https://cpw.cvlcollections.org/files/original/16d7b5c86998cf9898fc6db9eeb7c126.pdf 8e5fde525f6d2166dcae01b0c0707d3b PDF Text Text The research in this publication was partially or fully funded by Colorado Parks and Wildlife. Dan Prenzlow, Director, Colorado Parks and Wildlife • Parks and Wildlife Commission: Marvin McDaniel, Chair • Carrie Besnette Hauser, Vice-Chair Marie Haskett, Secretary • Taishya Adams • Betsy Blecha • Charles Garcia • Dallas May • Duke Phillips, IV • Luke B. Schafer • James Jay Tutchton • Eden Vardy �The Journal of Wildlife Management 82(6):1135–1148; 2018; DOI: 10.1002/jwmg.21476 Research Article Mortality of Mule Deer Fawns in a Natural Gas Development Area MARK E. PETERSON,1,2 Department of Fish, Wildlife, and Conservation Biology, Colorado State University, 1474 Campus Delivery, Fort Collins, CO 80523, USA CHARLES R. ANDERSON, JR., Mammals Research Section, Colorado Parks and Wildlife, 317 W. Prospect Road, Fort Collins, CO 80526, USA JOSEPH M. NORTHRUP,3 Department of Fish, Wildlife, and Conservation Biology, Colorado State University, 1474 Campus Delivery, Fort Collins, CO 80523, USA PAUL F. DOHERTY, JR., Department of Fish, Wildlife, and Conservation Biology, Colorado State University, 1474 Campus Delivery, Fort Collins, CO 80523, USA ABSTRACT Recent natural gas development has caused concern among wildlife managers, researchers, and stakeholders over the potential effects on wildlife and their habitats. Specifically, understanding how this development and other factors influence mule deer (Odocoileus hemionus) fawn (i.e., 0–6 months old) mortality rates, recruitment, and subsequently population dynamics have been identified as knowledge gaps. Thus, we tested predictions concerning the relationship between natural gas development, adult female, fawn birth, and temporal (weather) characteristics on fawn mortality in the Piceance Basin of northwestern Colorado, USA, from 2012–2014. We captured and radio-collared 184 fawns and estimated apparent cause-specific mortality in areas with relatively high or low levels of natural gas development using a multi-state model. Mean daily predation probability was similar in the high versus low development areas. Predation was the leading cause of fawn mortality in both areas and decreased from 0–14 days old. Black bear (Ursus americanus; 22% of all mortalities, n ¼ 17) and cougar (Felis concolor; 36% of all mortalities, n ¼ 6) predation was the leading cause of mortality in the high and low development areas, respectively. Predation of fawns was negatively correlated with the distance from a female’s core area to a producing well pad on winter or summer range. Contrary to expectations, predation of fawns was positively correlated with rump fat thickness of adult females. Well pad densities and development activity were relatively low during our study, indicating that the observed intensity of development did not appear to influence daily predation probability. Our results suggest maintaining development activity thresholds at levels we observed to potentially minimize the effects of development on fawn mortality. However, we caution that higher development intensity and drilling activity in flatter, less rugged areas with less concealment cover could influence fawn mortality. Managers should maintain low development densities in areas where topography and vegetation offer less concealment. Overall, region-specific data (e.g., development intensity, topography, predator assemblages, and associated predation risk) are needed to better understand the effects of natural gas development on fawn mortality. Ó 2018 The Wildlife Society. KEY WORDS anthropogenic disturbances, Colorado, mule deer, natural gas development, Odocoileus hemionus, population dynamics, predation, survival. Wildlife managers, researchers, and public stakeholders have heightened concern about the potential effects of natural gas development on wildlife populations and their habitats (Walker et al. 2007, Webb et al. 2011a, Christie et al. 2015). Effects from development on mule deer (Odocoileus hemionus) populations and their habitat are of particular interest Received: 18 August 2017; Accepted: 3 March 2018 1 E-mail: mark.peterson313@gmail.com Present address: South Dakota Department of Game, Fish and Parks, 4130 Adventure Trail, Rapid City, SD 57702, USA 3 Present address: Wildlife Research and Monitoring Section, Ontario Ministry of Natural Resources and Forestry, 2140 East Bank Drive, Peterborough, ON, Canada K9L 1Z8 2 Peterson et al. � Mule Deer Fawn Mortality because of the species’ recreational, social, and economic importance as a game species (Sawyer et al. 2006, Lendrum et al. 2012, Northrup et al. 2015). Despite these concerns with natural gas development, relatively little research has focused on how development influences demography and population dynamics (Johnson et al. 2016, Sawyer et al. 2017), and there has been no research on the influences of development on the survival of mule deer fawns (i.e., 0–6 months old). Estimating mule deer fawn mortality rates and evaluating cause-specific mortality is useful for modeling population dynamics and to guide management in areas where disturbances such as natural gas development are occurring. Ungulate fawn survival is typically low and variable, consequently annual variation in survival and recruitment of this age class can strongly influence 1135 �population dynamics (Gaillard et al. 1998, 2000; Forrester and Wittmer 2013). Yet, few studies have examined mule deer fawn survival even though this period may be when most mortality occurs in the species (Pojar and Bowden 2004, Bishop et al. 2009, Hurley et al. 2011, Monteith et al. 2014). Natural gas development is increasing worldwide and projections show an increasing trend into the future (United States Energy Information Administration 2017), but research examining fawn mortality in natural gas development areas is lacking. Past research provides insight into natural gas development factors that can influence mule deer behavior (Sawyer et al. 2006; Northrup et al. 2015, 2016) and possibly fawn mortality. This research suggests contrasting forces of development on mule deer, with potential influences of development on mule deer forage, mortality risk, and perceived mortality risk. Fawn mortality may be negatively influenced by the distance an adult female’s core area (i.e., 50% home range area) is from natural gas development and roads because of increased noise and human presence in these areas. However, the mechanism linking development to fawn mortality is unknown. Some studies suggest deer generally use habitat farther from roads (Rost and Bailey 1979, Webb et al. 2011b, Lendrum et al. 2012) and well pads (Sawyer et al. 2006, 2009; Northrup et al. 2015). However, others suggest deer use habitat closer to well pads and during spring migration because disturbed topsoil near well pads possibly provides enhanced foraging opportunities during the early growing season (Webb et al. 2011b, Lendrum et al. 2012). Increased human presence associated with development might provide a refuge from predators (Berger 2007, Dussault et al. 2012), although deer might perceive areas close to development as risky because of the relatively high levels of noise and human activity (Frid and Dill 2002, Lynch et al. 2014). Consequently, fawn mortality could be positively or negatively influenced by distance from development. Adult female characteristics (e.g., body condition, fetal production, and age) have potential to strongly influence fawn mortality (Lomas and Bender 2007, Bishop et al. 2011) and also may be influenced by development activity. Maternal nutrition plays a fundamental role in the reproductive success of deer (Johnstone-Yellin et al. 2009, Parker et al. 2009, Tollefson et al. 2011) and nutrition is an important determinant of adult female body condition (Robinette et al. 1973). Availability of high quality forage is necessary to support adult body condition and fetal and fawn growth (Keech et al. 2000, Tollefson et al. 2011), particularly during the critical periods of the last trimester and lactation (Robbins and Robbins 1979). However, natural gas development may alter availability of nutritious forage on winter and summer range and subsequently influence body condition of females and fawn mortality (Cook et al. 2004, Monteith et al. 2014, Shallow et al. 2015). Litter size also can influence fawn mortality because twins or triplets weigh less than singletons (Robinette et al. 1973, 1977; Monteith et al. 2014), increasing the risk of starvation for lighter fawns from larger litters. In addition, twins and triplets tend to be more active than singletons, potentially increasing their predation risk (Riley 1136 and Dood 1984, Johnstone-Yellin et al. 2009). Finally, maternal age also can influence fawn mortality; white-tailed deer (Odocoileus virginianus) offspring produced by primeaged and older females (3–10 years old) versus younger females had lower mortality because of improved rearing skills, antipredator behavior, and selection of prime habitat (Ozoga and Verme 1986; Grovenburg et al. 2009, 2012). Fawn mortality also can be influenced by individual fawn characteristics including mass, age, and date of birth. For instance, mass can interact with age (Lomas and Bender 2007, Bishop et al. 2009) and date of birth (Testa 2002) to influence survival; heavier and older fawns tend to have higher survival. Fawns are most vulnerable to mortality events from birth to 8 weeks old (Lomas and Bender 2007, Monteith et al. 2014, Shallow et al. 2015). Earlier date of birth may allow adult females access to improved forage conditions during the early growing season (Parker et al. 2009, Lendrum et al. 2014), which increases fawn growth and strength to elude predators (Testa 2002). However, predator swamping can decrease mortality for peak-born fawns and increase mortality of late-born fawns because predators need time to develop a search image to prey on vulnerable fawns during a birth pulse (Whittaker and Lindzey 1999, Petroelje et al. 2014). Thus, birth characteristics and subsequent mortality are also influenced by habitat and adult female body condition (Monteith et al. 2014, Shallow et al. 2015). Temporal (weather) characteristics, namely winter precipitation (i.e., during the season before parturition), summer precipitation, and temperature can influence fawn mortality. Precipitation can indirectly affect maternal condition and subsequent fawn birth mass and mortality through forage growth and quality (Lomas and Bender 2007, Monteith et al. 2014, Shallow et al. 2015). Additionally, summer ambient temperatures can influence fawn mortality if exposure to cold, wet weather occurs shortly after birth (Gilbert and Raedeke 2004, Hurley et al. 2011). Our objectives were to assess cause-specific mortality and test predictions about how natural gas development, adult female, fawn birth, and temporal characteristics influence fawn mortality. We addressed these objectives in 2 study areas, one with relatively low development activity and another with relatively high development activity. We predicted that predation would be the primary proximate cause of fawn mortality and mortality would be higher in the high versus low natural gas development study areas (because there would be less cover and more fragmented habitat in the high development area making fawns more vulnerable). More specifically, we tested the predictions that fawn mortality would be higher closer to producing well pads and roads (because of less cover and increased mortality risk); higher for younger and lighter fawns (because of their reduced ability to evade predators and reduced vigor); and higher for late-born (because of a decreased growth rate), male (because they are more active), and twin fawns (because they would likely have lower mass and be more active). We also predicted mortality would be higher for fawns produced by females with decreased rump fat thickness (i.e., body The Journal of Wildlife Management � 82(6) �condition) and younger (�3.5 years old) females because they would likely have poor rearing skills. Further, we predicted that increased previous winter precipitation, and decreased summer precipitation, and temperature would increase fawn mortality because of decreased body condition, decreased forage and cover, and increased thermoregulation, respectively. Ultimately, we hypothesized that natural gas development would negatively influence mule deer fawn mortality because of changes in behavior and habitat structure modification (e.g., fragmentation, removal of concealment cover). STUDY AREA During 2012–2014, we examined fawn mortality in the Piceance Basin, northwest Colorado, USA (39.7898 N, �108.3298 W and 40.0918 N, �108.1418 W). Dominant human activity in the area included natural gas extraction, ranching, and hunting during the fall. The area varied topographically, ranging from drainage bottoms that have been converted to irrigated, grass fields to moderately steep canyons and high mountains. Large herbivores in the Piceance Basin included mule deer, elk (Cervus canadensis), cattle, and wild horses. Predators included the black bear (Ursus americanus), bobcat (Lynx rufus), cougar (Puma concolor), coyote (Canis latrans), golden eagle (Aquila chrysaetos), and bald eagle (Haliaeetus leucocephalus). Our winter range study area included 4 study units in the Piceance Basin (Fig. 1) that were part of a larger research project (Anderson 2016). Deer occupied 2 winter range study units with relatively high levels of natural gas development (0.60–0.90 well pads/km2) and 2 winter range study units with low levels of development (0–0.10 well pads/km2). Average deer density ranged from 8.60–9.30 deer/km2 and 10.10–19.30 deer/km2 on the high and low development winter range units, respectively (Anderson 2016). Predator density in the study units was unknown. The winter range area was dominated by two-needle pinyon pine (Pinus edulis) and Utah juniper (Juniperus osteosperma) woodlands and mountain shrublands. Winter study unit elevations ranged from 1,860 m to 2,250 m. During our study, winter (Oct–Apr; Hurley et al. 2011) precipitation averaged 22.4 cm and mean winter temperatures ranged from �148 C to 148 C at the Rifle 23 NW weather station (National Climatic Data Center 2015; Fig. 1). Deer from winter range units with high levels of natural gas development generally migrated to the Roan Plateau summer range (Lendrum et al. 2013) where deer potentially encountered natural gas development (0.04–0.06 well pads/ km2; Fig. 1). Hereafter we refer to deer inhabiting the winter and summer range study units with relatively high levels of development as being in the high development study area. Deer from winter range units with low levels of natural gas development generally migrated towards the Flat Tops Wilderness Area summer range (Lendrum et al. 2013) and encountered minimal natural gas development (0–0.01 well pads/km2; Fig. 1). Hereafter we refer to deer inhabiting the winter and summer range study units with relatively low levels of development as being in the low development study area. Summer range for both groups of deer were dominated by Gambel oak (Quercus gambelii), quaking aspen (Populus tremuloides), two-needle pinyon pine, Utah juniper, and mountain shrublands. Dominant vegetation types were Figure 1. Mule deer winter and summer range study areas and National Climatic Data Center (NCDC) weather stations in the Piceance Basin, Colorado, USA, 2012–2014. Peterson et al. � Mule Deer Fawn Mortality 1137 �interspersed with Douglas-fir (Pseudotsuga menziesii), Engelmann spruce (Picea engelmannii), and subalpine fir (Abies lasiocarpa) forests (Garrott et al. 1987). Shrubs, forbs, and grasses common to the area are listed in Bartmann (1983). Summer study unit elevations ranged from 1,900 m to 3,150 m. During our study, summer (May–Sep) precipitation averaged 20.3 cm and mean summer temperatures ranged from 28 C to 318 C at the Rifle 23 NW or Hunter Creek weather station (Fig. 1) depending on available weather data (National Climatic Data Center 2015). The summer period encompassed the growing season of the area. METHODS Capture and Handling During December 2011–2013, a helicopter crew captured adult (�1.5 years old) female mule deer in each of the 4 winter range study units using a net gun (Webb et al. 2008, Jacques et al. 2009). The helicopter crew blindfolded and hobbled deer, and chemically sedated them with 0.5 mg/kg of Midazolam and 0.25 mg/kg of Azaperone given intramuscularly. For each captured deer, we estimated age by examining tooth wear (Robinette et al. 1957). We fit each captured deer with a global positioning system (GPS) radiocollar (Model G2110D, Advanced Telemetry Systems, Isanti, MN, USA). We used ultrasonography to measure thickness of subcutaneous rump fat for each adult female using a portable ultrasound machine (Cook et al. 2001, Stephenson et al. 2002). During March 2012–2014, a helicopter crew recaptured radio-collared adult females on winter ranges using helicopter net gunning. We recorded rump fat thickness as described above and performed transabdominal ultrasonography to determine pregnancy status and in utero fetal count (Bishop et al. 2007). If an adult female was pregnant, we inserted a vaginal implant transmitter (VIT; Model M3930, Advanced Telemetry Systems) following VIT insertion procedures described in detail by Bishop et al. (2011) and Peterson (2016). During parturition (late May–mid-Jul), we monitored radio-collar and VIT signals daily from a fixed-wing aircraft. Once VITs were expelled, ground crews used radiotelemetry to simultaneously locate the radio-collared female and VIT and searched for fawns and a birth site near (�400 m) the female and expelled VIT for up to 1 hour. Crews identified birth sites based on detection of fawns and VIT at a birth site or by characteristics known to typify mule deer birth sites (Barbknecht et al. 2011, Bishop et al. 2011, Rearden et al. 2011). During 2012 and 2013, ground crews captured fawns by hand and located birth sites in the high and low development study areas. During 2014 crews captured fawns and located birth sites predominantly in the high development study areas and sporadically in the low development study areas because VIT photo sensors malfunctioned. We focused our effort in the high development areas because it was logistically more difficult to monitor deer from the ground and capture fawns in the low development areas because of 1138 the spatial distribution of deer and land access issues. Crews handled each captured fawn with nitrile gloves, blindfolded them, and placed them in a cotton bag to measure body mass (�0.10 kg). Crews measured hind foot length (�0.50 cm), sexed each fawn, and estimated age primarily based on VIT expulsion date and secondarily based on hoof characteristics, condition of the umbilical cord, pelage, and behavior (Haugen and Speake 1958, Sams et al. 1996). Crews fit each fawn with a very high frequency radio-collar (Model M4210, Advanced Telemetry Systems) equipped with an 8-hour mortality sensor. Crews modified radio-collars for temporary attachment by cutting the collar in half and splicing the ends with 2 lengths of rubber surgical tubing (5.7 cm each) that deteriorated over time, eventually allowing the collar to fall off. All capture, handling, radio-collaring, and VIT insertion procedures were approved by the Institutional Animal Care and Use Committee at Colorado Parks and Wildlife (protocol numbers 17-2008 and 01-2012) and followed guidelines of the American Society of Mammalogists (Sikes et al. 2016). Fawn Monitoring and Cause-Specific Mortality We monitored radio-collared fawns from a fixed-wing aircraft daily during the parturition period, weekly until deer migrated from summer range, and daily from the ground when deer arrived on winter range. We monitored radiocollared fawns from birth until death, until collars were shed, or until the end of the fawn survival period of interest (i.e., 0–6 months old) on 15 December 2012, 2013, or 2014. Daily monitoring during the parturition period, when a majority of mortalities occurred, allowed us to investigate mortalities typically within 24 hour; thus, we classified causes of death as proximate and not ultimate. When we detected a mortality signal, ground crews located the fawn or radio-collar and conducted a mortality site investigation and field necropsy, if possible, to determine cause-specific mortality or scavenging. During the mortality site investigation, crews documented GPS coordinates of the site, predator tracks, predator scat, drag trails, blood, hair, and any other signs (e.g., matted vegetation or broken shrub branches) that could help determine cause-specific mortality or scavenging (Stonehouse et al. 2016). During the field necropsy, crews documented position of carcass, presence of tooth marks and distance between canines, and presence of hemorrhaging, bruising, and fractures. Crews used predation site characteristics and signs of predatory feeding (e.g., canine spacing) reported in the literature to help assign a specific predator (White 1973, Wade and Bowns 1982, Stonehouse et al. 2016). However, there will always be uncertainty in assigning cause of death unless a mortality event was witnessed firsthand, which is extremely rare. Thus, we identified likely causes of death using the following descriptions of predation site characteristics and predatory feeding behavior. We classified proximate causes of mortality into the following categories: black bear predation, bobcat predation, cougar predation, coyote predation, raptor predation, unknown predation, malnutrition, and unknown mortality. The Journal of Wildlife Management � 82(6) �We identified black bear predation based on characteristic canine punctures; massive hemorrhaging around neck, top of cranium, or tops of shoulder; bruising on hind legs; crushed ribs; blood in nose and mouth; hide peeled back to head; bear hair, scat, or tracks; and bear beds typified by a large area of flattened vegetation surrounding a carcass. We identified bobcat and cougar predation based on characteristic canine punctures; hemorrhaging on the back or side of neck or base of skull; broken neck; claw marks on neck or back of shoulders; plucked hair; cached carcass; drag trail marks to a cache site; and felid hair, scat, or tracks. We identified coyote predation based on characteristic canine punctures; hemorrhaging on the neck, throat, or cranium; torn tissue on hind legs; canid hair, scat, or tracks; signs indicating a chase or struggle; blood scattered across the ground or vegetation; patches of loose hair or torn pieces of hide; and buried carcass. Carcasses of neonates killed by coyotes occasionally consisted of detached portions scattered across the site, but we did not rely on this observation alone to confirm coyote predation. We identified raptor predation based on deep penetrating talon wounds in a triangular pattern, an intact carcass, and raptor feathers, scat, or tracks. We identified malnutrition based on an intact carcass with minimal or no fat deposits on the heart and kidneys (Kistner et al. 1980) and femur marrow fat (Cheatum 1949) and no sign of predation, hemorrhaging, disease, or injury. Unknown mortality included cases where we could not attribute death to predation or malnutrition. Mule Deer Core Area Estimation We programmed radio-collars deployed on adult females to release 16 months post-capture. We retrieved collars once released or from mortality sites, and downloaded GPS data following collar recovery. Using helicopter net gunning to capture deer can potentially influence mule deer behavior (Northrup et al. 2014); thus, we censored location data for 4 days following capture. Deer migrate between winter and summer range in this area; thus, we classified winter range locations as those occurring from post-capture to departure from winter range and summer range locations as those occurring from arrival to departure or date of fawn death. We determined departure from winter or summer range as the first location outside the winter or summer range for each female and arrival on summer range as the first location inside summer range for each female. To determine migration patterns, we derived 95% kernel density estimates of winter and summer range for each adult female using the Geospatial Modeling Environment (Beyer 2012). We also derived 50% kernel density estimates of core areas and centroids for each adult female on winter and summer ranges using the Geospatial Modeling Environment (Beyer 2012). Multi-State Mark-Recapture Mortality Analyses and Model Set We analyzed apparent daily mortality probability from parturition until 15 December 2012–2014 using a multistate model (Brownie et al. 1993, Lebreton et al. 2009) as implemented in Program MARK (White and Burnham 1999, White et al. 2006). We considered our encounter data Peterson et al. � Mule Deer Fawn Mortality to be in 1 of 5 states represented by alive in the high development study areas (H), alive in the low development study areas (L), death by predation (K), death by malnutrition (M), or death by unknown mortality (U; Fig. 2). In addition, we assigned each encounter history to 1 of 3 groups represented by 2012, 2013, or 2014. Multi-state models estimate 3 parameters including survival (S), detection (p), and transition probabilities (Lebreton et al. 2009). In our case, we modeled alive and dead states and estimated apparent mortality rates as the transition probability � ^ from an alive to a dead state (Lebreton and Pradel 2002, c Devineau et al. 2010). Specifically, we modeled transitions from alive in the high or low development study areas to death by malnutrition, or unknown mortality � HKpredation, � HM HU LK LM LU ^ ^ ^ ^ ^ ^ c ; c ; c ; c ; c ; c ; Fig: 2 . Because survival is the complement of mortality and we estimated mortality with the transition probabilities, we fixed survival � � rates in the high S H and low S L development area, � � predation S K , malnutrition S M , and unknown mortality � S U states to 1 (Devineau et al. 2010, 2014). Transitions from a dead to an alive state or a dead to a dead state could not happen and we fixed those parameters to zero. In addition, transitions from an alive state in the high development areas to an alive state in the low development areas did not occur and we fixed those parameters to zero. We defined detection probability as the probability that a radio-collared fawn was detected in an alive or mortality state during each monitoring survey. We modeled mortality probability as a function of winter range development, summer range development, adult female, fawn birth, and temporal covariates that we predicted would influence mule deer fawn mortality based on previous mule deer studies (Pojar and Bowden 2004, Bishop et al. 2009, Hurley et al. 2011, Monteith et al. 2014). Winter and summer range development covariates included the mean distance (km) from the centroid of an adult female’s core area to the nearest drilling well pad, producing well pad (http:// cogcc.state.co.us), or road. We classified each well in the high and low development winter and summer ranges as either being drilled or producing natural gas using a procedure described in Northrup et al. (2015). Using the classified well pad data, we calculated mean daily distance (km) from the centroid of each female’s core area to the nearest drilling and producing well pad on their specific winter and summer range study areas. We fit models using a distance threshold model structure (i.e., data truncated beyond a maximum distance) that accounted for distances that have been shown to illicit behavioral responses by adult female deer in this study area (Northrup et al. 2015). Specifically, we truncated data beyond 0.80 km from a drilling well pad and beyond 0.40 km from a producing well pad because these distances can alter habitat selection patterns of deer (Northrup et al. 2015). We also created a road network map by digitizing all roads visible on 2013 National Agricultural Imagery Program imagery and calculated mean distance (km) from the centroid of each female’s core area to the nearest road on their specific winter and summer range study areas. Because 1139 �Figure 2. Multi-state model schematic representing alive and dead states for mule deer fawns in the Piceance Basin, Colorado, USA, 2012–2014. Fawns transitioned from high (H) or low (L) development study areas to a cause-specific death by predation (K), malnutrition (M), or unknown mortality (U) state � � cHK ; cHM ; cHU ; cLK ; cLM ; or cLU . Fawns remained in an alive cHH ; cLL state or were absorbed in a cause-specific death by predation, malnutrition, � KK MM UU state. Fawns were captured at �3 days old and recaptured in an alive state or cause-specific death state or unknown mortality c ; c ; or c � pH ; pL ; pK ; pM ; or pU . development is dynamic, with wells transitioning between drilling and producing, we averaged distances from a female’s core area to well pads. We calculated distance based on a female’s daily status from the capture date to departure from winter range for winter range development covariates using the R statistical software (R Core Team 2015). We calculated distance from summer range development covariates on a fawn’s date of birth. Adult female-specific covariates included rump fat thickness (mm) of females measured in March, in utero fetal count documented in March, and female age estimated in December. Fawn-specific covariates included age (days old), estimated mass at birth (kg), Julian date of birth, and sex (Bishop et al. 2009, Hurley et al. 2011, Monteith et al. 2014). We incorporated fawn age into models by fitting a model that allowed transition probabilities to vary by an age trend from 0–14 days old and constant thereafter. Because we did not capture individuals at birth, we estimated fawn mass at birth by regressing fawn capture mass as a function of age for each year separately (Bishop et al. 2008, 2009) using a linear model in the R statistical software. We represented Julian date of birth as the number of days following the earliest detected birth in each year. Temporal covariates included total precipitation (cm) during the previous winter season before parturition (1 Oct–30 Apr), daily precipitation (cm) and daily temperature (8C) during the parturition period, and the 7-day average of precipitation and temperature after the parturition period until 15 December (Hurley et al. 2011, Monteith et al. 2014). Prior to modeling, 1140 we calculated a pairwise correlation matrix to test for collinearity among covariates (|r| � 0.60) and retained the more biologically plausible covariate if 2 covariates were correlated. We calculated mean, standard deviation, and range of all continuous variables included in multi-state models (Appendix A). We used a 2-stage modeling approach to assess covariate importance and used stage 1 to identify and exclude unsupported covariates from stage 2 (Arnold 2010; Monteith et al. 2011, 2014). For stage 1, we conducted separate mortality analyses for winter range development, summer range development, adult female, fawn birth, or temporal covariates while holding detection probabilities constant. We also ran a separate analysis where we modeled detection probability as a function of year and migration (i.e., different probability before and after autumn migration) while holding transition probabilities constant. Because of memory limitations in Program MARK, we fit all possible combinations of additive models (Doherty et al. 2010) with a maximum of 6 winter range development, 3 summer range development, 4 adult female, 3 fawn, and 4 temporal covariates (Tables S1–S5, available online in Supporting Information). For the detection probability analysis, we fit all possible combinations of additive and interactive models (Doherty et al. 2010, Table S6). For each analysis, we calculated the sum of Akaike’s Information Criterion adjusted for small sample size (AICc) weights for models containing each covariate of interest (Burnham and Anderson 2002). We considered covariates with a cumulative The Journal of Wildlife Management � 82(6) �AICc weight �0.50 as important (Barbieri and Berger 2004) and retained these variable for stage 2 of model selection (Appendix B). For stage 2 of model selection, we fit all possible combinations of additive models (Doherty et al. 2010, Table S7) and calculated cumulative quasi-likelihood using Akaike’s Information Criterion adjusted for small sample size (QAICc) weights to help identify important variables (Burnham and Anderson 2002). Following suggestions of Barbieri and Berger (2004), we constructed a prediction model that contained all covariates with a cumulative QAICc weight �0.50. Unless otherwise noted, we used the prediction model when presenting estimates. We attempted to capture and radio-collar all fawns documented in utero for each radio-collared adult female. This potentially caused overdispersion because of sibling dependence; thus, we followed methods described by Bishop et al. (2008) to test for overdispersion ð^c Þ. We conducted a bootstrap analysis in Program MARK by resampling litters of adult females (n ¼ 117) instead of individual fawns (n ¼ 183) to generate 1,000 replicates (Bishop et al. 2008). We used the most parametrized model from stage 2 of model selection described above for the bootstrap and calculated the mean and standard deviation for each of the 1,000 mortality estimates using mean covariate values. The dependence among litters is reflected in the standard deviation of the mortality estimates and yielded an empirical sampling variance estimate. We estimated overdispersion by dividing the empirical (i.e., bootstrap) estimate of standard deviation by the theoretical (i.e., observed) standard error from the mortality estimate of the top model. If the estimate of ð^c Þ was >1.00, we adjusted ð^c Þ in Program MARK and calculated QAICc weights (Burnham and Anderson 2002). RESULTS During March 2012–2014, we inserted VITs in 331 pregnant females. During 29 May–30 June 2012–2014, we captured and radio-collared 128 (2012, n ¼ 61; 2013, n ¼ 33; 2014, n ¼ 34) and 56 fawns (2012, n ¼ 20; 2013, n ¼ 31; 2014, n ¼ 5) in the high and low development study areas, respectively. We captured fawns from 85 (43 or 42 with 1 or 2 collared fawns) and 33 (21, 11, or 1 with 1, 2, or 3 collared fawn) females in the high and low development study areas, respectively. In the high and low development study areas, mortality was attributed to black bear predation (n ¼ 17 and 5), cougar predation (n ¼ 9 and 6), coyote predation (n ¼ 10 and 1), bobcat predation (n ¼ 1 and 4), felid predation (n ¼ 3 and 1), raptor predation (n ¼ 1 and 0), unknown predation (n ¼ 18 and 5), malnutrition (n ¼ 4 and 3), vehicle (n ¼ 1 and 0), and unknown mortality (n ¼ 13 and 5). We censored 3 fawns from the mortality analyses because their deaths were related to capture and right-censored 12 additional fawns during the study because of prematurely shed radio-collars. Cause-Specific Mortality of Fawns We estimated ð^c Þ as 1.04 � 0.15 (SE) and assessed relative importance of each covariate for predicting probability of predation using cumulative QAICc weights (Table 1). Rump fat thickness of adult females, distance (0–0.40 km) an adult female’s core area was from a producing well pad on winter or summer range, and a 14-day fawn age trend and constant thereafter all had a cumulative QAICc weight > 0.50 (Table 1), suggesting support for influencing predation of fawns. The daily predation probability of fawns increased as ^ ¼ 0.20, 95% CI female rump fat thickness increased (b ¼ 0.07 to 0.32). In addition, predation of fawns decreased as the distance from a female’s core area to a producing well pad ^ ¼ �2.14, 95% CI ¼ �4.21 to �0.06) or on winter (b ^ summer (b ¼ �6.22, 95% CI ¼ �10.59 to �1.84; Fig. 3) range increased from 0–0.40 km, and decreased as fawn age ^ ¼ �0.05, 95% CI ¼ �0.10 increased from 0–14 days old (b to �0.01). Overall, predation was the leading cause of fawn mortality in both areas and mean daily predation probability Table 1. Cumulative weights for quasi-likelihood Akaike’s Information Criterion adjusted for small sample size (QAICc), for all variables included in the second stage of analysis of mule deer fawn mortality in the Piceance Basin, Colorado, were probability of transitioning from an � USA, 2012–2014. Parameters � alive state in the high or low development study areas to a death by predation c:K or malnutrition c:M state and detection probability (p). We considered cumulative weights >0.50 to be important. Parameter Predation c � :K Malnutrition c:M � Detection probability (p) a Covariatea Cumulative QAICc weight Rump fat Summer distance to well pad Fawn age 0–14 days old Winter distance to well pad Temperature Fawn age 0–14 days old Winter distance to road Temperature Year Migration Year � migration 0.96 0.89 0.71 0.68 0.41 0.88 0.68 0.61 1.00 1.00 1.00 Covariates include rump fat thickness (mm) of adult females measured in March (rump fat), distance (km) from nearest producing well pad on summer range with data truncated beyond 0.40 km (summer distance to well pad), fawn age trend from 0–14 days old and constant thereafter (fawn age 0–14 days old), distance (km) from nearest producing well pad on winter range with data truncated beyond 0.40 km (winter distance to well pad), daily temperature (8C) during the parturition period and 7-day average of temperature after the parturition period until 15 December (temperature), distance (km) from nearest road on winter range (winter distance to road), each year of the study (year), and before and after autumn migration (migration). Peterson et al. � Mule Deer Fawn Mortality 1141 �DISCUSSION Figure 3. Estimated daily predation probability (�95% CI) of mule deer fawns as a function of distance an adult female’s core area was from a producing well pad on summer range in the high and low development study areas in the Piceance Basin, Colorado, USA, 2012–2014. of fawns was similar between the high and low development areas ranging from 0.011 � 0.003 to 0.012 � 0.002 (Fig. 4). We also assessed relative importance of each covariate for predicting probability of death by malnutrition using cumulative QAICc weights (Table 1). A 14-day fawn age trend and constant thereafter, distance an adult female’s core area was from a road on winter range, and temperature all had a cumulative QAICc weight > 0.50 (Table 1), suggesting support for influencing death by malnutrition. The daily probability of death by malnutrition decreased as fawn age ^ ¼ �0.18, 95% CI ¼ �0.36 increased from 0–14 days old (b to 0.01), increased as the distance from a female’s core area to ^ ¼ 2.17, 95% CI ¼ 0.35 to a road on winter range increased (b ^ ¼ �0.12, 4.00), and decreased as temperature increased (b 95% CI ¼ �0.25 to 0.01), but this covariate was weakly supported. Overall, mean daily probability of death by malnutrition was the same between the high and low development areas ranging from 0.001 � 0.001 to 0.002 � 0.003 (Fig. 4). Lastly, variation in detection probability was best explained by an interaction between year and an autumn migration effect (cumulative QAICc weight ¼ 1.00; Table 1). Detection probability ranged from 0.93 � 0.01 to 0.99 � 0.003 before migration and from 0.52 � 0.04 to 0.81 � 0.04 after migration (Table 2). Figure 4. Mean daily probability of death by predation, malnutrition, or unknown mortality (�95% CI) of mule deer fawns from 0–6 months old in the high and low development study areas in the Piceance Basin, Colorado, USA, 2012–2014. 1142 Contrary to our prediction, predation of fawns was similar between the high and low development areas, indicating that at the investigated intensity of development, the resulting landscape modification did not appear to influence predation risk to fawns. However, well pad densities and development activity thresholds were relatively low during our study and thus might be below levels that could strongly influence predator-prey relationships. Directional drilling (i.e., drilling multiple wells from 1 well pad) in the Piceance Basin resulted in low to moderate well pad densities relative to coal-bed methane development in Wyoming (Sawyer et al. 2013) and the most disruptive phase of development, drilling of wells, was minimal during our study. Thus, the influence of development on predation of fawns may be stronger with higher intensity (e.g., coalbed methane) development and increased drilling activity. Ultimately, true before-aftercontrol-impact studies (Manly 2001) are needed in areas with moderate to intense well pad density and drilling activity to better understand the effects of natural gas development on fawn mortality at higher activity levels. Predation was the primary cause of mule deer fawn mortality in the high and low development areas but decreased as fawn age increased from 0–14 days old. However, the dominant predator varied by study area, with black bear predation the leading cause of fawn mortality in the high development areas (22% of all mortalities) compared to cougar predation in the low development areas (36% of all mortalities). Fawns �14 days old rely on a hiding strategy with cryptic coloration and sedentary behavior to minimize predation risk (Walther 1965, Lent 1974, Geist 1981). Consequently, an annual birth pulse of fawns provides predators with an eruption of vulnerable prey after predators develop a search image (Whittaker and Lindzey 1999, Testa 2002, Petroelje et al. 2014). Bears prey on mule deer fawns during the first few weeks after birth when fawns are most vulnerable (Monteith et al. 2014, Marescot et al. 2015, Shallow et al. 2015), but hiding cover can reduce predation (Panzacchi et al. 2010, Shallow et al. 2015). Patchy habitat further fragmented by development possibly contributes to reduced hiding cover or increased edge effects (e.g., abrupt change in habitat between treed and open areas) in the high development area, leading to increased predation. However, our results suggest that predation of fawns was similar between the high and low development areas. Our result of age-specific vulnerability to predation is similar to other studies examining mortality of mule deer fawns (Bishop et al. 2009, Hurley et al. 2011, Monteith et al. 2014, Shallow et al. 2015). In contrast to our results, coyote predation (Whittaker and Lindzey 1999, Bishop et al. 2009, Hurley et al. 2011) or malnutrition (Pojar and Bowden 2004, Lomas and Bender 2007) has been reported to be the primary cause of fawn mortality in other studies. Of note, we monitored fawns weekly instead of daily after mid-July. Weekly surveys increased the time between field necropsies and increased the likelihood of unknown mortality. Consequently, we suspect many unknown mortalities of fawns The Journal of Wildlife Management � 82(6) �Table 2. Estimated detection probability before (psummer) and after autumn migration (pwinter), associated standard error (SE), and upper and lower 95% confidence limits (CL) of mule deer fawns in the Piceance Basin, Colorado, USA, 2012–2014. Year Parameter 2012 2012 2013 2013 2014 2014 psummer pwinter psummer pwinter psummer pwinter Estimate SE Lower 95% CL Upper 95% CL 0.95 0.56 0.99 0.52 0.93 0.81 0.01 0.04 0.003 0.04 0.01 0.04 0.94 0.48 0.98 0.43 0.91 0.72 0.96 0.64 0.99 0.60 0.95 0.88 during the daily period (22%) were from predation, particularly by bears when fawns were <4 weeks old, because the fawns had reduced mobility (Steigers and Flinders 1980, Riley and Dood 1984). More mortalities were attributed to unknown causes during the weekly monitoring period (66%) than the daily monitoring period (23%), but some mortality events might have been from cougar predation when fawns were >4 weeks old as fawns increase activity with limited agility to evade predators (Riley and Dood 1984, Lingle and Pellis 2002). Bear predation likely declined during weekly surveys, but felid and coyote predation may have been similar to that observed during daily surveys. Although we did not see an overall influence of development on predation of fawns, predation of fawns was negatively correlated with the distance from a female’s core area to a producing well pad on winter or summer range as predicted. Deer can temporarily alter their behavior to use habitat closer to producing well pads during the night (Northrup et al. 2015), which is speculated to provide foraging benefits (Webb et al. 2011b, Lendrum et al. 2012). Deer foraging in openings closer to producing well pads and associated pipelines could positively influence maternal nutrition and condition and subsequently the birth mass and growth rate of fawns (Lomas and Bender 2007, Monteith et al. 2014, Shallow et al. 2015). However, foraging in openings can increase predation risk of hiding fawns (Rearden et al. 2011) in sparse cover, especially at night when deer are primarily feeding and predators are generally active (Rogers 1970, Anderson and Lindzey 2003). Thus, habitat closer to producing well pads could be beneficial for adult female condition, but detrimental to fawns, especially at night. However, deer reduce habitat selection in this area within 200 m of producing well pads at night on winter range (Northrup et al. 2015), potentially limiting access to refuge from predators. Whether similar behavioral processes or other unknown processes influence fawn mortality near producing well pads on summer range is unknown, but we presume the process is different on summer range in our study system. Thus, we recommend continued monitoring across a range of different development scenarios (e.g., development intensity and drilling activity) to better understand the effects of natural gas development on fawn mortality. We recognize that our study design was unbalanced, but this fact was unavoidable because of VIT failures during 2014 and the unpredictability of capturing wildlife over large areas. The unbalanced design could potentially influence our interpretation of results and we accounted for this issue. Peterson et al. � Mule Deer Fawn Mortality Varied sample sizes between the areas was accounted for in the variance around estimates (Fig. 4). A model with no difference between the areas ranked higher than a model with a difference, but the high development area with a large sample size could have overwhelmed the low development area with a smaller sample size. Thus, we acknowledge that our results might have been different if we had a balanced study design and larger sample sizes. Contrary to our prediction, rump fat thickness of adult females was positively correlated with predation of fawns. Poor nutritional condition of maternal females contributes to lower birth mass (Robinette et al. 1973), which inhibits fawn growth (Tollefson et al. 2011, Shallow et al. 2015) and increases fawn mortality (Bishop et al. 2009, Hurley et al. 2011, Monteith et al. 2014). However, ad libitum nutrition supplementation only marginally decreases fawn mortality (Bishop et al. 2009). Further, female condition may have limited influence on early fawn survival as some predation of vulnerable fawns (<28 days old) is expected regardless of maternal condition (Hamlin et al. 1984, Ballard et al. 2001) and others suggest the influence of maternal condition on fawn mortality primarily occurs when lactation demands increase (�28 days old; Monteith et al. 2014). Our finding is counterintuitive to previous studies and there is a plausible biological explanation for our unexpected result. Specifically, female condition may be related to past reproductive success (i.e., fawn recruitment) whereby females exhibiting reduced condition may be indicative of continuous reproductive success when compared to females exhibiting improved condition and potentially intermittent reproductive success (Monteith et al. 2013, Bergman et al. 2018). Unfortunately, we were unable to test this hypothesis because we did not capture and monitor the same females each year. However, a longitudinal study reported survival of juvenile elk was higher in females with lower body condition in spring (D. A. Clark, Oregon Department of Fish and Wildlife, unpublished data). A post hoc analysis suggested that elk with continuous reproductive success might explain this relationship compared to females with intermittent reproductive success (D. A. Clark, unpublished data). Past research suggests birth mass contributes to increased mortality in mule deer (Bishop et al. 2009, Monteith et al. 2014, Shallow et al. 2015) because lighter fawns are likely to die from malnutrition and fawns <2.50 kg at birth are nonviable (Lomas and Bender 2007). However, our findings suggest birth mass did not influence fawn mortality (Appendix B) and there are possible reasons for this contradictory result. First, 1143 �mild winters may have contributed to high forage availability for adult females and subsequently increased birth mass. Second, adult female body condition and December fawn mass suggest that most deer were not nutritionally stressed or limited by habitat on winter range in the Piceance Basin during this study (Anderson 2016). Lastly, deer populations in the high and low development area winter ranges have been increasing since 2009 (Anderson 2016) and reproductive parameters (e.g., pregnancy rate and in utero fetal count) were similar between the 2 areas (Peterson et al. 2017). This finding suggests the populations are below carrying capacity. Future studies should explicitly quantify forage availability and quality and hiding cover. The influence of these factors on maternal body condition and subsequently fawn mass should also be quantified to fully comprehend the influence of development on fawn mortality. MANAGEMENT IMPLICATIONS At the intensity and activity we observed, and under the environmental conditions acting during our study, natural gas development did not appear to influence fawn mortality overall. However, there might have been effects to deer when closer to producing well pads. These results indicate that under these conditions, development and mule deer might be compatible, requiring no major management action. Managers should be aware of how habitat characteristics and fragmentation associated with development may influence predation of fawns under different conditions. Our results suggest that maintaining development activity thresholds at the levels we observed will guard against potential negative effects. Our threshold values could be used as potential guidelines until further research clarifies the relationship between weather and higher development thresholds. We caution that higher development intensity and drilling activity in flatter, less rugged areas could influence fawn mortality. Managers should maintain low development densities in areas where topography and vegetation offer less concealment. ACKNOWLEDGMENTS L. L. Wolfe, E. J. Bergman, and C. J. Bishop assisted with ultrasonography and VIT insertion. We thank numerous field technicians, personnel from Colorado Parks and Wildlife Area 6, and volunteers for project coordination and conducting field work. Fixed wing pilots L. L. Gepfert and L. A. Coulter provided assistance with aerial telemetry flights and Quicksilver Air assisted with deer captures. E. J. Bergman, C. J. Bishop, C. N. Jacques, A. W. Maki, P. J. Meiman, D. W. Tripp, G. Wittemyer, and 4 anonymous referees improved the manuscript through constructive reviews. Project funding was provided by Colorado Parks and Wildlife, Exxon Mobil Production/XTO Energy, Williams/WPX Energy, Shell Exploration and Production, EnCana Corporation, Marathon Oil Corporation, Federal Aid in Wildlife Restoration, the Colorado Mule Deer Foundation, the Colorado Mule Deer Association, the Boone and Crockett Club, Colorado Chapter of the Wildlife Society, and the Colorado State Severance Tax. 1144 LITERATURE CITED Anderson, C. R. Jr. 2016. Population performance of Piceance Basin mule deer in response to natural gas resource extraction and mitigation efforts to address human activity and habitat degradation. Federal Aid in Wildlife Restoration Job Progress Report W-185-R, Colorado Parks and Wildlife, Fort Collins, Colorado, USA. Anderson, C. R. Jr., and F. G. Lindzey. 2003. Estimating cougar predation rates from GPS location clusters. Journal of Wildlife Management 67:307–316. Arnold, T. W. 2010. Uninformative parameters and model selection using Akaike’s Information Criterion. Journal of Wildlife Management 74:1175–1178. Ballard, W. B., D. Lutz, T. W. Keegan, L. H. Carpenter, and J. C. deVos. 2001. Deer-predator relationships: a review of recent North American studies with emphasis on mule and black-tailed deer. Wildlife Society Bulletin 29:99–115. Barbieri, M. M., and J. O. Berger. 2004. Optimal predictive model selection. Annals of Statistics 32:870–897. Barbknecht, A. E., W. S. Fairbanks, J. D. Rogerson, E. J. Maichak, B. M. Scurlock, and L. L. Meadows. 2011. Elk parturition site selection at local and landscape scales. Journal of Wildlife Management 75:646–654. Bartmann, R. M. 1983. Composition and quality of mule deer diets on pinyon-juniper winter range, Colorado. Journal of Range Management 36:534–541. Berger, J. 2007. Fear, human shields and the redistribution of prey and predators in protected areas. Biology Letters 3:620–623. Bergman, E. J., C. R. Anderson Jr., C. J. Bishop, A. A. Holland, and J. M. Northrup. 2018. Variation in ungulate body fat: Individual versus temporal effects. Journal of Wildlife Management 82:130–137. Beyer, H. L. 2012. Geospatial modelling environment. Version 0.7.3.0. http://www.spatialecology.com/gme. Accessed 22 Jul 2015. Bishop, C. J., C. R. Anderson Jr., D. P. Walsh, E. J. Bergman, P. Kuechle, and J. Roth. 2011. Effectiveness of a redesigned vaginal implant transmitter in mule deer. Journal of Wildlife Management 75:1797–1806. Bishop, C. J., D. J. Freddy, G. C. White, B. E. Watkins, T. R. Stephenson, and L. L. Wolfe. 2007. Using vaginal implant transmitters to aid in capture of mule deer neonates. Journal of Wildlife Management 71:945–954. Bishop, C. J., G. C. White, D. J. Freddy, B. E. Watkins, and T. R. Stephenson. 2009. Effect of enhanced nutrition on mule deer population rate of change. Wildlife Monographs 172:1–28. Bishop, C. J., G. C. White, and P. M. Lukacs. 2008. Evaluating dependence among mule deer siblings in fetal and neonatal survival analyses. Journal of Wildlife Management 72:1085–1093. Brownie, C., J. E. Hines, J. D. Nichols, K. H. Pollock, and J. B. Hestbeck. 1993. Capture-recapture studies for multiple strata including nonmarkovian transitions. Biometrics 49:1173–1187. Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference a practical information-theoretic approach. Springer, New York, New York, USA. Cheatum, E. L. 1949. Bone marrow as an index of malnutrition in deer. NY State Conservationist 3:19–22. Christie, K. S., W. F. Jensen, J. H. Schmidt, and M. S. Boyce. 2015. Longterm changes in pronghorn abundance index linked to climate and oil development in North Dakota. Biological Conservation 192:445–453. Cook, J. G., B. K. Johnson, R. C. Cook, R. A. Riggs, T. Delcurto, L. D. Bryant, and L. L. Irwin. 2004. Effects of summer-autumn nutrition and parturition date on reproduction and survival of elk. Wildlife Monographs 155:1–61. Cook, R. C., J. G. Cook, D. L. Murray, P. Zager, B. K. Johnson, and M. W. Gratson. 2001. Development of predictive models of nutritional condition for Rocky Mountain elk. Journal of Wildlife Management 65:973–987. Devineau, O., W. L. Kendall, P. F. Doherty, T. M. Shenk, G. C. White, P. M. Lukacs, and K. P. Burnham. 2014. Increased flexibility for modeling telemetry and nest-survival data using the multistate framework. Journal of Wildlife Management 78:224–230. Devineau, O., T. M. Shenk, G. C. White, P. F. Doherty Jr., P. M. Lukacs, and R. H. Kahn. 2010. Evaluating the Canada lynx reintroduction programme in Colorado: patterns in mortality. Journal of Applied Ecology 47:524–531. Doherty, P. F., G. C. White, and K. P. Burnham. 2010. Comparison of model building and selection strategies. Journal of Ornithology 152(Supplement 2):S317–S323. The Journal of Wildlife Management � 82(6) �Dussault, C., V. Pinard, J. P. Ouellet, R. Courtois, and D. Fortin. 2012. Avoidance of roads and selection for recent cutovers by threatened caribou: fitness-rewarding or maladaptive behaviour? Proceedings of the Royal Society B-Biological Sciences 279:4481–4488. Forrester, T. D., and H. U. Wittmer. 2013. A review of the population dynamics of mule deer and black-tailed deer Odocoileus hemionus in North America. Mammal Review 43:292–308. Frid, A., and L. Dill. 2002. Human-caused disturbance stimuli as a form of predation risk. Conservation Ecology 6:art11. Gaillard, J. M., M. Festa-Bianchet, and N. G. Yoccoz. 1998. Population dynamics of large herbivores: variable recruitment with constant adult survival. Trends in Ecology and Evolution 13:58–63. Gaillard, J. M., M. Festa-Bianchet, N. G. Yoccoz, A. Loison, and C. Toigo. 2000. Temporal variation in fitness components and population dynamics of large herbivores. Annual Review of Ecology and Systematics 31:367–393. Garrott, R. A., G. C. White, R. M. Bartmann, L. H. Carpenter, and A. W. Alldredge. 1987. Movements of female mule deer in northwest Colorado. Journal of Wildlife Management 51:634–643. Geist, V. 1981. Behavior: adaptive strategies in mule deer. Pages 157–224. in O. C. Wallmo, editor. Mule and black-tailed deer of North America. University of Nebraska Press, Lincoln, Nebraska, USA. Gilbert, B. A., and K. J. Raedeke. 2004. Recruitment dynamics of blacktailed deer in the western Cascades. Journal of Wildlife Management 68:120–128. Grovenburg, T. W., J. A. Jenks, C. N. Jacques, R. W. Klaver, and C. C. Swanson. 2009. Aggressive defensive behavior by free-ranging whitetailed deer. Journal of Mammalogy 90:1218–1223. Grovenburg, T. W., K. L. Monteith, R. W. Klaver, and J. A. Jenks. 2012. Predator evasion by white-tailed deer fawns. Animal Behaviour 84:59–65. Hamlin, K. L., S. J. Riley, D. Pyrah, A. R. Dood, and R. J. Mackie. 1984. Relationships among mule deer fawn mortality, coyotes, and alternate prey species during summer. Journal of Wildlife Management 48:489–499. Haugen, A. O., and D. W. Speake. 1958. Determining age of young fawn white-tailed deer. Journal of Wildlife Management 22:319–321. Hurley, M. A., J. W. Unsworth, P. Zager, M. Hebblewhite, E. O. Garton, D. M. Montgomery, J. R. Skalski, and C. L. Maycock. 2011. Demographic response of mule deer to experimental reduction of coyotes and mountain lions in southeastern Idaho. Wildlife Monographs 178:1–33. Jacques, C. N., J. A. Jenks, C. S. Deperno, J. D. Sievers, T. W. Grovenburg, T. J. Brinkman, C. C. Swanson, and B. A. Stillings. 2009. Evaluating ungulate mortality associated with helicopter net-gun captures in the Northern Great Plains. Journal of Wildlife Management 73:1282–1291. Johnson, H. E., J. R. Sushinsky, A. Holland, E. J. Bergman, T. Balzer, J. Garner, and S. E. Reed. 2016. Increases in residential and energy development are associated with reductions in recruitment for a large ungulate. Global Change Biology 23:578–591. Johnstone-Yellin, T. L., L. A. Shipley, W. L. Myers, and H. S. Robinson. 2009. To twin or not to twin? Trade-offs in litter size and fawn survival in mule deer. Journal of Mammalogy 90:453–460. Keech, M. A., R. T. Bowyer, J. M. Ver Hoef, R. D. Boertje, B. W. Dale, and T. R. Stephenson. 2000. Life-history consequences of maternal condition in Alaskan moose. Journal of Wildlife Management 64:450–462. Kistner, T. P., C. E. Trainer, and N. A. Hartmann. 1980. A field technique for evaluating physical condition of deer. Wildlife Society Bulletin 8:11–17. Lebreton, J. D., J. D. Nichols, R. J. Barker, R. Pradel, and J. A. Spendelow. 2009. Modeling individual animal histories with multistate capturerecapture models. Advances in Ecological Research 41:87–173. Lebreton, J. D., and R. Pradel. 2002. Multistate recapture models: modelling incomplete individual histories. Journal of Applied Statistics 29:353–369. Lendrum, P. E., C. R. Anderson Jr., R. A. Long, J. G. Kie, and R. T. Bowyer. 2012. Habitat selection by mule deer during migration: effects of landscape structure and natural-gas development. Ecosphere 3:art82. Lendrum, P. E., C. R. Anderson Jr., K. L. Monteith, J. A. Jenks, and R. T. Bowyer. 2013. Migrating mule deer: effects of anthropogenically altered landscapes. PLoS ONE 8:e64548. Lendrum, P. E., C. R. Anderson Jr., K. L. Monteith, J. A. Jenks, and R. T. Bowyer. 2014. Relating the movement of a rapidly migrating ungulate to spatiotemporal patterns of forage quality. Mammalian Biology 79: 369–375. Peterson et al. � Mule Deer Fawn Mortality Lent, P. C. 1974. Mother-infant relationship in ungulates. Pages 14–55 in V. Geist, and F. R. Walther, editors. The behaviour of ungulates and its relation to management. International Union for Conservation of Nature and Natural Resources, Morges, Switzerland. Lingle, S., and S. M. Pellis. 2002. Fight or flight? Antipredator behavior and the escalation of coyote encounters with deer. Oecologia 131:154–164. Lomas, L. A., and L. C. Bender. 2007. Survival and cause-specific mortality of neonatal mule deer fawns, north-central New Mexico. Journal of Wildlife Management 71:884–894. Lynch, E., J. M. Northrup, M. F. McKenna, C. R. Anderson Jr., L. Angeloni, and G. Wittemyer. 2014. Landscape and anthropogenic features influence the use of auditory vigilance by mule deer. Behavioral Ecology 26:75–82. Manly, B. F. J. 2001. Statistics for environmental science and management. Chapman & Hall/CRC, Boca Raton, Florida, USA. Marescot, L., T. D. Forrester, D. S. Casady, and H. U. Wittmer. 2015. Using multistate capture-mark-recapture models to quantify effects of predation on age-specific survival and population growth in black-tailed deer. Population Ecology 57:185–197. Monteith, K. L., V. C. Bleich, T. R. Stephenson, B. M. Pierce, M. M. Conner, J. G. Kie, and R. T. Bowyer. 2014. Life-history characteristics of mule deer: effects of nutrition in a variable environment. Wildlife Monographs 186:1–62. Monteith, K. L., V. C. Bleich, T. R. Stephenson, B. M. Pierce, M. M. Conner, R. W. Klaver, and R. T. Bowyer. 2011. Timing of seasonal migration in mule deer: Effects of climate, plant phenology, and lifehistory characteristics. Ecosphere 2:art47. Monteith, K. L., T. R. Stephenson, V. C. Bleich, M. M. Conner, B. M. Pierce, R. T. Bowyer, and I. Montgomery. 2013. Risk-sensitive allocation in seasonal dynamics of fat and protein reserves in a long-lived mammal. Journal of Animal Ecology 82:377–388. National Climatic Data Center. 2015. NOAA’s National Centers for Environmental Information. http://www.ncdc.noaa.gov/. Accessed 23 Jan 2015. Northrup, J. M., C. R. Anderson Jr., and G. Wittemyer. 2014. Effects of helicopter capture and handling on movement behavior of mule deer. Journal of Wildlife Management 78:731–738. Northrup, J. M., C. R. Anderson Jr., and G. Wittemyer. 2015. Quantifying spatial habitat loss from hydrocarbon development through assessing habitat selection patterns of mule deer. Global Change Biology 21:3961–3970. Northrup, J. M., C. R. Anderson Jr., and G. Wittemyer. 2016. Environmental dynamics and anthropogenic development alter philopatry and space-use in a North American cervid. Diversity and Distributions 22:547–557. Ozoga, J. J., and L. J. Verme. 1986. Relation of maternal age to fawn-rearing success in white-tailed deer. Journal of Wildlife Management 50:480–486. Panzacchi, M., I. Herfindal, J. D. C. Linnell, M. Odden, J. Odden, and R. Andersen. 2010. Trade-offs between maternal foraging and fawn predation risk in an income breeder. Behavioral Ecology and Sociobiology 64:1267–1278. Parker, K. L., P. S. Barboza, and M. P. Gillingham. 2009. Nutrition integrates environmental responses of ungulates. Functional Ecology 23:57–69. Peterson, M. E. 2016. Reproductive success, habitat selection, and neonatal mule deer mortality in a natural gas development area. Dissertation, Colorado State University, Fort Collins, USA. Peterson, M. E., C. R. Anderson Jr., J. M. Northrup, and P. F. Doherty Jr. 2017. Reproductive success of mule deer in a natural gas development area. Wildlife Biology 2017:wlb.00341. Petroelje, T. R., J. L. Belant, D. E. Beyer, G. M. Wang, and B. D. Leopold. 2014. Population-level response of coyotes to a pulsed resource event. Population Ecology 56:349–358. Pojar, T. M., and D. C. Bowden. 2004. Neonatal mule deer fawn survival in west-central Colorado. Journal of Wildlife Management 68:550–560. R Core Team. 2015. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Rearden, S. N., R. G. Anthony, and B. K. Johnson. 2011. Birth-site selection and predation risk of Rocky Mountain elk. Journal of Mammalogy 92:1118–1126. Riley, S. J., and A. R. Dood. 1984. Summer movements, home range, habitat use, and behavior of mule deer fawns. Journal of Wildlife Management 48:1302–1310. 1145 �Robbins, C. T., and B. L. Robbins. 1979. Fetal and neonatal growth patterns and maternal reproductive effort in ungulates and subungulates. American Naturalist 114:101–116. Robinette, W. L., C. H. Baer, R. E. Pillmore, and C. E. Knittle. 1973. Effects of nutritional change on captive mule deer. Journal of Wildlife Management 37:312–326. Robinette, W. L., N. V. Hancock, and D. A. Jones. 1977. The Oak Creek mule deer herd in Utah. Utah Division of Wildlife Resources Publication 77-15, Salt Lake City, USA. Robinette, W. L., D. A. Jones, G. Rogers, and J. S. Gashwiler. 1957. Notes on tooth development and wear for Rocky Mountain mule deer. Journal of Wildlife Management 21:134–153. Rogers, L. L. 1970. Black bear of Minnesota. Naturalist 21:42–47. Rost, G. R., and J. A. Bailey. 1979. Distribution of mule deer and elk in relation to roads. Journal of Wildlife Management 43:634–641. Sams, M. G., R. L. Lochmiller, E. C. Hellgren, W. D. Warde, and L. W. Varner. 1996. Morphometric predictors of neonatal age for white tailed deer. Wildlife Society Bulletin 24:53–57. Sawyer, H., M. J. Kauffman, A. D. Middleton, T. A. Morrison, R. M. Nielson, T. B. Wyckoff, and N. Pettorelli. 2013. A framework for understanding semi-permeable barrier effects on migratory ungulates. Journal of Applied Ecology 50:68–78. Sawyer, H., M. J. Kauffman, and R. M. Nielson. 2009. Influence of well pad activity on winter habitat selection patterns of mule deer. Journal of Wildlife Management 73:1052–1061. Sawyer, H., N. M. Korfanta, R. M. Nielson, K. L. Monteith, and D. Strickland. 2017. Mule deer and energy development—long-term trends of habituation and abundance. Global Change Biology 23:4521–4529. Sawyer, H., R. M. Nielson, F. Lindzey, and L. L. McDonald. 2006. Winter habitat selection of mule deer before and during development of a natural gas field. Journal of Wildlife Management 70:396–403. Shallow, J. R. T., M. A. Hurley, K. L. Monteith, and R. T. Bowyer. 2015. Cascading effects of habitat on maternal condition and life-history characteristics of neonatal mule deer. Journal of Mammalogy 96:194–205. Sikes, R. S., and The Animal Care and Use Committee of the American Society of Mammalogists. 2016. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. Journal of Mammalogy 97:663–688. Steigers, W. D., and J. T. Flinders. 1980. Mortality and movements of mule deer fawns in Washington. Journal of Wildlife Management 44:381–388. Stephenson, T. R., V. C. Bleich, B. M. Pierce, and G. P. Mulcahy. 2002. Validation of mule deer body composition using in vivo and post-mortem indices of nutritional condition. Wildlife Society Bulletin 30:557–564. Stonehouse, K. F., C. R. Anderson Jr., M. E. Peterson, and D. R. Collins. 2016. Approaches to field investigations of cause-specific mortality in mule deer (Odocoileus hemionus). Colorado Parks and Wildlife Technical Report No. 48, Fort Collins, USA. 1146 Testa, J. W. 2002. Does predation on neonates inherently select for earlier births? Journal of Mammalogy 83:699–706. Tollefson, T. N., L. A. Shipley, W. L. Myers, and N. Dasgupta. 2011. Forage quality’s influence on mule deer fawns. Journal of Wildlife Management 75:919–928. United States Energy Information Administration. 2017. International energy outlook 2017. United States Department of Energy, Washington, D.C., USA. Wade, D. A., and J. E. Bowns. 1982. Procedures for evaluating predation on livestock and wildlife. Texas Agricultural Extension Service, Texas A&M University, San Angelo, USA. Walker, B. L., D. E. Naugle, and K. E. Doherty. 2007. Greater sage-grouse population response to energy development and habitat loss. Journal of Wildlife Management 71:2644–2654. Walther, F. R. 1965. Verhaltensstudien an der grantsgazelle (Gazella granti Brooke, 1872) im Ngorongoro-Krater. Zeitschrift fu ̈r Tierpsychologie 22:167–208. Webb, S. L., M. R. Dzialak, S. M. Harju, L. D. Hayden-Wing, and J. B. Winstead. 2011a. Effects of human activity on space use and movement patterns of female elk. Wildlife Society Bulletin 35:261–269. Webb, S. L., M. R. Dzialak, R. G. Osborn, S. M. Harju, J. Wondzell, L. Hayden-Wing, and J. B. Winstead. 2011b. Using pellet groups to assess response of elk and deer to roads and energy development. Wildlife Biology in Practice 7:32–40. Webb, S. L., J. S. Lewis, D. G. Hewitt, M. W. Hellickson, and F. C. Bryant. 2008. Assessing the helicopter and net gun as a capture technique for white-tailed deer. Journal of Wildlife Management 72:310–314. White, G. C., and K. P. Burnham. 1999. Program MARK: survival estimation from populations of marked animals. Bird Study 46: 120–139. White, G. C., W. L. Kendall, and R. J. Barker. 2006. Multistate survival models and their extensions in Program MARK. Journal of Wildlife Management 70:1521–1529. White, M. 1973. Description of remains of deer fawns killed by coyotes. Journal of Mammalogy 54:291–293. Whittaker, D. G., and F. G. Lindzey. 1999. Effect of coyote predation on early fawn survival in sympatric deer species. Wildlife Society Bulletin 27:256–262. Associate Editor: Christopher Jacques. SUPPORTING INFORMATION Additional supporting information may be found in the online version of this article at the publisher’s website. The Journal of Wildlife Management � 82(6) �APPENDIX A. DESCRIPTIVE STATISTICS FOR MULE DEER FAWN MORTALITY. Table A1. Descriptive statistics for all continuous covariates included in multi-state analyses of mule� deer fawn mortality in the Piceance Basin, Colorado, � USA, 2012–2014. Parameters were probability of transitioning from an alive state in the high cHH or low cLL development study areas to a death by � � � predation c:K , malnutrition c:M , or unknown mortality c:U state. Covariatea �x SD Min. Max. Winter distance to well pad Winter distance to road Summer distance to well pad Summer distance to road Female age Rump fat Birth date Mass Previous precipitation Precipitation Temperature Winter distance to well pad Winter distance to road Summer distance to well pad Summer distance to road Female age Rump fat Birth date Mass Previous precipitation Precipitation Temperature Winter distance to well pad Winter distance to road Summer distance to well pad Summer distance to road Female age Rump fat Birth date Mass Previous precipitation Precipitation Temperature Winter distance to well pad Winter distance to road Summer distance to well pad Summer distance to road Female age Rump fat Birth date Mass Previous precipitation Precipitation Temperature 0.35 0.24 0.39 0.26 4.94 2.46 11.21 3.29 20.73 0.39 23.71 0.40 0.41 0.40 0.49 4.75 2.25 11.88 2.88 20.73 0.39 23.71 0.38 0.23 0.40 0.16 5.50 2.06 13.44 3.19 20.73 0.39 23.71 0.37 0.22 0.40 0.24 4.76 1.85 11.34 3.42 20.73 0.39 23.71 0.11 0.22 0.05 0.35 2.29 1.92 5.14 0.65 6.31 0.86 8.04 0.00 0.49 0.00 0.67 2.31 0.71 8.81 0.91 6.31 0.86 8.04 0.05 0.22 0.00 0.13 1.78 1.16 6.77 0.87 6.31 0.86 8.04 0.08 0.17 0.00 0.36 1.85 1.02 6.45 0.68 6.31 0.86 8.04 0.00 0.00 0.10 0.00 1.50 0.00 1.00 1.80 12.40 0.00 �10.24 0.40 0.10 0.40 0.10 1.50 1.00 2.00 1.60 12.40 0.00 �10.24 0.20 0.00 0.40 0.00 2.50 1.00 2.00 1.40 12.40 0.00 �10.24 0.00 0.00 0.40 0.00 1.50 0.00 0.00 1.10 12.40 0.00 �10.24 0.40 1.60 0.40 1.90 10.50 12.00 24.00 5.20 27.60 5.10 35.00 0.40 1.60 0.40 2.10 9.50 3.00 25.00 3.90 27.60 5.10 35.00 0.40 0.90 0.40 0.40 8.50 5.00 30.00 4.70 27.60 5.10 35.00 0.40 0.70 0.40 2.10 8.50 4.00 32.00 5.30 27.60 5.10 35.00 Parameter � Predation c:K Malnutrition c:M � Unknown mortality c:U Alive cHH ; cLL a � � Covariates include distance (km) from nearest producing well pad on winter range with data truncated beyond 0.40 km (winter distance to well pad), distance (km) from nearest road on winter range (winter distance to road), distance (km) from nearest producing well pad on summer range with data truncated beyond 0.40 km (summer distance to well pad), distance (km) from nearest road on summer range (summer distance to road), age of adult females estimated in December during adult female capture (female age), rump fat thickness (mm) of adult females measured in March (rump fat), Julian date of birth represented as the number of days following the earliest detected birth in each year (birth date), estimated birth mass (kg) of fawns (mass), total precipitation (cm) during the previous winter season before parturition (1 Oct–30 Apr; previous precipitation), daily precipitation (cm) during the parturition period and 7-day average of precipitation after the parturition period until 15 December (precipitation), and daily temperature (8C) during the parturition period and 7day average of temperature after the parturition period until 15 December (temperature). Peterson et al. � Mule Deer Fawn Mortality 1147 �APPENDIX B. FIRST STAGE OF MULE DEER FAWN MORTALITY MODELING. Table B1. Cumulative weights for Akaike’s Information Criterion adjusted for small sample size (AICc), for all covariates included in the first stage of analysis of mule deer fawn mortality in the Piceance Basin, Colorado, USA, 2012–2014. � � Parameters were probability � of transitioning from an alive state in the high or low development study areas to a death by predation c:K , malnutrition c:M , or unknown mortality c:U state. We considered cumulative weights >0.50 to be important. Covariatea Parameter Winter range development characteristics � Predation c:K � Malnutrition c:M Unknown mortality c:U � Summer range development characteristics � Predation c:K � Malnutrition c:M � Unknown mortality c:U Adult female characteristics � Predation c:K Malnutrition c:M � Unknown mortality c:U � Fawn birth characteristics � Predation c:K Malnutrition c:M � Unknown mortality c:U � Temporal characteristics � Predation c:K Malnutrition c:M � Unknown mortality c:U a � Winter Winter Winter Winter Winter Winter distance distance distance distance distance distance to to to to to to well pad road road well pad well pad road Cumulative AICc weight 0.92 0.50 0.73 0.48 0.33 0.27 Summer distance to well pad Summer distance to well pad Summer distance to well pad 0.95 0.30 0.35 Rump fat Female age Fetal count Fetal count Rump fat Female age Female age Rump fat Fetal count 0.98 0.22 0.21 0.35 0.25 0.21 0.37 0.21 0.21 Fawn age 0–14 days old Mass Birth date Sex Fawn age 0–14 days old Mass Sex Birth date Fawn age 0–14 days old Birth date Mass Sex 0.61 0.13 0.05 0.04 0.98 0.26 0.09 0.03 0.31 0.21 0.10 0.10 Temperature Previous precipitation Precipitation Temperature Previous precipitation Precipitation Temperature Precipitation Previous precipitation 0.62 0.49 0.31 0.99 0.21 0.18 0.26 0.17 0.15 Covariates include distance (km) from nearest producing well pad on winter range with data truncated beyond 0.40 km (winter distance to well pad), distance (km) from nearest road on winter range (winter distance to road), distance (km) from nearest producing well pad on summer range with data truncated beyond 0.40 km (summer distance to well pad), rump fat thickness (mm) of adult females measured in March (rump fat), age of adult females estimated in December during adult female capture (female age), in utero fetal count documented in March during capture (fetal count), fawn age trend from 0–14 days old and constant thereafter (fawn age 0–14 days old), estimated birth mass (kg) of fawns (mass), Julian date of birth represented as the number of days following the earliest detected birth in each year (birth date), sex of captured fawns (sex), daily temperature (8C) during the parturition period and 7-day average of temperature after the parturition period until 15 December (temperature), total precipitation (cm) during the previous winter season before parturition (1 Oct–30 Apr; previous precipitation), and daily precipitation (cm) during the parturition period and 7-day average of precipitation after the parturition period until 15 December (precipitation). 1148 The Journal of Wildlife Management � 82(6) � https://cpw.cvlcollections.org/files/original/9c186dfc0667ec35193f12c7006cb54a.pdf 9007c4ecb35d9312d25f5589c537bd0a Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Journal Articles Description An account of the resource CPW peer-reviewed journal publications Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Mortality of mule deer fawns in a natural gas development area Rights Information about rights held in and over the resource <a href="http://rightsstatements.org/vocab/InC-NC/1.0/" target="_blank" rel="noreferrer noopener">In Copyright - Non-Commercial Use Permitted</a> Date Created Date of creation of the resource. 2018-05-01 Subject The topic of the resource Anthropogenic disturbances Colorado Mule deer Natural gas development <em>Odocoileus hemionus</em> Population dynamics Predation Survival Description An account of the resource Recent natural gas development has caused concern among wildlife managers, researchers, and stakeholders over the potential effects on wildlife and their habitats. Specifically, understanding how this development and other factors influence mule deer (<em>Odocoileus hemionus</em>) fawn (i.e., 0–6 months old) mortality rates, recruitment, and subsequently population dynamics have been identified as knowledge gaps. Thus, we tested predictions concerning the relationship between natural gas development, adult female, fawn birth, and temporal (weather) characteristics on fawn mortality in the Piceance Basin of northwestern Colorado, USA, from 2012–2014. We captured and radio-collared 184 fawns and estimated apparent cause-specific mortality in areas with relatively high or low levels of natural gas development using a multi-state model. Mean daily predation probability was similar in the high versus low development areas. Predation was the leading cause of fawn mortality in both areas and decreased from 0–14 days old. Black bear (<em>Ursus americanus</em>; 22% of all mortalities, n = 17) and cougar (<em>Felis concolor</em>; 36% of all mortalities, n = 6) predation was the leading cause of mortality in the high and low development areas, respectively. Predation of fawns was negatively correlated with the distance from a female's core area to a producing well pad on winter or summer range. Contrary to expectations, predation of fawns was positively correlated with rump fat thickness of adult females. Well pad densities and development activity were relatively low during our study, indicating that the observed intensity of development did not appear to influence daily predation probability. Our results suggest maintaining development activity thresholds at levels we observed to potentially minimize the effects of development on fawn mortality. However, we caution that higher development intensity and drilling activity in flatter, less rugged areas with less concealment cover could influence fawn mortality. Managers should maintain low development densities in areas where topography and vegetation offer less concealment. Overall, region-specific data (e.g., development intensity, topography, predator assemblages, and associated predation risk) are needed to better understand the effects of natural gas development on fawn mortality. Creator An entity primarily responsible for making the resource Peterson, Mark E. Anderson Jr, Charles R. Northrup, Joseph M. Doherty Jr, Paul F. Language A language of the resource English Is Part Of A related resource in which the described resource is physically or logically included. The Journal of Wildlife Management Extent The size or duration of the resource. 14 pages Bibliographic Citation A bibliographic reference for the resource. Recommended practice is to include sufficient bibliographic detail to identify the resource as unambiguously as possible. Peterson, M.E., C. R. Anderson Jr, J. M. Northrup, and P. F. Doherty Jr. 2018. Mortality of mule deer fawns in a natural gas development area. The Journal of Wildlife Management 82:1135–1148. <a href="https://doi.org/10.1002/jwmg.21476" target="_blank" rel="noreferrer noopener">https://doi.org/10.1002/jwmg.21476</a> Format The file format, physical medium, or dimensions of the resource application/pdf Type The nature or genre of the resource Article