Montana Statewide Mule Deer Study

We also used multi-scale resource selection analysis to study how partially migratory mule deer balanceselection for forage and avoidance of predation risk during summer. We compared the availability offorage (kcal/m2) and predation risk from wolves (Canis lupus) and mountain lions (Puma concolor)between summer ranges of study area and according to migratory strategy, then assessed how selectionfor those factors at the home range (second order) and within‐home range (third order) scales varied.As forage availability increased among mule deer summer ranges and individual home ranges, selectionfor forage decreased at the second-order (P 0.052) and third‐order (P 0.081) scales, respectively, butavoidance of predators varied weakly. In 1 study area, summer range of residents contained lowerforage and higher risk than summer range of migrants, but residents compensated for this disadvantagethrough stronger selection of forage and avoidance of risk at finer spatial scales. In the other 2 studyareas, summer range of migrants contained lower forage and higher risk than residents, but migrantsdid not compensate through stronger selection for beneficial resources. The majority of mule deer inour study system were migratory, suggesting partial migration may persist in populations even whenmeasurable benefits in terms of forage and predation risk were not evident.Winter habitat relationships (Section 7): Mule deer in northwest Montana inhabit the northern forestsecoregion, where the winter season can pose relatively challenging climatic conditions. We also studiedresource selection during winter, with particular attention to the effects of snow on deer behavior. Wefound mule deer avoided areas with deeper snow while seeking out patches with greater solar radiationpotential. Additionally, interaction between responses to snow and canopy cover showed deer withinthe deepest snow conditions showed increased selection for high forest canopy, with associated highsnow-intercept. This result mirrored our diet results, showing increased use of conifer in deep snowenvironments. These differences suggest the importance of locally adapted behaviors as mule deernavigate energetic constraints of winter landscapes.Fall migration: timing and initiating factors (Section 8): Deer in all three study areas exhibited summermovements into different hunting districts or management jurisdictions than those occupied duringwinter. Mule deer captured in the Cabinet-Salish winter range of HD103 were spread across 3 huntingdistricts during summer, with some not returning to HD103 until the end of general rifle season. Deercaptured in the Rocky Mountain Front study area were spread across 8 hunting districts and the SunRiver Game Preserve when archery season began in early September. While many deer returned duringarchery season, some remained in wilderness until after the general rifle season concluded. Lastly,many Whitefish Range deer left winter range in HD109 to summer in HD110, British Columbia, or GlacierNational Park, in some cases being fully inaccessible during portions of the hunting season.Given the management implications of fall migrations, we conducted an additional analysis to explorewhat drives the initiation of these migrations back to winter range. We hypothesized that factorsbehind fall migration timing might include precipitation (ie., snow), cold temperatures, senescence ofplant forage, human hunting pressure, and the migration distance separating summer and winterranges. Results supported several of these variables, with the most parsimonious model indicating thatthe initiation of fall migration is correlated with increased precipitation over the previous week,decreased daily minimum temperatures, and begins earlier with longer migration distances.Integrated population modeling in the prairie-breaks (Section 9): MFWP biologists are charged withmanaging hunting regulations at hunting district scales yet, aside from harvest estimates, they lackpopulation data sampled at this scale. Here, we used integrated population models (IPMs) to evaluatedeer monitoring data within the framework of biological models of deer population growth, and wewere able to estimate an important, but previously unavailable metric: harvest rate of antlered buckMule deer ecology and population modeling, Final Report 6

mule deer. We used harvest and trend area population data from Region 7 during 1999–2017 and builtan IPM founded on data from trend areas. This model returned estimates of population vital rates (e.g.,age- and sex-specific survival) that were consistent with prior studies in Montana and elsewhere. It alsoprovided estimates of mule deer antlered buck harvest rates, averaging 46%, that could be used toextrapolate harvest data to HD-scale abundance. We compared these results to those generated fromextrapolation of deer densities observed in trend areas to the entire area of surrounding HDs, and wefound results to be similar during some periods but quite different during others. Generally, this newIPM-based approach is founded on the assumption that harvest rates, but not necessarily density, areequivalent for those portions of the population within vs. not within trend areas. We feel thisassumption may fit many but not all mule deer populations in Montana, and thus warrants furtherinvestigation towards robust estimation of HD-scale abundance and sex-age composition of mule deer.Management Implications (Section 10):— Migration imposes a disconnect between population unitsmanaged and sampled during the hunting season (via harvest regulations and harvest survey data) andpopulation units monitored during the winter (via post-season and spring trend surveys). This crossboundary movement of animals may emphasize broad-scale, multi-district, interpretation of bothharvest and survey data over finer-scale, single district monitoring data. Season structures that increaseharvest opportunity during different portions (early vs. late) of the season may have unequal effects onsubgroups of deer, depending upon migration timing and the spatial continuity of regulations. Formigratory deer, mirroring regulations in both summer and winter ranges over time may be necessary toavoid undue impacts on those that return early vs. late, though such regulations may ultimately bedifficult to implement. Regardless of migration considerations, the high natural mortality rates observedhere, driven primarily by predation from mountain lions, leave little to no room for additional huntingopportunity of antlerless deer if mortality from hunting is even partially additive to other natural causes.However, pulses of mortality of animals in poor nutritional condition during early spring do suggestsome potential for harvest mortality to be compensatory with winter mortality in these systems, as theeffects of winter severity were apparent on both adult and fawn survival.Responses of summer plant communities to disturbance were highly variable both within and amongstudy areas, and less desirable outcomes such as increased invasive species biomass and decreasedfloristic quality were also associated with some disturbances. Although different management outcomesled to varying optimal results in each study area, our methods, including SMART analysis, allowed us topredict outcomes in multiple settings and account for uncertainty in vegetative responses todisturbance. Decision analyses suggested that low-severity timber harvest was an appealing forestmanagement action in that it best balanced increased forage nutrition while limiting invasive speciesand maintaining floristic quality. Regarding the placement of targeted forest management for muledeer habitat, the high fidelity of mule deer to summer ranges and more consistent selection for forageat fine- than broad-scales may suggest limited utility of management-directed forage improvements toindividual deer unless placed within existing, often high-elevation, home ranges. Variation in snowconditions across our three study areas was also a primary driver of variation in winter mule deerecology. In deep-snow environments like the Whitefish Range, management to reduce forest canopymight create additional forage (e.g., shrub) resources but those resources appear to becomeinaccessible when and where snow is deep enough to inhibit movement or bury plants. In more mildwinter environments, habitat management to promote shrubs may be less costly to energy expenditureand increase the biomass of accessible forage, with favorable implications for deer population dynamics.Lastly, the population modeling approach we pioneered allows novel estimation of a previouslyunavailable and management-relevant metric: the harvest rate of antlered buck mule deer.Mule deer ecology and population modeling, Final Report 7

Furthermore, it facilitates absolute deer population estimates at HD- or broader-scales using existingmonitoring data collected by MFWP, the use of which has previously been limited to yielding trend indexcounts for subsets of area and deer populations (i.e., trend survey areas). We recommend furtherevaluation and validation of this methodology, including parsimonious incorporation of weathercovariates to potentially improve or smooth predictions.1. INTRODUCTIONOver the past century, mule deer (Odocoileus hemionus) have experienced periods of population growthand decline throughout their range (Mackie et al. 1998, Pierce et al. 2012, Bergman et al. 2015). Studiesof mule deer population dynamics have revealed a suite of interacting factors which influence annualvariation and trends in population growth (Mackie et al. 1998, Unsworth et al. 1999, Pierce et al. 2012,Monteith et al. 2014, Hurley et al. 2014, Ciuti et al. 2015). The complexity of mule deer populationdynamics creates a challenge for biologists seeking to monitor local deer populations and respond withappropriate management decisions in a timely manner (White and Bartmann 1998, Bishop et al. 2005).Mule deer population trends are of particular concern in Montana, where significant declines inabundance and hunter harvest (correlated) have been documented in many areas throughout the state.Wildlife managers are tasked with the difficult mission of maintaining or recovering deer populations,dampening the magnitude of potential future declines, and stabilizing populations and subsequenthunter opportunity. Therefore, improved quantitative understanding of mule deer ecology andpopulation dynamics is of relevance across Montana.Mule deer in the northern forests.—With this study, Montana Fish, Wildlife and Parks (MFWP), incooperation with the University of Montana, used detailed studies of mule deer ecology to addressseveral information gaps concerning mule deer within Montana’s portion of the “Northern Forests”ecoregion (Sections 2–8; Hayden et al. 2008). The Northern Forests ecoregion is the northernmost of 7ecoregions within which the mule deer of North America have been grouped according to similarities inenvironmental conditions and management challenges. The Northern Forests includes much of westernMontana but spans from portions of California across British Columbia and into the Yukon and Alaska.Basic ecological information for mule deer in northwest Montana is limited, and available populationmonitoring data have indicated declines. Topics of study in this report include vital rate monitoring,detailed assessment of seasonal habitat selection, diet and forage nutrition, and evaluation of seasonalspace use, migration routes, and drivers of fall migration timing.Integrated population modeling.—The methods by which MFWP currently monitors and managesmule deer were established in 2001 with the adoption of the Adaptive Harvest Management (AHM)system (MFWP 2001). This system included four primary components: 1) population objectives, 2)monitoring program, 3) hunting regulation alternatives, and 4) population modeling. The populationmodeling component of AHM was initially designed to predict future deer dynamics given a suite ofharvest and weather scenarios. Despite being founded upon very powerful data sets, Pac and Stewart(2007) found the AHM population models achieved mixed results and subsequently recommended theyremain in an experimental phase rather than be implemented as a management tool (AHM 2021). Withthis p

Nutritional condition and vital rates (Section 4): We measured nutritional condition in the form of percent body fat, estimated from ultrasonic rump fat measurements and body condition scoring. Nutritional condition varied widely across individual deer, and body fat declined significantly over time as the winter season progressed.