Fine-scale genetics, population dynamics, and management of suburban white-tailed deer (Odocoileus virginianus)

Overabundant populations of white-tailed deer Odocoileus virginianus can cause broad declines in native biodiversity, the lack of advanced woody regeneration, and shifts in successional trajectories. These problems are especially pronounced in the suburbs of the Northeastern U.S. In the northern suburbs of New York City, land managers have begun implementing small scale (< 20 km2), bow-only hunts to reduce overabundant deer herds. The success of these controlled hunts will depend on deer socio-population dynamics, the efficacy of bow hunting, and the sustained participation of bow hunters. I used a multi-disciplinary approach to address these issues and to evaluate the utility of bow hunting as a tool for managing suburban deer.;I used surveys to evaluate why hunters from Westchester, NY and Fairfield, CT participated in controlled hunts. Members were primarily motivated by the chance to see wildlife, opportunities for recreation, and a passion for archery. Most (71%) reported that their enjoyment had increased since first joining their controlled hunt. Nevertheless, I documented several trends that threaten the long-term sustainability of these programs. First, 78.2% of survey respondents were over the age of 40, possibly suggesting few young recruits. Second, the opportunity to hunt previously unhunted land, a transitory incentive, was the most common reason for participating in controlled hunts. Third, respondents whose doe harvest was limited by choosing to spend time hunting outside of the controlled hunts were also more likely to have seen fewer deer when participating in those controlled hunts (G-test = 13.2, df = 4, P = 0.01). This suggests that if herd reduction is successful, and opportunities for seeing and harvesting deer become fewer, that hunter participation and effort may decline as well.;To evaluate the impact of deer management activities, land managers need accurate measures of deer abundance. I proposed modifications to Jacobson et al.'s (1997) camera trap method to estimate the abundance of the Mianus River Gorge Preserve (MRGP; Westchester County, NY) deer herd. This method uses photographs to provide a minimum count of distinctive branch-antlered males, and then uses the photographic rate of males, females, and fawns to estimate demographic ratios. These ratios are used to extrapolate from the number of individual branched-antlered males to the number of females, fawns, and spike males. I modified this technique to 1) generate measures of uncertainty for parameter estimates via bootstrapping camera stations, and 2) address the concern that demographic ratios will be biased if groups of animals differed in their probability of being photographed. For each demographic group, I standardized photographic rates by a measure of detection probability using linear regression. I evaluated the performance of using standardized vs. raw photographic rates by estimating female abundance using both sets of ratios. I compared the results to an independent estimate of MRGP female abundance based on mark-resight methods. Using standardized rates generated a female estimate (x¯ = 60.43 44, 95% PI = 44.78-77.38) closest to the mark capture estimate (x¯ = 49.50, 95% PI = 38.70-63.40), although precision intervals were wide regardless of whether raw or standardized ratios were used.;The concept that deer can be successfully managed at fine-scales (<20 sq km) is based on the rose-petal hypothesis (RPH). Under this theory, female deer are believed to exist in spatially exclusive, matrilineal social-units whose members are philopatric and exhibit low rates of dispersal. In theory, repopulation of the removal area should be slow as immigration is low and because deer immediately adjacent to the removal area will not shift their movements towards the void. Previous studies have demonstrated that the RPH model leaves a genetic fingerprint such that female deer separated by short distances (< 1.0 km) are genetically related. I used molecular techniques and Moran I's spatial autocorrelation analyses to evaluate whether an overabundant deer herd in suburban Westchester County was structured following the RPH. At the MRGP, tissue collected from deer that were sampled at the same spatial coordinate were genetically similar (r2=0.21). However, Moran's I index of genetic relatedness became non-significant (e.g. no different from 0) by 230 m. High deer densities may have resulted in overlapping ranges of non-related social units thus weakening a broader genetic signature.;To evaluate the short and long-term impact of bow hunting, I used projection models to simulate harvests on deer herds exhibiting density-dependent growth and survival across a range of carrying capacities and immigration rates. For scenarios believed to be representative of the MRGP (low immigration, ∼ 7%, and carrying capacity = 13.8 female km-2), maximum herd reduction was modest (∼20%) when simulated harvest rates approximated those observed at the MRGP. Sustaining harvests over multiple decades is the biggest challenge facing bow-only hunts. Hours per female harvest increased rapidly as population declined (CF(t) =0.0031NF(t); R2=0.56). As controlled bow hunts are executed by volunteer sportsman, realized reductions will be contingent hunters' incapability or unwillingness to increase effort. Consequently, bow hunting will likely result in deer densities lower than historical peak values, yet higher than is currently assumed necessary for forest regeneration.