T.J. Brinkman | USGS Science

USGS2011

Influence of landscape characteristics on migration strategies of white-tailed deer

A trade-off exists for migrating animals as to whether to migrate or remain residents. Few studies have documented relationships between landscape variables and deer migration strategies. From 2000 to 2007 we captured 267 adult female white-tailed deer ( Odocoileus virginianus ) at 7 study sites in Minnesota and South Dakota and monitored 149 individuals through ≥3 seasonal migration periods (585 deer-migration seasons). All deer classified as obligate migrators with ≥3 migrations (range 3–9 migration seasons) maintained their obligate status for the duration of the study. Multinomial logistic odds ratios from generalized estimating equations indicated that the odds of being a resident increased by 1.4 and 1.3 per 1-unit increase in forest patch density and mean area, respectively, compared to migrating deer. Odds of being an obligate migrator increased by 0.7 and 0.8 per 1-unit decrease in forest patch density and mean area, respectively, compared to resident or conditional migrating deer. Areas inhabited by resident deer were characterized by greater number of forest patches per 100 ha and larger mean forest patch area than conditional and obligate migrant areas. Odds of migrating increased by 1.1 per 1-unit increase in deer winter severity index. Migration behavior of white-tailed deer varied among regions, and land-cover and landscape characteristics provided predictive indicators of migration strategies for deer that could have important implications for conservation, metapopulation dynamics, and species management.

Journal of Mammalogy↗

USGS2011

Survival of white-tailed deer neonates in Minnesota and South Dakota

Understanding the influence of intrinsic (e.g., age, birth mass, and sex) and habitat factors on survival of neonate white-tailed deer improves understanding of population ecology. During 2002–2004, we captured and radiocollared 78 neonates in eastern South Dakota and southwestern Minnesota, of which 16 died before 1 September. Predation accounted for 80% of mortality; the remaining 20% was attributed to starvation. Canids (coyotes [ Canis latrans ], domestic dogs) accounted for 100% of predation on neonates. We used known fate analysis in Program MARK to estimate survival rates and investigate the influence of intrinsic and habitat variables on survival. We developed 2 a priori model sets, including intrinsic variables (model set 1) and habitat variables (model set 2; forested cover, wetlands, grasslands, and croplands). For model set 1, model { S age-interval } had the lowest AIC c (Akaike's information criterion for small sample size) value, indicating that age at mortality (3-stage age-interval: 0–2 weeks, 2–8 weeks, and >8 weeks) best explained survival. Model set 2 indicated that habitat variables did not further influence survival in the study area; β-estimates and 95% confidence intervals for habitat variables in competing models encompassed zero; thus, we excluded these models from consideration. Overall survival rate using model { S age-interval } was 0.87 (95% CI = 0.83–0.91); 61% of mortalities occurred at 0–2 weeks of age, 26% at 2–8 weeks of age, and 13% at >8 weeks of age. Our results indicate that variables influencing survival may be area specific. Region-specific data are needed to determine influences of intrinsic and habitat variables on neonate survival before wildlife managers can determine which habitat management activities influence neonate populations.

Journal of Wildlife Management↗