Long-term monitoring of large carnivores is important for maintaining species viability and evaluating management actions. Yet the spatial and temporal scales at which large carnivores operate (e.g. reproduction, dispersal) results in logistical and economic challenges to applying monitoring techniques (Pollock et al. 2012). In states where large carnivores are hunted, harvest indices are often used to evaluate coarse-scale population trends and adjust subsequent hunting limits. For cougars (Puma concolor), these data are readily available, cost-effective, and validated as an effective monitoring metric (Wolfe et al. 2016). Even so, it is important to assure monitoring continues and metrics best describe population size (Pollock et al. 2012). This is particularly important as scrutiny of management plans increases under pressure from a growing number of the public that oppose hunting of large carnivores. Continuous monitoring of adult female survival provides an excellent baseline for calibrating the accuracy of such metrics (Wolfe et al. 2016, and references therein), and additional demographic and behavioral information can refine estimate precision and help predict population trends (Rankin and Kokko 2007). Reproductive rates and juvenile survival are essential population parameters for predicting population trends in long-lived mammals (Lambert et al. 2006; Martorello and Beusoleil 2003). For cougars, previous research indicates that reproduction is highly variable, and kitten survival is virtually unknown. Moreover, harvest can reduce adult survival rates. Thus, robust estimates of reproductive rates and dispersal routes (i.e. population connectivity) could serve as calibrating measures to increase the accuracy of population trend estimates. However, few studies for cougars in Utah - or elsewhere - exist that focus on these metrics (but see Sinclair et al. 2001). Although many cougar studies have obtained fecundity rates (mean litter size, inter-birth intervals), few have followed offspring to dispersal to determine population impacts. At the same time, genetic connectivity studies indicate a mismatch in scale between sampling and connectivity (Sinclair et al. 2001). Thus, estimating population parameters under a range of environmental and management conditions is critical for developing robust, defensible population estimates. Cougar populations in Utah are monitored using annual harvest indices (UDWR 2015). The two primary indices of sustainable harvest are the proportion of adult females in the harvest (≤ 40%), and that ≥ 15% of the harvest is comprised of adults ≥ 5 years old (UDWR 2015). In areas where mule deer (Odocoileus hemionus) and bighorn sheep (Ovis canadensis) survival are a primary management objective, only the proportion of females in the harvest, is used (UDWR 2015). Cougars primarily prey on mule deer in Utah, however, they incorporate other prey species into their diet as availability dictates, and some individuals learn to specialize on select species (Ross et al. 1997, Elbroch et al. 2013). Other prey commonly found in cougar diet include elk (Cervus canadensis), bighorn sheep, pronghorn antelope (Antilocapra americana), moose (Alces alces), wild horses (Equus ferus), and livestock. Many states offer compensation for livestock losses to large carnivores, but these programs are costly, do not necessarily increase positive perceptions towards carnivores, and may not accurately reflect economic losses (van Tassell et al. 1999, Wagner et al. 1997). In fact, many western states use a multiplier to determine compensation payments for livestock depredation (e.g., for every livestock verified as killed, the owner is compensated at a rate that includes X additional livestock that are likely killed but undetected). Compensation schemes have recently been advocated for adoption in Utah but there is limited to no data from which to set the multiplier rate. In areas where cougars overlap bears (Ursus americanus), they may be displaced from prey caches. This may increase kill rates or focus predation on secondary species. Elbroch et al. (2014) found a 48% increase in ungulate kills per week by cougars during summer when black bears were active. Such effects could result in underestimating the impacts of cougars on their primary prey if relying on predator population indices alone. In Elbroch et al. (2014), cougars lacked spatial refugia from bears. Further, another study of interactions between other large carnivores found contradictory evidence (Taillan et al. 2017). In Utah, cougar and bear habitat show a close correspondence, so additional information from Utah systems would clarify factors influencing the relationship between kill rates and cougar-black bear competition and provide specific recommendations for managing these species.

The Utah Cougar Management Plan (2015) outlined several research objectives focused on estimating cougar populations, obtaining population parameters of cougars, prey switching, and impacts to mule deer. Based on UDWR top needs, our goal is to obtain population parameters to improve estimates of cougar abundance and population growth rates to inform management decisions. To meet this goal, our proposed study will include fieldwork on cougars in three areas of Utah to address three primary objectives: 1. Estimate adult female cougar survival and reproductive parameters to more accurately quantify population estimates; 2. Determine if seasonal migration of native ungulates affects cougar movement or prey switching; and 3. Identify if black bears affect cougar predation patterns and how this alters total predation, mule deer survival rates, and population parameters associated with population estimates. Our goal and objectives match the top priorities from the Utah Cougar Management Plan (2015) and those articulated in several meetings by UDWR personnel, including the Mammals Program Coordinator and Wildlife Section Chief. In addition, this proposal answers several questions within each of these primary objectives. For example, we anticipate the possibility of also gaining an improved understanding of kitten survival to dispersal under objective 1, actual versus accounted take of livestock during depredation events under objective 2, and importance of ungulate biomass in the foraging patterns of black bears under objective 3.

Because large carnivores are difficult to observe in the field, researches have been diligently trying to devise better methods to estimate population densities. This study will implement techniques that could be used across the State to get better estimates. Also because large carnivore regularly target livestock, one way to try to build tolerance for these species on the landscape has been to compensate livestock producers for losses. One complaint with this process has been that Wildlife Services can only account for a fraction of actual losses to predators. This study will try to determine the numbers of livestock that go unaccounted for when losses are verified by Wildlife Services.

The objectives from this study come directly from the Utah Cougar Management Plan.

N/A

N/A

N/A

Description of study design To meet these three objectives, we will first compile, organize, and clean all spatially relevant extant cougar, black bear, and mule deer telemetry data. We will obtain hunter harvest data from UDWR to update estimates of cougar abundance and/or population trends using the indices and models detailed in Wolfe et al. (2016). Harvest data include the sex/age class, date, and location of each cougar, which we will use to assess population demography. We will obtain similar harvest records on black bears. We also will use extant cougar GPS telemetry data from five sites, largely restricted to management units in the Central and Southern Regions. These data include ~ 50 previously GPS collared cougars from the Oquirrh, Stansbury, Monroe, Boulder, and Zion units that will be used to calculate baseline information on home ranges, survival, and reproduction. We will also obtain similar data from black bears, for which several sites across the state have radio-collared female bears to calculate baseline information on den sites and reproduction. The abundance and distribution of native prey are the most important predictors of cougar presence and abundance (Pierce et al. 2012). Accordingly, we will work with UDWR to acquire estimates of mule deer abundance and annual fawn counts, by management unit. All existing data will be used to target and design new field efforts. Field-based research We will use field-based research to obtain survival and reproductive data for cougars across three study sites in Utah (Objective 1), evaluate movement and diet of cougars in relation to migratory patterns of ungulates and availability of alternative prey including domestic species (Objective 2), and determine the impacts of black bears on cougar foraging and population parameters (Objective 3). This proposed fieldwork is important for several reasons. First, no cougar GPS data exist from the Northern or Southeastern regions, and this proposed fieldwork will fill that gap. Utah has four distinct ecoregions and at least three of these, including the Northern and Southeastern regions, contain areas of overlap among highly valued populations of cougars, bears, mule deer, and livestock. The northern region is particularly important because it contains the most private land and the highest deer, cougar, and livestock densities statewide. Second, extant data were not collected specifically to answer the questions posed in this proposal and there is currently a mismatch in scales between available and needed data. We selected study sites after extensive conversations with UDWR employees to ensure the sites represented state management issues while meeting study design parameters. These sites include the Book Cliffs, Wasatch/Manti, and Cache management units. The sites represent three of the four ecoregions, have concurrent ungulate monitoring programs, all receive levels of cougar hunting pressure commensurate with the statewide average, and they offer a reasonable representation of the range of environmental and management conditions that UDWR managers are faced with. Further, two factors, black bear population sizes and ungulate seasonal range limitations, vary among sites. These factors provide a quasi-experimental design for comparisons, which is lacking in previous studies. Black bears are abundant in the Book Cliffs and Wasatch/Manti but rare in the Cache. Deer and other ungulates are winter-limited in the Cache and Wasatch/Manti, whereas they are summer limited in the Book Cliffs. We will work with UDWR to obtain data from harvested cougars within our three study sites. This includes tissue sample for genetic analysis, sex, age estimate, and GPS (or map) location of harvest. We will use sex and age estimates to provide baseline data for modeling population dynamics using the techniques described by Wolfe et al. (2016). We will process tissue samples at Karen Mock's laboratory at USU to determine relatedness of individuals within and across units. These data will improve population modeling by adding critical information on population connectivity (e.g., Sinclair et al 2001; Atwood et al. 2011, Andreasen et al. 2012). Further, we can leverage these data for more information -- Idaho Fish and Game started a genetic mark-recapture monitoring program of cougars within the units that abut our northern field site. At the 2017 Mountain Lion Workshop, I spoke with the program lead and he is willing to share genetic data with our project. Because the Bear River Range within the Cache unit provides contiguous habitat across the Utah-Idaho border, their data will greatly enhance connectivity models at no extra cost to UDWR, and may even reduce the effort needed at our study site to gain the same quantity of data. This will directly address the need for a larger geographic scale highlighted by Sinclair et al. (2001). We will capture and fit GPS-collars on cougars using hounds at all three study sites. The collars will be satellite-based or remote-download, so that information on space use is available in ~ real time. To detect the presence of bears and compile sufficient data to address objectives 2 and 3, real-time location data are needed. Traditionally, researchers search clusters of GPS-locations to identify cougar kills. In many cases, these sites are searched months or even a year later, once the collar drops off a cougar and data are downloaded. Small prey items including neonate ungulates and lambs can be completely consumed within 1-2 days, leaving very little evidence (Mitchell 2013). Our focus is to both to evaluate where cougars kill domestic and native animals at a fine scale (e.g., Kauffman et al. 2007) and parameterize habitat and kill site selection among demographic classes at a larger scale (e.g., Marucco and McIntire 2010). At a subset of the domestic prey carcasses we will also use detection dogs to thoroughly investigate the area to detect if any other sheep were killed but not found. This will provide UDWR with tools for calculating livestock loss when discussing management strategies with regional advisory councils and setting the depredation multiplier. To detect the presence of bears and compile sufficient data to address objectives 2 and 3, real-time location data are needed. First, dietary information will be obtained by sampling remains found at GPS clusters identified using Animal Site Fidelity (rASF) in program R (Mahoney and Young 2016). Then, we will set camera traps at new kill sites of native prey to observe cougars that use the carcasses and observe any potential interactions or displacement events related to black bears. We will employ these methods when possible to obtain an appropriate sample, but we recognize we will not be able to do this for all kills throughout the study. We will use location data from collared cougars for several other purposes relevant to our objectives. For one, we will be able to monitor survival of collared cougars and, because of the use of satellite-collars, obtain cause-specific mortality. Second, GPS-collars will assist with monitoring cougars with dependent kittens and facilitate capture and collaring of kittens. To adequately address kitten and juvenile survival, we will need to mark kittens as early as possible, and recapture as needed to follow until dispersal. We will follow well-established techniques to capture and monitor kittens (Clark et al. 2015, Cooley et al. 2009). Briefly, at young ages (< 2 months) cougar kittens can be safely and effectively captured at nursery sites without use of immobilizing drugs. Capture of kittens at nursery sites does not negatively affect survival or result in abandonment (Logan and Sweanor 2001). Throughout capture procedures, biologists wear sterile gloves to minimize transfer of human scent to kittens. After capture, kittens are restrained in a breathable bag (e.g., burlap sack) to minimize stress. Kittens are weighed, gender determined, and an estimate of age obtained using pelage spotting progression (Shaw 1986). One ear of a kitten is permanently marked with an ear tattoo applied with a small animal tattoo kit. Prior to administering an ear tattoo, the ear of the kitten is sterilized with an alcohol wipe. Kittens > 4 weeks of age can have a temporary, expandable radio-collar attached. These collars were developed and deployed during research projects in Colorado and have minimal to no effects of survival of kittens nor do they result in injuries to kittens (K. Logan, Colorado Division of Wildlife, personal communication). They have also successfully been used in Oregon (B. Orning, personal communication, Clark et al. 2015). The collars incorporate an expandable joint (made of elastic) that allows the collar to expand over time as the kitten grows. The elastic joint represents a weak point in the collar that will fail over time, allowing the collar to fall off the cougar after approximately 6-12 months (K. Logan, Colorado Division of Wildlife, personal communication). Collars fitted to young kittens (~ 4 weeks old; 3 kg [Laundré and Hernández 2002]) weigh approximately 70 g. Transmitters attached to young kittens will represent at most 2.3% of the individual's body mass, which is below suggested guidelines for mammals (5-10% of body mass; Wilson et al. 1996). After handling, kittens are placed in the nursery site where they were initially captured. Finally, we will also use GPS-collar data to estimate home range size in our study area and determine if home range size or movement patterns shift in response to migration patterns of native prey (e.g. Pierce et al. 1999). We will likely use resource-selection function (RSF) models framework so that other relevant factors can be accounted for (Manly et al. 2007). Many native ungulates within our study sites are fitted with GPS-collars and monitored as part of the migration studies conducted by UDWR. With no additional cost to this study, we will use those data to inform models of space use by cougars (Stoner et al. 2013a). For similar information on domestic prey, we will work with UDWR biologists, USDA Wildlife Services, other agencies, and ranchers to map locations of livestock herds. While this technique does not permit a fine-scale evaluation of movement, we anticipate a high-level of accuracy in the location data. Dr. Young has worked with ranchers across five states on another study of livestock guard dogs. In that study, sheep were fitted with GPS tags but (at a larger scale) the information on livestock space use gained from the collars matched that of the herders and ranchers. We will utilize UDWR's black bear monitoring program to gain additional information on black bear-cougar dynamics. Where possible, we will replace VHF collars on already-monitored black bears with refurbished or new GPS-satellite collars in conjunction with the migration program. This will increase data quantity and quality to evaluate whether black bears affect foraging patterns of cougars at sites with and without high densities of bears. The southeastern units suffer from chronically low fawn survival (Smith 1983), while the central units have higher livestock depredation by bears, suggesting native prey may be limited or difficult to capture. These units also have the highest bear densities, relatively low cougar harvest, and a large amount of refugia (Stoner et al. 2013b). Thus, the primary question of interest here is the degree to which predation may underlie low fawn survival, and whether bear predation is additive to cougar predation and how that may affect management decisions related to cougars.

Animals will be fitted with GPS collars and monitored for the duration of the study.

Utah State University

Management implications Products derived from this research will include estimates of cougar abundance and growth rates, understanding of how ungulate migration affects cougar space use and kill rates of native prey, and knowledge of how black bears influence cougar space use and kill rates of native prey and how that can be used to inform population estimates of cougars and subsequent management decisions. Cougar management has been hindered by the lack of monitoring techniques independent of harvest data. Results of this effort will provide the UDWR with expected cougar density estimates on a management unit basis. When used in conjunction with current standard survey protocols (Wolfe et al 2016), UDWR personnel can measure deviation from this baseline to track population trends through time. Management implications of these products are better estimates and monitoring cougar populations and their impacts on mule deer and alternative prey. This study will adequately predict cougar habitat use and therefore ungulate vulnerability to predation. The product from this effort will benefit the Mammals Program by updating the current distribution and seasonal habitat maps for cougar prey. It will benefit the Big Game Program for the same reasons and ultimately benefit sportsmen. The assessment of livestock vulnerability to predation will be a value-added product that will enable DWR managers to predict when (seasonal variation), where (spatial distribution), and how many (multiplier data) livestock are susceptible to cougar depredation. To the extent that depredation can be reduced, more cougars can be removed through sport hunting rather than by Wildlife Services. A clear benefit to the UDWR, sportsmen and livestock producers. Conservation benefits More accurate estimates of cougar populations, population growth rates, and the impacts of cougars on native prey can better inform harvest management decisions that help maintain viable populations of cougars and their prey. By including information on alternative factors that may influence cougar space use and predation patterns, such as migration of native and domestic prey and effects of black bears, this study takes a broad scale ecological approach that is rare but recognized as critical (Sih et al. 1998; Creel et al. 2017; Tallian et al. 2017). Results will therefore not only be useful to UDWR management of cougars, but also place UDWR as leaders in applied cougar research. Thus, this study should allow greater hunting opportunities for the public while placing UDWR at the forefront of large carnivore ecology and management. Moreover, it will provide defensible monitoring criteria to address criticisms from various user-groups. Expected outcomes Expected outcomes include an improved index to monitor cougar populations and management recommendations. Specific outcomes include estimates of cougar population parameters such as reproduction, survival, immigration, predation rates and competition with bears.

The livestock portion of the study could provide data that can inform Utah's compensation program and may help build tolerance for large carnivores on the landscape. It may also help producers limit predation by avoiding likely times and locations of predation events.