During the last decade, losses of semi-domestic reindeer Rangifer tarandus have increased in central Norway. Natural mortality in a semi-domestic reindeer herd was studied by use of mortality sensing transmitters. From 15 April 1995 to 15 April 1996,135 of 612 animals equipped with radio collars were found dead. Adult females (>2 years old) suffered a mortality of 18.3%, of which 40.5% was due to predation. Yearlings suffered a yearly mortality of 20.2% of which 66.7% was due to predation. Calf mortality from 6 August 1995 to 15 April 1996 was 31.0%, of which 75.3% was due to predation. Predation by Eurasian lynx Lynx lynx was the predominant cause of death, with 39.3% of the total mortality. Accidents were the second most important cause of death, with 16.0% of the total mortality. Peaks in mortality were registered in September, November and January.
Losses of semi-domestic reindeer Rangifer tarandus increased in Norway through the 1980s and early 1990s. Predation is suspected to be the primary cause of mortality in wild reindeer and caribou (Bergerud 1980, Filonov 1980, Miller, Broughton & Gunn 1988, Hearn & Luttich 1989, Bjärvall, Franzén, Nordkvist & Åhman 1990, Mahoney, Abbott, Russell & Porter 1990, Whitten, Gamer, Mauer & Harris 1992, Adams, Singer, & Dale 1995). Although reindeer owners have always recognised that carnivores inflict losses on their herds, quantitative studies are lacking (Mysterud & Mysterud 1995). ‘Unknown cause’ is by far the most used label on carcasses of lost animals in reindeer husbandry. The aim of this study was to identify, date and quantify the causes of mortality in a semi-domestic reindeer herd in an area in central Norway known to hold a high density of large carnivores, especially the Eurasian lynx Lynx lynx.
The study took place in the northwestern part of the county of Nord-Trøndelag, situated in central Norway between 64°35′-65°00′N and 12°00′-13°00′E. The study area is mountainous with peaks reaching up to 1,100 m a.s.l. The treeline is found at 300.–400 m a.s.l. in coastal areas and at 400.–500 m a.s.l. inland. Boreal spruce forest, bogs and alpine plants dominate the vegetation. Norway spruce Picea abies occurs up to the treeline. In the subalpine area Scots pine Pinus sylvestris occur together with ericaceous heather and lichens. In 1969, Lyftingsmo (1974) found that parts of the area had good grazing with favourable snow conditions and lichens growing abundantly among heather and on bogs. The large bogs in the subalpine areas also provided early green grazing. The winter grazing area had a carrying capacity of 6.25 reindeer per km2 in 1969 (Lyftingsmo 1974). Investigations of reindeer pastures have not been given priority in this area in recent years due to the obvious needs in Finnmark, where overgrazing, in contrast to the situation in central Norway, constitutes a severe problem in reindeer husbandry.
The period of snow cover varies depending upon elevation and distance from the sea. In winter 1995/96 the first snow came on 1 November at 400 m a.s.l., but at the coast it melted away several times in November and December. At altitudes above 800 m the snow came in October and did not melt until May. The snow cover that winter was about 70.–90% of the normal snow cover in the area. The normal annual precipitation is 1,137 mm ranging within 934–1,340 mm (Norwegian Meteorological Institute, pers. comm.). The normal temperature in the area ranges from -7.2°C in January to 11.9°C in July (DNMI, pers. comm.).
Lynx, wolverine Gulo gulo and golden eagle Aquila chrysaetos are present in the area. Brown bear Ursus arctos irregularly roam the area, and a wolf Canis lupus was last seen in 1986. Moose Alces alces and roe deer Capreolus capreolus are common below the treeline.
Central Norway is the southernmost part of the reindeer herding area in Norway. Reindeer husbandry is the basis of the ancient culture of the Saami, who have status as aboriginal citizens in accordance with ILO-convention no 169. The owners guard the semi-domestic reindeer with varying intensity all year round. Three to four times a year the reindeer are fenced up for marking and slaughtering, and on these occasions weighing, radio-collaring and counting can be carried out.
Material and methods
During 15 April 1995 - 15 May 1996,612 reindeer (including 323 calves and 54 yearlings (of both sexes) and 235 adult females) were fitted with mortality-sensing radio collars (TELEVILT®) operating on the 142 MHz band evenly distributed on six different frequencies. This kind of transmitter remains silent as long as the bearer is active, but emits a signal after the animal has remained motionless for about two hours.
Reindeer were classified by sex and age as calves, both sexes (< 1 year old), yearlings, both sexes (1–2 years old), and adult females (>2 years old), and were weighed and ear-tagged during routine seasonal roundups of the herd. Non-calf reindeer were weighed in box scales, and calves were held and weighed on bathroom scales (weight of person with reindeer calf minus weight of person). Calves were ear-tagged and radiocollared in August only.
Radio-collared reindeer were monitored daily from the ground, and once a week from aircraft. All carcasses were closely examined using standard autopsy methods, and the sites at which carcasses were found were searched for clues as to the cause of death, including signs of predators such as tracks, scats and hair. The cause of death was classified as 1) killed by lynx, 2) killed by other predators (wolverine, golden eagle, or undetermined predator), 3) ‘other death causes’ (accidents or disease) and 4) ‘unknown’.
The ages of dead reindeer were determined by tooth development for calves and yearlings, and by counts of cementum annulations in reindeer older than two years (Reimers & Nordby 1968).
As reindeer were radio-collared on different dates throughout the year, ‘the product/limit method’ (Kaplan & Meier 1958, Cox & Oakes 1984) modified by Pollock, Winterstein & Conroy (1989) was used to calculate survival rates for radio-collared reindeer from 15 April 1995 to 15 April 1996. Annual survival rates and ‘reindeer-days’ (one ‘reindeer-day’ = one radio-collared reindeer out for one day) were calculated using the computer programme ‘Kaplan-Meier survivalship analysis ver. 1.0’ (Pollock et al. 1989). Mortality rates distributed by cause of deaths were calculated by means of the computer program ‘Micromort ver. 1.3’ (Heisey & Fuller 1985). Differences between survival rates were tested by use of ‘Contrast’ (Sauer & Williams 1989). All other statistical tests were carried out by use of SPSS ver. 7.0 for Windows. The data were examined for statistical significance at a critical level of 5%.
Of the 612 radio-collared reindeer. 135 were found dead during 15 April 1995 - 15 April 1996. Predation accounted for 65.2% (N = 88) of the total reindeer mortality. The recorded mortality was not evenly distributed over the year. Predation by lynx peaked in September, November and January and was lowest in October and December. Accidents peaked in January (Table 1). All accidents except two cases of collisions with cars happened during periods with snow cover.
Annual survival rates and monthly survival of radio-collared reindeer in central Norway during 15 April 1995 - 15 April 1996.
Adult females and yearlings
Adult females (>2 years old) suffered a mortality of 18.3% (see Table 1). For adult females, accidents were the most important cause of death with 45.9% of the total mortality. Predation accounted for 40.5% of the mortality.
Yearlings suffered a yearly mortality of 20.2% of which 66.7% was due to predation (see Table 1). Annual survival rates for the two age classes (yearlings and adult females) of reindeer ≥l year old were not significantly different (X2 = 1.433, df = 2, P = 0.488), and pooling these groups together as ‘reindeer ≥l year old’ resulted in an annual survival rate of 82.3%. Predation was the cause of 48.2% of the total mortality in this group. Of 10 documented cases of predation on reindeer by golden eagle, three were adult females 5,7 and 9 years of age. The 5- and 7-year-old reindeer weighed 65 and 71 kg, respectively. The mean weight in December of adult females in the herd was 66.5 kg (mean = 66.5 kg, SD = 7.5, N = 484).
Of 323 radio-collared reindeer calves, collared after 6 August 1995, 89 died during the research period. Calf mortality from 6 August 1995 to 15 April 1996 was 31.0%, of which 75.3% was due to predation (see Table 1). The data on calves were tested for sex-specific predation risk, but no indication was found that predators showed any preference for a particular sex (X2 = 0.009, df = 1, P = 0.923). In the period 16 August 1995 - 15 April 1996 survival rates of calves and adults were significantly different (X2 = 11.521, df = 1, P < 0.001). In the same period lynx killed more calves (15.6%) and yearlings (7.9%) than adult females (1.7%).
For a considerable proportion of the autopsied carcasses the cause of death could not be determined (‘unknown cause of death’). There is no reason to believe that these cases should be distributed differently on real causes of death than those with known causes of death. The distribution of causes of death should accordingly be based on the number of cases with known causes of death, and not on the total number of radiocollared carcasses.
Of the total mortality in calves, 89.3% of the cases with known causes of death was due to predation, and predation by lynx was the dominant cause of death reported in our study. Only for adult females did ‘other causes’ account for more deaths than lynx predation. Lynx accounted for at least 53 of 135 lost reindeer, which is 42.4% of cases in which the cause of death was known, and the wolverine accounted for 6.9% of the total mortality in our study.
The main cause for the high reindeer mortality record-ed in our study was an inflated lynx population in North-Trøndelag. Based on hunting statistics the population size more than doubled from 1990 to 1996, when 5.3 lynx per 1,000 km2 was estimated (Kvam 1997). Recent studies report reindeer to constitute 28.1% of the lynx diet in the area, while roe deer constitutes 31% (Sunde & Kvam 1997). Telemetry studies indicate that lynx in the study area tend to kill one medium sized ungulate (reindeer or roe deer plus sheep in summer) every 4–6 days (Kvam, Sunde & Overskaug 1998, Sunde, Kvam, Bolstad & Bronndal 2000).
The wolverine is often dependent on favourable snow conditions to be able to kill reindeer (Haglund 1966). Four of 10 reindeer documented to have been killed by a wolverine were adults. The biggest adult female reindeer killed by a wolverine weighed 73 kg in December 1995. As the lynx tends to leave a larger part of its prey uneaten, it is probable that the numerous lynx population has made it possible for the wolverine to benefit from scavenging lynx prey.
The design of the present study was similar to that of the four-year programme monitoring reindeer calf mortality in Sweden in the 1980s (Bjärvall et al. 1990). The total mortality recorded in our study was higher (31.0%) than the total mortality (14.3%) recorded in Umbyn, Sweden. Predation accounted for a higher part of the total mortality recorded in our study (75.3%) than in the study in Sweden (65%). Caribou herds exposed to predation may lose 50% of the annual calf crop before calves reach 6 months of age (Bergerud 1980), and predation has been reported to constitute up to 93% of the total annual mortality in calves (Mahoney et al. 1990). The annual mortality recorded in our study is, nevertheless, higher than earlier reported from Scandinavia, and supports the hypothesis that predation is a key element in the mortality regime of semi-domestic reindeer in central Norway.
Other causes of death
Besides predation, accidents were the most important cause of death in our study, and were the primary cause of death for adult reindeer. Most accidental deaths were due to falls off cliffs, avalanches, or getting trapped in scree. It is currently discussed to what extent reindeer falling off cliffs were chased by predators. If this were the case one might expect the dead reindeer to be eaten by the predator. In most accident cases, however, the carcasses were ‘intact’ when found. Thus nothing in our study seems to indicate that predators affected the number of accidents happening. We often observed that reindeer which were later lost in accidents were foraging on foliose lichens in the scree or near the rock crack where they later got stuck. With no chance of getting up again they starved to death there before they were located by radio telemetry. The reindeer tended to forage in steep areas in periods with deep snow, as the snow cover is not so deep there and it is easier to find food. In our investigation, accidents and disease constituted 8.1% annual mortality in adult female reindeer. Losses of this size could be expected in a hypothetic situation with no predation. The recorded mortality rate for calves (31.0%) obviously underestimates annual mortality because some calves must have died before radio- collaring was carried out. Calves were collared on 6 August, and normal parturition of reindeer in the study area is in May (Ozaga & Clute 1988, Whitten et al. 1992).
The mortality peaked in September, November and January. Herd structure and habitat use of the reindeer vary throughout the year. During the calving season in May and through the summer season reindeer go up to higher parts of their range, normally above the treeline. In the calving season the reason may be to avoid predators, such as the lynx, which is the predominant predator species in central Norway. The lynx prefers forest areas in lower altitudes rather than open alpine areas (Moa, Negård & Kvam 1998). In autumn reindeer spread out and migrate to lower elevations to search for mushrooms. This means that habitat use by lynx and reindeer overlaps in autumn more than in summer, thus causing increased predation.
As many as 52% of the accidents and 34% of the lynx kills were recorded during 16 December – 15 February, which is the period with the deepest snow cover in the area. Many of the accidents were the result of reindeer falling off precipices, getting trapped in screes, or being taken by avalanches. The deep snow may also partly explain the high losses due to lynx predation in January, since the deep snow may have led to favourable hunting conditions for the lynx.
Age and sex
Although adult female reindeer weighing up to 74 kg, which is more than three times the body weight of the lynx, were recorded to have been killed by lynx, calves seem to be the lynx's favourite prey. Lynx, wolverine and golden eagle are relatively small predators, and it seems reasonable for them to prefer calves to adults.
Biased sex ratios in adult reindeer have been reported from a number of studies (Bergerud 1980, Reimers 1977, Parker 1972), and are assumed to be a result of predation and intraspecific strife. Male calves are supposed to attract more attention from predators than females due to their more independent and ‘careless’ behaviour. However, no selection for male calves was found in our study.
Research and management implications
Although some risks are connected with generalising the results of this one-year investigation, our results are in agreement with those of earlier comparable studies. The economic consequences for reindeer husbandry in an area with predation rates as high as those recorded in our study makes it questionable to what extent reindeer husbandry in its present form can be combined with legislation and international treaties which demand that large carnivores be maintained in viable populations.
By ratification of the Bern Convention, Norway became responsible for survival of large carnivore species in viable populations. In 1990 the ratification of ILO-Convention no 169 made Norway responsible for enforcing aboriginal citizens' right to their own culture. The Saami people have status as aboriginal citizens, and have based their culture on reindeer husbandry for centuries. This leads to a conflict between the two conventions. The protected large carnivores threaten Saami reindeer husbandry, which constitutes the framework of an aboriginal culture. It is therefore a management challenge to find large carnivore population sizes that keep the predation rate of semi-domestic reindeer at an acceptable level, and at the same time maintain viable populations of the large carnivore species.
Studies of carnivores and predation on reindeer have mostly been descriptive, without examining more closely the factors that may explain the results obtained. Hence, the results of such studies are of limited value for practical management because they merely deal with a single situation, at a specific time without saying anything about the general causes of losses. Important aspects may be the intensity of herding, scatteredness, grazing conditions, weather conditions in different seasons of the year, and interspecific factors like habitat overlap between predators and prey species. Future research should focus on interspecific interactions and apply models of population dynamics based on the interplay between species and their biology. This will improve the general basis for management and enable us to manage the natural heritage with a more long-term perspective.
this project was supported by grants from the Norwegian Research Council, the Norwegian Research Council of Semi-domestic Reindeer Husbandry (Rein-driftens forskningsråd), the Directorate for Nature Management, and the Norwegian Institute for Nature Research. We thank Karl Brøndbo, Torsten Ingerslev, Charlotte Moshøy, Stig Lasse Rosendal and Kai Magne Rosendal for field assistance and Nicholas Tyler for his helpful review on an early version of the manuscript.