A study investigated factors that may predispose white-tailed deer (Odocoileus virginianus) to surplus and excessive killing by gray wolves (Canis lupus) during winters of 1990-91 to 1995-96. Delgiudice reports surplus and excessive killing of deer by wolves was observed only during winter 1995-96.
Surplus and excessive killing is characterized by a predator killing its typical prey with no or low consumption of the carcass. It is atypical in that it appears to be in excess of the predator's short-term requirements for sustenance. Environmental conditions and nutritional status of prey have been implicated as factors contributing to their increased vulnerability to predation in cases of surplus and excessive killing, but our understanding of this phenomenon has been based primarily on circumstantial evidence....
Surplus and excessive killing of whitetailed deer (Odocoileus virginianus) by wolves (Canis lupus) have occurred during isolated severe winters in northeastern Minnesota and Ontario, but condition of snow or deer killed by wolves were not evaluated. During six winters that were mild to moderately severe in northcentral Minnesota, Fuller (1991) reported no variation in predation or rates of consumption of deer carcasses by wolves.
Ambient temperatures and snow conditions in winter during my ongoing study (1991-1996) of white-tailed deer in northcentral Minnesota have varied appreciably. I predicted that surplus or excessive killing would occur during prolonged periods of snow depths of 70 cm or more. I present data on environmental and physiological factors that may predispose deer to surplus and excessive killing by wolves.
MATERIALS AND METHODS
Study area -- Data were collected along the eastern and southern boundaries of the Chippewa National Forest in northcentral Minnesota....Six to seven wolf packs have been observed, and mean pack size ranged from four to six wolves during the study.
Minimum and maximum ambient temperatures and depth and penetrability of snow were recorded daily in open habitat on the study area from early January to late March, 1991-1996. Depth and penetrability of snow were measured at three to six locations each day....Ambient temperatures and depths of snow for November-December and late MarchMay were obtained from a weather station at Grand Rapids, Minnesota (National Oceanic and Atmospheric Administration, 1990-1996, in litt.).
Capturing, collaring, and monitoring deer -- Deer were captured primarily in Clover traps from January to mid-March. Captured female deer were immobilized...ear-tagged, weighed, radiocollared, and physically examined....
Radio-collared deer were located one to three times per week from fixed-wing aircraft. If a radio signal was in mortality mode (120 pulses/ min versus live mode of 45 pulses/min), the location of the collar was plotted on a topographic map....Subsequently, a ground crew located the kill-site and carcass.
Investigating winter wolf-killed deer -- I defined winter as 1 November through 31 May. Deer killed by wolves were identified based on evidence from the site (wolf tracks, feces, signs of a chase, blood) and carcass characteristics (wounds, patterns of consumption). Date of death was estimated from the most recent date that a radio frequency was detected in live mode, the first date that it was detected in mortality mode, and kill-site evidence (e.g., recent snowfall, tracks). Degree of consumption of the carcass was estimated visually. Ground crews also examined sites and carcasses of non-collared deer killed by wolves during winter 1995-1996. When available, a femur or other long bone was collected to measure marrow-fat content....
Weather -- Generally, mean weekly depth and penetrability of snow were greatest during January-April 1996 compared with this period in 1991-1995. Based on mean weekly depth of snow, annual differences in winter severity from 1991 to 1995 were apparent only from January to mid-late February, but they did not persist beyond late February. Total numbers of days with depths of snow 41 cm or more, a depth at which movements of deer may become severely restricted, were 1, 36, 61, 59, 9, and 116 days for November-May from 1990-1991 to 1995-1996, respectively. Coincidently, mean daily minimum and maximum temperatures were lowest for a longer period during winter 1995-1996.
Radio-collared deer -- From 1991 to 1996, 172 female deer were captured and radio-collared. During each winter, the number of radio-collared deer monitored for part (depending on date of death) or all of the season ranged from 22 to 75; median ages of the collared deer ranged from 1.9 to 6.7 years. Forty-one of the deer were killed by wolves during November-May; relative mortality ranged from 3.3 to 22.7% during the 6 winters. Seventy-one percent of the kills made by wolves during winter occurred in March-May....
During the first 5 winters of the study, remains of all 24 radio-collared deer killed by wolves were limited to rumen contents; scattered, small pieces of hide and bone or portions of hide; and several, mostly intact, bones (pelvis, scapula, various leg bones, vertebra, or ribs) stripped of tissue. Occasionally, the upper or lower portion of the vertebral column would be found intact and still attached to the skull or pelvis, respectively, but stripped clean of all soft tissue. In three to four instances per year, lower portions of dismembered legs were found with hide intact. In no cases were intact or largely intact carcasses recovered.
Surplus and excessive killing was observed only during winter 1995-1996; 11 relatively intact carcasses out of 17 (64.7%) radio-collared deer were recovered. During November-December, three of four death sites of deer killed by wolves were assessable. Remains of two of the three deer carcasses were typical of remains that characterized deer killed by wolves during the first five winters. However, the third deer, which was killed on 14 December 1995, was less thoroughly consumed by wolves. The hindquarters were eaten and the back and right ribs were stripped clean of meat, but the remainder of the carcass was intact.
No kills of radio-collared deer were made by wolves from 15 December 1995 to 9 March 1996, but 10 deer were killed between 10 March and 5 April. Consumption of nine carcasses was 0-33% and was 0-5% for five of those nine. Specific parts of carcasses that were consumed varied. Wolves ate only small parts of the hindquarters of two deer, the uterus (and fetus presumably) of 1 deer, and the heart, liver, and lungs of another deer. Wolves consumed ca. 80% of a tenth collared deer killed on 29 March. Three additional collared deer were killed by wolves from 12 April to 31 May, with nearly complete consumption of the carcass that typified the first 5 winters of the study....
Non-collared deer -- During winter 1995-1996, an additional 20 carcasses and kill-sites of non-collared deer killed by wolves were examined....Median age of those deer was 0.8 years. The degree of consumption varied. Of the 15 assessable carcasses, six were consumed to a degree typical of use during mild and average winters as described for radio-collared deer killed by wolves. One kill was made during December, three during early February, and one each during early and late March. Nine of the deer killed between mid-February and mid-March were underutilized, with 65% or less consumption of select parts of the carcass....
Typically, wolves consume all or nearly all of the deer that they kill, as was the case during the first 5 winters of this study. The surplus and excessive killing of deer observed during winter 1995-1996 is rare. Because the average length of time elapsed between the estimated date of death and examination of the carcass was similar (3-4 days) during all 6 years, the high incidence of surplus and excessive killing in winter 1995-1996 cannot be explained as an artifact of inconsistent follow-up time contributing to varied consumption by wolves. To my knowledge, previous investigations did not rule out this alternative hypothesis.
A preponderance of circumstantial evidence has indicated that unique or atypical environmental conditions may predispose, directly or indirectly, prey to surplus and excessive killing by predators. Mech et al. (1971) observed surplus and excessive killing of deer by wolves during a winter when depth of snow was 70 cm or more for ca. 12-13 weeks but not during a previous winter when snow was 60-72 cm for 7-8 weeks. It has been suggested that the debilitating effects of deep snow on condition, vigor, and mobility of deer were primarily responsible for the excessive killing. My data are consistent with this explanation. Surplus and excessive killing occurred only during winter 1995-1996, the only winter in which mean weekly depths of snow 70 cm or more (70-83 cm) persisted for a prolonged period (7-8 weeks). As was the case when Mech et al. (1971) reported surplus killing, daily depths ranged up to 96 cm. Furthermore, wolf-caused mortality of collared deer was highest during this severe winter (22.7%) compared with the previous 5 winters.
The mean weekly depths of snow in winter 1993-1994 were 55-65 cm between early January and mid-February but subsequently diminished precipitously. Wolf-caused mortality of deer was relatively high (18.2%), but there was no evidence of surplus or excessive killing. Seven of the nine deer killed by wolves that were aged were 7.5 or more years old; six of those were 9.7-16.9 years old. On average, older deer were killed by wolves during winter 1993-1994. Evidence suggests that older ungulates are more commonly debilitated by arthritis and periodontitis. During 7 winters in Ontario, Pimlott et al. (1969) reported varying degrees of excessive killing of deer by wolves, particularly during the severe winter of 1958-1959. Conditions of snow were not reported, but the mortality of deer that was due to poor nutrition was estimated at 8.5 deer/km2....
The interaction between penetrability and depth of snow, also has been implicated as contributing to energy loss of deer. From early January to late February, 1996, penetrability of snow indicated that a walking deer would have sunk 90-100% of the full depth of the snow. With a chest height of 51-61 cm (Kelsall, 1969), deer were forced to plow snow with their chests and resort to energetically expensive bounding (Moen, 1976). At snow depths 41 cm or more, deer movements are restricted severely....After early February, penetrability of snow began to decline; however, it did not decrease to less than 50 cm until late March (25 cm) and probably afforded an advantage to wolves pursuing deer, because they have a weight-on-track load less than one-half that of deer....
Few femurs or other long bones were available at kill sites of collared deer during winters when surplus and excessive killing did not occur and mortality by wolves was relatively low. However, comparison of mean fat content of a small number of long bones collected during late winters 1993-1994 (78.2%) and 1995-1996 (27.0%) indicated that during the latter winter, body fat reserves of deer nearly were exhausted, significant body protein had been lost, and many of the deer were probably moribund. The high number of days with snow depths 41 cm or more (116 days) suggested that during winter 1995-1996, energy reserves of deer were depleted more severely than during the previous 5 winters....Conversely, when snow conditions are moderate, marrow fat in deer killed by wolves is usually higher.
Figure 1: Mean weekly (a) depth and (b) penetrability of snow in open habitat during January-April 1991to 1996 in northcentral Minnesota.
Figure 2: Mean (a) daily minimum and (b) maximum ambient temperatures during winters (Nov.-May) 1990-1991 to 1995-1996.
Table 1: Number of radiocollared female deer monitored, those killed by wolves, and their median ages in northcentral Minnesota during winters (Nov.-May) 1990-1991 to 1995-1996.