|
|
Post by grrraaahhh on Nov 4, 2010 7:07:14 GMT -9
 Use of ungulates by Yellowstone grizzly bears Ursus arctos (Mattson) Brief extracts: Elk and Moose
Among the ungulates, elk were by far the most important prey. This was basically a density-dependent relationship, since elk outnumbered all other ungulates combined. Elk were most vulnerable at three periods during the annual life cycle: the rutting season, the end of winter, and at calving.
During and after the rutting season, bull elk are more vulnerable to grizzly bear attacks than at other times when grizzlies may selectively hunt them. Bulls become vulnerable in fall because of their habit of bugling, their strong rutting scent, and their preoccupation with mating and holding a harem. Many also receive incapacitating wounds in combat with other contending males. Any of these situations may trigger predatory behavior in the bear, which then uses the combination of skills most appropriate to the specific opportunity.
Anecdotal accounts from field observers illustrate how grizzly bears capture elk during these periods and how they behave at their kills. In addition to having great power and speed, grizzlies are known to stalk with great skill and work cooperatively to pursue and kill large prey. As "loners," they kill and defend.
Some grizzlies develop specific hunting strategies. Number 88 (Patch-eye), alpha male at the Trout Creek ecocenter in 1967, moved to the Firehole-Madison River drainage after the closure of Trout Creek in 1970. We observed him on three elk kills during the spring of 1971. Two of the three elk (both bulls) appeared to have been stalked and captured with a simple hunting strategy. Both carcasses were found in heavy timber less than 45 meters (50 yards) from the banks of the Madison River. We found drag-trails where the grizzly had hauled the elk from the river. These observations suggested that Patch-eye, from concealment in the timber, attacked the elk wading the river, his superior strength enabling him to overpower them in the water. Bone marrow compression tests showed that neither elk was malnourished at the time of death. In the same area, we observed an adult female grizzly on a fresh elk kill in the middle of the river. We believed the female used a similar hunting strategy.
Even in thermal areas such as the geyser basins, spring snow was often 0.5 to 1 meter (2-3 ft) deep in the timber. Winter-weakened elk were able to travel more easily in the river or along its banks than in the deep, wet snow of adjacent timber. Thus, kills tended to be concentrated in riparian habitat.The Grizzly Bears of Yellowstone (Craighead). |
|
|
|
Post by warsaw on Mar 12, 2014 13:42:19 GMT -9
GAME MANAGEMENT UNIT 13 BROWN BEAR STUDIES By Ted H. Spraker Warren B. Ballard and Sterling D. Miller STATE OF ALASKA Jay S. Hammond, Governor "...During spring and fall 1978, 23 radio-collared bears were observed on 78 kills. Moose of all age classes comprised 87 percent of the kills. Calf moose comprised 57 percent of the moose kills and 47 percent of the total kill..." "...Identification of brown bears as significant predators of moose will create problems for game managers attempting to manage moose (Ballard et al. 1980). If bears of all age and family classes are preying upon moose to the extent indicated in this study, then simple manipulation of bear sport hunting regulations will not likely reduce the impact of bear predation on moose..." Bear 208, a 12.5 year old female without cubs, was observed for 33 days in which she preyed on 4 adult moose, 1 adult caribou, 9 moose calves, and 1 small mammal or unidentified species. Estimated mass: 91 kg Length over curves: 180.3 cm Shoulder height: 107.6 cm Neck circumference: 59.1 cm Girth: 106.0 cm Body length: 104.1 cm Head width: 22.2 cm Head length: 35.2 cm Length upper left canine: 21.0 mm Width upper left canine: 15.2 mm Length lower left canine: 20.8 mm Width lower left canine: 14.0 mm Bear 205, a 4.5 year old male observed for 29 days, in which he preyed on 6 adult moose and 3 moose calves. Estimated mass: 205 kg Length over curves: 229.2 cm Shoulder height: 128.6 cm Neck circumference: 77.2 cm Girth: 124.1 cm Body length: 111.8 cm Head width: 21.6 cm Head length: 38.7 cm Length upper left canine: 22.0 mm Width upper left canine: 16.0 mm Length lower left canine: 23.0 mm Width lower left canine: 16.0 mm www.arlis.org/docs/vol2/hydropower/APA_DOC_no._3148.pdf |
|
|
|
Post by warsaw on Mar 14, 2014 14:13:46 GMT -9
Predation on moose and caribou by radio-collared grizzly bears in east central Alaska R. D. BOERTJE, W. C. GASAWAY, Alaska Department of Fish and Game, 1300 College Road, Fairbanks, AK 99701, U.S.A. Received December 17, 1987 D. V. GRANGAARD, and D. G. KELLEYHOUSE BOERTJE, radio-collared grizzly bears in east central Alaska. Can. J. Zool. 66: 2492 - 2499. Radio-collared grizzly bears (Ursus arctos) were sighted daily for approximately 1-month periods during spring, summer, and fall to estimate predation rates. Predation rates on adult moose (Alces alces) were highest in spring, lowest in summer, and intermediate in fall. The highest kill rates were by male grizzlies killing cow moose during the calving period. We estimated that each adult male grizzly killed 3.3 - 3.9 adult moose annually, each female without cub(s) killed 0.6 -0.8 adult moose and 0.9 - 1.0 adult caribou (Rangifer tarandus) annually, and each adult bear killed at least 5.4 moose calves annually. Grizzly predation rates on calves and grizzly density were independent of moose density and are probably more related to area-specific factors, e.g., availability of alternative foods. An important implication of our results is that managers should not allow moose densities to decline to low levels, because grizzlies can have a greater relative impact on low- than on high-density moose populations and because grizzly predation can be difficult to reduce. Grizzly bears were primarily predators, rather than scavengers, in this area of low prey availability (1 1 moose/grizzly bear); bears killed four times more animal biomass than they scavenged. R. D., GASAWAY, W. C., GRANGAARD, D. V., and KELLEYHOUSE, D. G. 1988. Predation on moose and caribou by BOERTJE, radio-collared grizzly bears in east central Alaska. Can. J. Zool. 66 : 2492 -2499. L'observation quotidienne d'Ours bruns (Ursus arctos) munis au cou d'un tmetteur radio pendant des ptriodes d'environ 1 mois au printemps, B l'ttt et B l'automne, a permis d'estimer les taux de prtdation. Les taux de prtdation d'originaux adultes (Alces alces) ttaient maximaux au printemps, minimaux durant l'ttt et intermtdiaires B l'automne. Les taux les plus tlevts ont ttt enregistrts chez des ours miiles prtdateurs de femelles nourricikres. Nous avons estimt que chaque ours miile adulte tue en moyenne 3,3 -3,9 orignaux adultes annuellement, chaque femelle sans petit, 0,6 -0,8 orignal adulte et 0,9 - 1,O caribou (Rangifer tarandus) adulte chaque annte, et chaque ours adulte, au moins 5,4 petits orignaux annuellement. Les taux de prtda- tion des ours sur les petits et la densitt des ours sont indtpendants de la densitt des orignaux et sont probablement relits et plus prks B des facteurs de l'environnement immtdiat, p.ex., la disponibilitt d'aliments de rechange. Les rtsultats dtmontrent que les spkcialistes de l'amtnagement ne doivent jamais laisser les densitts d'orignaux atteindre des niveaux trop bas parce que les ours ont un impact relatif plus tlevt sur les populations d'orignaux de densitt faible que sur les populations de densitt tlevte et parce que la prtdation par les ours peut &re difficile B enrayer. Les Ours bruns se sont avtrts plus prtdateurs que dttritivores dans cette region oh la disponsibilitt des proies est faible (1 1 originaux/ours); 80% de la matikre animale de leur regime prove- nait de la prtdation. R. D., GASAWAY, W. C., GRANGAARD, D. V., et KELLEYHOUSE, D. G. 1988. Predation on moose and caribou by [Traduit par la revue] Introduction system; they killed 52% of 33 collared calves (Boertje et al. 1987). The objectives of this study were (i) to estimate grizzly bear predation and scavenging rates on moose and caribou (Rangi- fer tarandus), (ii) to further define the ecological niche of the grizzly bear, and (iii) to determine some implications of grizzly predation for moose management. Quantitative studies on the predatory nature of brown or grizzly bears (Ursus arctos) are clearly lacking. Grizzly bears have been implicated as major predators on radio-collared moose (Alces alces) calves 1 6 weeks old in south central Alaska (Ballard et al. 1981, 1988) and the southern Yukon Territory (Larsen et al. 1987). Before those studies, grizzly bears in North America were often considered primarily scavengers of ungulates (Craighead and Craighead 1972; Jonkel 1978), although considerable circumstantial evidence and anecdotal accounts of grizzly bears killing livestock or wild ungulates are reported in the literature (e.g., Cole 1972; Haglund 1974; Ballard and Larsen 1987). No previous study of collared grizzly bears has distinguished grizzly predation from scavenging on adult ungulates, nor has grizzly bear activ- ity previously been monitored daily during extended spring, summer, and fall periods. The present study is part of an investigation of the factors limiting moose population growth in the Fortymile River drainage and adjacent areas of east central Alaska (Boertje et al. 1985, 1987; Gasaway et al. 1986). Low moose densities, 86 + 23 moose (90 % CI)/ 1000 km2 in 7500 km2 during 198 1, have prevailed in the area since the mid-1970's despite low moose harvest rates, low or moderate snow depths, and abun- dant browse. A moose calf mortality study during 1984 revealed that grizzly bears were major predators in this eco- Site description The grizzly bear study area (4000 km2), located in east central Alaska north of Tok (Fig. l), consists of rolling hills and seven sub- alpine and alpine mountains (1400- 1800 m) largely covered with mature black and white spruce (Picea mariana, Picea glauca) over- story below treeline (1000 m elevation). Subalpine vegetation consists primarily of dwarf birch (Betula nana) and willow (Salix spp.), inter- spersed with willow-lined drainages. Additionally, there are shrub- dominated burned areas, a 200-km2 poorly drained shrub-dominated flat (700 m), and meandering drainages bordered by willow, shrub birch (Betula spp.), alder (Alnus spp.), white birch (Betula papyri- fera), and white spruce. A wide diversity of berries is available at all but the highest elevations, primarily Vaccinium uliginosum, Vaccin- ium vitis-idaea, Empetrum nigrum, Arctostaphylos uva-ursi, Arcto- staphylos alpina, Arctostaphylos rubra, Vibumum edule, Shepherdia canadensis, Comus canadensis, Andromeda polifolia, and Oxycoccus microcarpus. The climate is subarctic and continental. Leaves emerged on most shrubs during June 1 - 7 in 1983 - 1986, and leaf senescence occurred Printed in Canada 1 Imprim6 au Canada Can. J. Page 2 BOERTJE ET AL. 2493 during the last 2 weeks of August. The average annual temperature near Tok (Fig. 1) is -4"C, and total annual precipitation averages 24 cm (National Climatic Data Center 1986). Large carnivores inhabiting the study area include wolves (Canis lupus), black bears (Ursus americanus) , and grizzly bears. Their prey include moose, caribou, beaver (Castor canadensis), snowshoe hares (Lepus americanus) , and hoary marmots (Marmota caligata) . Arctic ground squirrels (Citellus parryi) and salmon (Oncorhynchus spp.) are absent from the study area, and snowshoe hares have not been abundant in the study area since the early 1970's. Seasonal and annual distribution of the Fortymile Caribou Herd varies, but usually 300 -7500 caribou are in the study area; calving does not occur in the study area. Met hods This study was based on repeated observations of individual grizzly bears during fall 1985 and spring and summer 1986. Both grizzly bears and wolves were radio collared using methods described by Boertje et al. (1987). Seventeen grizzly bears were collared during spring 1985, one during late fall 1985, and four during spring 1986. Estimating predation rates on moose and caribou 2 I year old Spring, summer, and fall predation rates (number of bear-days/ number of kills) were calculated from daily observations (except 5 days) during radio-tracking flights from April 30 through June 10, 1986, July 9 through August 10, 1986, and September 18 through October 18, 1985, respectively. Of the 5 days on which flights were not made, no 2 were consecutive; therefore, based on observations of the length of time bears spent on yearling and adult kills, we included these 5 days when totaling the number of bear-days. Data were also included from five instances in which individual bears were obscured by fog or dense vegetation; however, the obscured bear was always sighted the following day to confirm whether a kill had been made. Bear-days excluded from the calculation of predation rates include the following (i) the first 5 days following immobilization because daily movements were obviously reduced for 4 days, and (ii) days that bears spent in or near dens (i.e., ~ 2 0 0 Student's t-test was used to test for differences in predation rates among seasons and between males and females. One pilot was responsible for all flights in a Piper Super Cub; he was accompanied by an observer. In most cases, bears were sighted on the first pass or circle. If bears were traveling when first observed, we searched the expected preceding travel path for carcasses and then relocated the bear before departing. To determine cause of death, we necropsied moose and caribou car- casses within 36 h of sighting them, except in one instance (72 h). Sites were examined for evidence of a struggle and other sign. Femurs and lower incisors were collected from each carcass when possible. To estimate annual grizzly predation rates, we extrapolated data from the three seasonal radio-tracking periods to May 1 through October 15 (168 days), based on mean den departure and entrance dates in the study area. A maximum annual kill rate was derived by minimizing the summer extrapolation to 56 days (June 26 through August 20) which maximized spring and fall extrapolations; con- versely, a minimum annual kill rate was derived by maximizing the summer extrapolation to 84 days (June 14 through September 5). Maximum and minimum annual kill rates were calculated because of uncertainty over when to begin and end the summer extrapolation. Dates are based on phenology, bear breeding behavior, and chronol- ogy of breeding (J. Hechtel, unpublished data). Data on den departure and entrance were only from bears for which annual predation rates were calculated, i.e., males and females without cub(s) of the year (hereafter referred to as females WOC). The term "females WOC" includes lone females, females with yearlings, and females with 2-year-old cubs. These groups were pooled because predatory behavior was relatively similar (Spraker et al. 1981 ; Miller 1985, 1986). Females with cub(s) of the year (hereafter referred to as females WC) were treated separately because of their restricted move- ment patterns. m from dens). The two-tailed FIG. 1. Grizzly bear study area (4000 krn2) in east central Alaska, 1985 - 1986. Derivation of ratios Ratios of moose to grizzly bear were derived using moose densities from November 1987 and grizzly bear densities from May 1, 1986 (Boertje et al. 1987). We estimated a moose population of 700 based on 475 moose counted in high-density areas during low-intensity searches with a Super Cub plus an estimate of 225 moose missed in surveyed areas and in areas not surveyed. Flight paths followed con- tours of slopes and moose were counted while circling each group. Estimating minimum predation rates on calves We estimated minimum bear predation rates on calves from obser- vations of radio-collared bears feeding on calf carcasses. We made these observations during the same flights used to estimate predation rates on moose and caribou 2 1 year old. Because calves were not regularly observed until May 22 in 1986, we calculated minimum spring predation rates only for May 22 through June 10. We did not land at all calf carcasses, but assumed that calves attended by radio-collared bears were killed by those bears. This assumption is based on data from necropsied calves attended by bears in 1984 (Boertje et al. 1985) and 1986 and on direct aerial observa- tions in 1985 and 1986 of bears killing and subsequently feeding on calves. We found no evidence from our study of radio-collared calves that grizzly bears were consuming calves that died of other causes; this observation agrees with other studies (Ballard et al. 198 1 ; Franz- mann et al. 1980). Results Predation on moose and caribou r 1 year old Predation rates by male bears were highest during spring (1 kill per 26 bear-days), lowest during summer (1 kill per 132 bear-days), and intermediate during fall (1 kill per 43 bear- days), but rates were not significantly different (p > 0.1; Table 1). Extrapolated annual kill rates for an adult male bear ranged from 3.3 to 3.9 adult moose with 0.8 and 6.6 as the extremes of 90 % confidence intervals. Estimated seasonal predation rates for female bears WOC Page 3 CAN. J. ZOOL. VOL. 66, 1988 were not significantly different (p > 0.1 ; Table 1). The extrapolated average annual kill rates for an adult female grizzly WOC ranged from 0.6 to 0.8 adult moose and from 0.9 to 1.0 caribou 2 1 year old. Extremes of 90 % confidence intervals on these averages were 0.1 and 1.4 for moose and 0 and 2.2 for caribou. Predation rates of the various reproductive classes of females WOC were as follows: lone females made four kills during 467 bear-days, a female with yearlings made 0 kills during 22 bear-days during fall, and a female with 2-year-old cubs made two kills during 72 bear-days during spring and summer. Females WC killed no moose or caribou 2 1 year old during 117 bear-days in spring and summer. Lack of kills was prob- ably due in part to restricted movements and low prey densities (Boertje et al. 1987). Adult male bears 2 8 years old killed adult moose at signifi- cantly greater rates (p < 0.1) than female bears 2 4 years old WOC, when data were combined for the three observation periods (Table 1). Differences in these kill rates may be due to age-specific differences between bears sampled. However, when predation data on moose and caribou 2 1 year old were combined, no differences (0.1 < p < 0.2) in predation rates were found between male grizzlies and female grizzlies WOC. Data suggest that some males 2 8 years old may be more predatory than others and that most males kill adult moose annually. Of the seven adult male bears, two killed three moose each in 65 and 72 bear-days, one killed two moose in 69 bear-days, one killed one moose in 49 bear-days, and three did not kill adult prey in 15, 3 1, and 74 bear-days, respectively. Therefore, four of five adult males that were observed for at least 49 days killed adult moose. However, no male bears were observed for twice the 42-day average interval between adult moose kills, which was the subjectively determined minimum interval required for assessing if a bear was likely to be a predator of adult moose. Certain females WOC may also be more predatory than others. However, data are inadequate to assess whether most adult female grizzlies kill adult moose and (or) caribou annu- ally. Of the l l female bears (Table l), 2 each killed l moose and 1 caribou in 94 and 106 bear-days, 1 killed 1 moose in 63 bear-days, 1 killed 1 caribou in 27 bear-days, and 7 did not kill adult prey in 5, 17, 22, 24, 47, 73, and 84 bear-days, respectively. Minimum predation rates on calves Our extrapolation of kill data in Table 2 indicates that each adult bear killed a minimum of 5.4 f 0.8 (SE) moose calves annually. We used the summer predation rate to extrapolate through the June 11 - July 8 gap in the data because radio- collared calves died at a faster rate before June 10 than there- after (Boertje et al. 1987). We also assumed that no calf predation occurred after August 10 because there were no calves killed during 323 fall bear-days. This calculation results in a minimum rate because we could have missed observing some calves (see section on biases in Discussion). However, an independently calculated annual kill rate of five calves per grizzly bear 2 3 years old suggests that bias is low in our study area. The calculation is based on numbers of calves produced in the study area, numbers of grizzly bears, and the fact that bears killed 52% of the radio-collared calves (Boertje et al. 1987). When all bears were combined, spring predation rates on moose calves (7 bear-days per kill) were significantly greater Page 4 BOERTJE ET AL. TABLE 2. Number of moose calf kills observed during spring (May 22 - June 10) and summer (July 9 - August 10) 1986 attended by 15 collared grizzly bears Minimum no. of calves killed Minimum no. of bear-days per killa (SE) No. of bear-days Spring Summer Spring Summer Spring Summer Total Males (n = 5) Females WOC (n = 6) Females WC (n = 4) Total 74 100 5 1 225 132 120 66 318 4 8 5 1 14 18(7) 5 (2) 7 (2) 7 (2) 16(5) 24 (5) 66 (66) 23 (5) 17(4) 9 (3) 14 (4) 12 (3) 20 7 3 1 NOTE: NO calf kills were observed during 323 bear-days during September 18 - October 18, 1985. Values were used to extrapolate to minimum annual grizzly predation rates for May 22 - August 10 in east central Alaska. "For females without cubs, spring and summer values differed ( p < 0.02). For data of all bears combined, spring and summer values differed ( p < 0.02). No other means differed significantly ( p > 0. I) when tested between seasons or among totals or classes of bears. (p < 0.02) than summer rates (23 bear-days per kill; Table 2). However, when classes were treated separately, only females WOC killed significantly more calves (p < 0.1) during spring than during summer. Actual differences between spring versus summer predation rates were probably greater than reflected by our estimates (see section on biases in Discussion). The three sex and reproductive classes of bears did not kill calves at significantly different rates (p > 0.1) when com- pared within a season or when seasons were combined for each class of bear (Table 2). All adult grizzlies studied killed moose calves, except two that were sampled over only 1 and 11 bear-days, respectively. Yet, in spring, some grizzlies were more predatory than others. Four of the nine bears sampled for 17 - 20 days in spring killed 21 (72%) of the 29 calves killed by these bears. Grizzly handling time for moose and caribou I Adult grizzly bears consumed adult moose in 7- 14 days. Accurate estimates of days required for grizzly bears to con- sume adult moose carcasses were obtained in four instances in which single bears or family groups were observed on or immediately adjacent to carcasses (< 100 m) during consecu- tive daily flights; the bears completely consumed the carcasses before departing. The handling times were as follows: (i) an 1 1-year-old male grizzly spent 14 days on an adult bull during October, (ii) an 1 1-year-old male spent 8 days on an adult cow during September, (iii) an adult female and one 2-year-old cub spent 7 days on an adult cow during June, and (iv) an adult female and two 2-year-old cubs spent 7 days on an adult bull during June. Grizzly bears consumed caribou r 1 year old in 2 - 3 days. An adult female bear completely consumed a 1.3-year-old caribou in 2 days, a female and two 2-year-old cubs consumed an adult female caribou in 3 days, and two lone grizzlies con- sumed a 2.3-year-old male caribou in 3 days (1 day by an adult female followed by 2 days by an adult male). 1 year old Food acquisition: predation versus scavenging Collared grizzly bears killed about four times more animal biomass than they scavenged. Data presented (Table 3) are total animal biomass available at observed scavenging and kill sites. Because we disturbed bears at feeding sites, they did not necessarily consume all animal biomass available. However, bears would likely have consumed equal proportions of bio- mass at scavenging and kill sites. Grizzlies appropriated and consumed more wolf kills than vice versa. Three of five adult moose killed by wolves were largely scavenged by grizzly bears; numbers include all spring and summer observations in the study area during 1984 through 1986. In contrast, only 1 of 12 adult moose killed by grizzly bears was largely scavenged by wolves. However, wolf-killed moose contributed only about 200 kg (3%) of 6640 kg of adult moose available to collared grizzly bears (Table 3). If either predator affected the other's kill rate, grizzlies elevated the wolf kill rate. One factor contributing to the higher rate of carcass appropriation by grizzlies than by wolves was the higher grizzly densities (16 grizzly bears/ 1000 km2 versus 5 wolves/1000 km2 during spring 1986; Boertje et al. 1987). Discussion Biases associated with the derivation of predation rates We discuss potential biases associated with estimating grizzly predation rates because identifying biases is vital for interpreting existing data and increasing comparability in future studies. However, biases discussed yield no quantitative basis for altering our results. - Potential biases associated with estimating grizzly predation rates on ungulates 2 1 year old Several factors indicate that we had a high probability of observing carcasses of moose and caribou 2 1 year old killed by collared grizzlies during the daily observation periods (Table 1). First, we visually sighted nondenned bears on 99 % of our attempts. Second, intervals between sightings of non- denned bears did not exceed 1 day, yet grizzly handling times were r 2 days (see Results). Third, all observed carcasses were sighted within 36 h of when bears made the kills (as evi- denced by daily map locations of bears). Fourth, before our disturbance of bears from carcasses (up to 72 h), bears were on or immediately adjacent to ( < 10 m) , and protective of, all but one observed carcass. In this exception, the grizzly was approximately 50- 100 m from the carcass the day of the kill; tracks in the snow indicated that the grizzly had killed the moose. Two biases may have caused overestimation of predation rates. First, a kill could have been attributed to a collared bear when in fact the collared bear could have displaced a non- collared bear from its kill. This bias affects predation rates of large males more than that of small males and females. Ballard et al. (1988) appropriately attributed 0.5 kills to a collared bear if collared and uncollared bears were initially seen at the kill site. We had no similar cases. Second, disturbing bears from Page 5 2496 CAN. J. ZOOL. VOL. 66, 1988 TABLE 3. Numbers and approximate total weights of animals preyed upon and scavenged by collared grizzly bears, east central Alaska, September 1985 - August 1986 Predation Scavenged Spring Summer Fall Spring Summer Fall Total wt. (kg) Total wt. (kg) Total wt. (kg) Total wt. (kg) Total wt. (kg) Total wt. (kg) n n n n n n Moose 2 1 year old Calf Caribou 2 1 year old Calf Gut piles Black bear Grizzly bear Adult female Cubs Total wt. (kg) No. of bear-days Animal biomass per bear-day (kg) NOTE: These data relate to total animal biomass available at observed feeding sites; however, because we disturbed them, collared bears did not necessarily consume biomass available. carcasses during our necropsy investigation could have increased kill rates. This bias potentially overestimates male kill rates more than female rates because we chased males off kills more frequently than females. Underestimation of the kill rate on ungulates r 1 year old may have resulted from three sources. First, a radio-collared bear could have been displaced from a kill before we observed it; this bias affects predation rates by females most. Second, collared bears could possibly have killed and consumed year- ling caribou between daily flights (<36 h), although this is unlikely as collared bears usually spent >36 h on moose calves observed in July, which weigh about 20 kg less than yearling caribou. This bias affects rates by large males most because of their shorter handling time. Third, radio collar malfunctions (one in our study) could have resulted from a moose's kick and, if the grizzly killed the moose, the predation rate would have been underestimated. This bias affects preda- tion rates by females most because they took longer to lull adult moose than males, as evidenced by sign around the kill site. To summarize our qualitative assessment, net bias tends to cause overestimation of predation rates by male grizzlies whereas net bias is unclear for predation rates by females. The previous predation studies of collared grizzlies (Ballard et al. 198 1, 1988) based predation rates on number of observa- tion days or visual sightings per kill or suspected kill ("kill"), rather than on total consecutive days of tracking as in this study. Predation rates based on observation days or visual sightings overestimate actual predation rates (Fuller and Keith 1980) on adult moose because the probability of observing a grizzly bear on an adult moose carcass is greater (up to 14 times greater in our study) than observing the bear the day on which the kill was made. For example, Ballard et al. (1981) reported an estimate of grizzly predation rates on adult moose as one adult moose "kill" per 16 observation days (n = 28 "kills"). Ballard et al. (1988) reanalyzed data on "kill" rates of adult moose by excluding data from the first day of observa- tion periods. They also summarized 2 additional years of data and reported that grizzly bears "kill" one adult moose every 43.7 bear-days (n = 13.5 "kills") in spring. Their revised estimate remains potentially an overestimate because (i) they did not confirm cause of death through necropsy and (ii) they observed collared bears revisiting carcasses (Spraker et al. 1981); therefore not all moose carcasses should have been counted as recent kills. In contrast, their revised estimate of the adult moose kill rate may be an underestimate because 20 addi- tional unidentifiable "kills" were noted, some of which could have been adult moose. To minimize biases in estimating grizzly predation rates on ungulates 2 1 year old, we recommend (i) that bears be observed daily or at least once every 2 days and (ii) that carcasses be examined on the ground within 24 h of discovery. Because grizzly bears often quickly bury adult ungulate car- casses (Mysterud 1973), it is often impossible for an observer in an aircraft to determine the date of kill or degree of consumption. Biases associated with documenting grizzly predation rates on moose calves Observed predation rates on moose calves (Table 2) are minimum rates, although observed summer rates are more accurate than spring rates. Bears killed and consumed neonatal calves between our daily flights in one or two of nine instances of predation on single collared calves in 1984. Furthermore, Ballard et al. (1988) found that most calf carcasses (28 of 32) were attended by bears for < 12 h, based on two observation flightslday during late May and the first few days of June. Also, D. Larsen (personal communication) noted that grizzly bears attended neonatal calf carcasses for <4 h in several instances in the Yukon. In contrast, larger summer calves were more reliably located; 10 of 12 single calf carcasses were attended by bears for two daily flights in our study area. Predation by bears Visual inspection of data (Fig. 2) suggests that predation on adult moose is greatest in spring, lowest in summer, and intermediate in fall.   |
|
