alberta, B.C., and Yukon. Use of snowbanks by caribou in summer is likely to be a response to insect harassment. The severity of insect harassment is weather related and observed climatic warming could add to the problem. Summer behaviour of caribou is greatly influenced by actions to reduce exposure to insects and insect-borne parasites. The effect of insect harassment on the fitness of forest-dwelling caribou is unknown. Physical condition and pregnancy incidence was inversely related to numbers of warble larvae in female barren-ground caribou >2 years old (Thomas and Kiliaan 1990). 

Also poorly known are the incidence and prevalence of internal parasites and their effects. Mature and old woodland caribou are likely to have a relatively high incidence and prevalence of hydatid cysts (Echinococcus granulosis). The adult tapeworm resides in canids and it cycles through snails, moose, and caribou. A large number of large cysts in the lungs could make a caribou susceptible to predators, thus completing the cycle.

The protostrongylid nematode, Parelaphostrongylus andersoni, is widely distributed in woodland and barren-ground caribou of mainland Canada (Lankester and Hauta 1989). Though it may not cause neurologic disease in wild cervids, its eggs and larvae develop in the lungs and an inflammatory reaction contributes to verminous pneumonia. A related meningeal nematode, P. tenuis, causes neurologic disease in cervids, including caribou. That parasite, benign in white-tailed deer, is potentially a limiting factor for woodland caribou (Pitt and Jordan 1994) and may be a threat to caribou in southern Ontario and west to Saskatchewan. Meningeal worms may be artificially spread in western Canada through game ranching (Samuel et al. 1992). A protostrongylid nematode introduced with reindeer from Norway, Elaphostrongylus rangiferi, has become established in Newfoundland caribou (Lankester and Fong 1998). It does not cause neurologic disease but it can induce pneumonia in young calves (Ball et al. 2001) and was implicated in a decline in the Avalon population from 7000 to 2500 individuals over 3 years (Lankester and Fong 1998).


Other Limiting Factors

Accidents always account for a small proportion of deaths though avalanches accounted for most deaths in one study of the Mount Revelstoke population in B.C. (Simpson et al. 1985). Caribou are excellent swimmers but drownings occur at rapids, when crossing lakes in rough water, and when falling through thin ice.  

Recreational activities such as snowmobiling, boating, horseback riding, hiking particularly with dogs, and hunting modify the distribution of caribou with unknown effects.