Mule deer (Odocoileus hemionus) are an ecologically, economically, and socially important species across much of the western United States. As such, populations and habitat are intensely managed by state game agencies. However, populations have been declining in recent decades and several factors have been implicated (e.g., climate, predation, competition, and habitat availability). Population dynamics of mule deer are driven by a combination of survival of adults and juveniles and reproductive rates. While adult female mule deer typically have consistently high annual survival rates (85% annually), juveniles are more easily affected by stressors (biotic and abiotic conditions) and therefore their annual survival rates are generally low and highly variable. In an effort to better understand the effects of management on—and the habitat needs for—reproduction and recruitment, we examined the potential effects of male-biased harvest on recruitment in populations of mule deer and the selection of sites for parturition by mule deer females. Changes in buck:doe ratio due to male-biased harvest may alter rates of pregnancy, timing of parturition, and synchrony of parturition if inadequate numbers of males are present to fertilize females during their first estrous cycle. If rates of pregnancy or timing of parturition are influenced by decreased buck:doe ratios, recruitment may be reduced. This results from fewer births, later parturition (resulting in lower survival of fawns), and a less synchronous parturition period (increasing the proportion of neonates exposed to predation). Our objectives were to compare rates of pregnancy, timing of parturition, and synchrony of parturition between exploited populations of mule deer with relatively high (Piceance Basin) and relatively low (Monroe Mountain) buck:doe ratios. We determined rates of pregnancy via ultrasonography and timing of parturition via expulsion of vaginal implant transmitters. We found no differences in rates of pregnancy, timing of parturition, or synchrony of parturition between Monroe Mountain and Piceance Basin. This suggests that the relatively low buck:doe ratios typical of heavily harvested populations do not have unintended or indirect impacts on population dynamics because recruitment remains unaffected. Because neonate ungulates are most vulnerable to predation during parturition and shortly thereafter, selecting sites for parturition can have direct fitness consequences. We investigated the selection of sites for parturition by mule deer. We utilized vaginal implant transmitters to identify sites of parturition. We then obtained and compared macro- and micro-habitat features between sites of parturition and associated random sites. Parturitient females selected sites based on topography, habitat-type, and obscurity. Enhanced understanding of habitat variables that are selected for parturition provides insight into the life history or behavior of a species and allows managers to ensure that suitable habitat is available for this stage of life-cycles.

Translocation of wildlife has become common practice for wildlife managers charged with management of animals on increasingly modified landscapes. Translocation can be used to reduce population density in the source area, supplement existing populations, reestablish extirpated populations, and establish new populations. Mule deer (Odocoileus hemionus) are a species of great interest to the public in western North America. Although translocations have been used to manage mule deer, very little has been done to document the outcomes of this management practice. The purpose of this research was to evaluate movement, site fidelity, space use, and survival of translocated mule deer in relation to the timing of release (early versus late winter) and to provide managers with information useful in judging the relative value of translocation as a management strategy for this species. We captured 102 mule deer in January and March 2013 and translocated them from winter range near Parowan, UT, to winter range along the Pahvant Mountain Range near Holden, UT (approximately 144 km north of the capture location). Each deer was fitted with a radio transmitter (21 GPS collars, 81 VHF collars) prior to release to document outcomes. In January 2013 and 2014 we also captured and marked a total of 70 resident deer (non-translocated deer; 9 GPS collars, 61 VHF collars) to serve as a reference group within our study area. Following release, we monitored deer weekly through March 2015. We found that translocated deer had lower annual survival rates than resident deer during the first year following release, but similar annual survival rates to resident deer during the second year following release. Additionally, we found that age strongly influenced the survival of translocated deer; young deer (e.g., 2.5 year olds) were more than twice as likely to survive the initial year following translocation than old deer (e.g., 7.5 year olds). We also found that translocated deer had larger home ranges compared to resident deer during the first and second years following release. However, the average size of translocated deer home ranges decreased from year 1 to year 2 following release. Despite these large home ranges and extended movements during the summer months, most surviving deer (96 %) returned (within < 30 km) to winter range where they were released. We found no difference in movement, site fidelity, or survival for transplanted deer released in January and March. Based on our findings, wildlife managers that elect to translocate mule deer should not expect a difference in survival between early and late winter releases, but will likely see high site fidelity, higher survival rates during the second year following translocation (compared to the first year), and higher survival rates for younger deer compared to older deer.

Monitoring population trend and estimating vital demographic parameters are essential for effective management of a mule deer (Odocoileus hemionus) population. Because of financial constraints, many wildlife agencies use computer models to obtain indirect indices of population size and trend as an alternative to annual field-based estimates of population size. These models are based primarily on herd composition counts and harvest rates from hunter-harvest surveys, and are rarely field validated. I developed an alternative method for monitoring population dynamics of wintering populations of mule deer. I designed a hybrid monitoring program that combined annual vital rate monitoring to track changes in population growth rate with a field-based approach for estimating population abundance. The program allocated resources optimally towards the most critical components of mule deer population dynamics, and consisted of 4 field surveys: annual monitoring of age ratios, overwinter fawn survival, and annual doe survival, with field-based estimates of population size only once every 4 years. Surveys were conducted from 2006 to 2008 in Wildlife Management Unit (WMU) 2, Utah, and cost $29,298 per year, prorated over 4 years. Unfortunately, financial constraints prohibit the implementation of this monitoring program in every WMU in Utah. Instead, the program can be implemented in select WMUs throughout the state, with survival data collected in these core units, providing estimates for nearby satellite units. To establish core-satellite unit pairs, I developed a proxy method for determining correlation in survival rates between core and satellite units using model-simulated estimates. I demonstrated this core-satellite method using WMU 2 as a core and WMU 3, an adjacent unit, as a satellite. Finally, I compared a multiple data sources (MDS) model with a herd composition-based population model, POP-II. The MDS model better approximated observed data, and provided statistical rigor. Overall, the hybrid program was less costly and provided more precise estimates of population trend than could be achieved with a monitoring program focused on abundance alone, and was more defensible than herd composition monitoring. After establishing correlations in doe and fawn survival between core and satellite units, data collected in core units via the hybrid program could then be used to model the mule deer population dynamics of other units using MDS modeling procedures. This combined approach could be an effective statewide program for monitoring mule deer populations.

For many species, population size data is difficult to obtain or even unavailable. Therefore, estimations or indirect abundance measures of populations are crucial for ungulate management. Hunting has an important role in wildlife management, and is a partnership between state, landowners, and hunting communities. In ungulate management harvest statistics, as well as ungulate-vehicle collisions (UVCs) and observations, are often used as proxies for population densities and provide detailed information on a spatial and temporal scale. A Bayesian approach was used to model hunters’ response to population densities of fallow deer (Dama dama), roe deer (Capreolus capreolus), and wild boar (Sus scrofa). The results indicate that the variability in number of individuals observed and the non-linearity of both hunters’ harvest and UVCs response varies between roe deer, fallow deer, and wild boar. Both hunters’ harvest and UVC display a sub-linear response to population densities to all three species. Additionally, roe deer show a low variability in the number of individuals observed compared to the other two species. Predictions of population densities by using proxies like harvest statistics and UVC statistics need to be species-specific to be reliable, and by considering a potential sub-linearity and temporal trends for the species of interest more confident and realistic estimates can be developed.