Winter severity, survival, and cause-specific mortality of female white-tailed deer in north-central Minnesota

Knowledge of age-specific survival of deer (Odocoileus spp.) and the underlying causes of mortality are essential to our understanding of their population dynamics. We examined age-specific survival and cause-specific mortality of female white-tailed deer (O. virginianus) in an area where wolves (Canis lupus) recently had become reestablished and assessed the influence of winter severity during 1991-1996 using new survival analysis procedures. Sufficient data permitted rigorous survival analyses on 153 of 179 radiocollared females greater than or equal to0.6 years old, whose age distribution remained stable with annual median ages of 5.9-6.7 years old during the last 5 years of the study Winter severities ranged from unusually mild to historically severe; 84% (81 of 97) of all mortalities occurred during November-May. The median age of deer Survival according to the iterative Nelson estimator (INE) was 3.6 years (90% confidence limits [CL] = 2.4, 4.0). The smoothed estimate of the hazard function was a U-shaped curve with an elevated risk of mortality at 0.6 years old, which declined progressively until 5 years old, then increased steadily with the highest age-specific risk of death for deer greater than or equal to10 years old. Fits of the Anderson-Gill proportional hazards model to our data indicated that survival was negatively associated (P<0.0001) with winter severity (primarily snow depth), but not with site of winter capture or body mass at capture. The risk of death increased progressively over the course of each winter, with maximum risk occurring in winters of greatest snow depth. Estimated life-time mortality rates (90% CL) due to hunting, wolf predation, and miscellaneous causes were 43.3% (29.5, 57.1%), 33.1% (20.1, 45.3%), and 23.6% (13.2, 34.0%), respectively. Within 5-11 years of reestablishment in our area, wolf predation became a leading cause of mortality for female deer, but its importance relative to limiting depended on annual fluctuations in winter severity. Competing risk analyses permitted comparison of age-specific hazards associated with hunter harvest, wolf predation, and miscellaneous causes of mortality. The nature of the observed interaction between the cause-specific hazards and the effects of winter severity v appears to validate the practice of regulating the issuance of antlerless deer permits based on population condition. Collective evidence suggests that as agencies formulate management strategies relative to wolf reintroduction or range expansion, the frequency of severe winters, snow depth, deer population goals, and the use of antlerless permits should be primary considerations.