https://cpw.cvlcollections.org/files/original/fce423b2d21f6469652fbd58896fc010.pdf 16acbca2c2b672982eac1838dc30fecd 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 �Wildlife Society Bulletin 43(4):717–725; 2019; DOI: 10.1002/wsb.1034 Tools and Technology A Noninvasive Automated Device for Remotely Collaring and Weighing Mule Deer CHAD J. BISHOP,1,2 Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA MATHEW W. ALLDREDGE, Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA DANIEL P. WALSH,3 Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA ERIC J. BERGMAN, Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA CHARLES R. ANDERSON, JR., Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA DARLENE KILPATRICK, Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA JOE BAKEL, Dynamic Group Circuit Design, Inc., 2629 Redwing Road, Suite 360, Fort Collins, CO 80525, USA CHRISTOPHE FEBVRE, Dynamic Group Circuit Design, Inc., 2629 Redwing Road, Suite 360, Fort Collins, CO 80525, USA ABSTRACT Wildlife biologists capture deer (Odocoileus spp.) annually to attach transmitters and collect basic information (e.g., animal mass and sex) as part of ongoing research and monitoring activities. Traditional capture techniques induce stress in animals and can be expensive, ineﬃcient, and dangerous. They are also impractical for some urbanized settings. We designed and evaluated a device for mule deer (O. hemionus) that automatically attached an expandable radiocollar to a ≥6‐month‐old fawn and recorded the fawn’s mass and sex, without physically restraining the animal. The device did not require on‐site human presence to operate. Students and faculty in the Mechanical Engineering Department at Colorado State University produced a conceptual model and early prototype. Professional engineers at Dynamic Group Circuit Design, Inc. in Fort Collins, Colorado, USA, produced a fully functional prototype of the device. Using the device, we remotely collared, weighed, and identiﬁed sex of 8 free‐ranging mule deer fawns during winters 2010–2011 and 2011–2012. Collars were modiﬁed to shed from deer approximately 1 month after the collaring event. Two fawns were successfully recollared after they shed the ﬁrst collars they received. Thus, we observed 10 successful collaring events involving 8 unique fawns. Fawns demonstrated minimal response to collaring events, either remaining in the device or calmly exiting. A fawn typically required ≥1 weeks of daily exposure before fully entering the device and extending its head through the outstretched collar, which was necessary for a collaring event to occur. This slow acclimation period limited utility of the device when compared with traditional capture techniques. Future work should focus on device modiﬁcations and altered baiting strategies that decrease fawn acclimation period, and in turn, increase collaring rates, providing a noninvasive and perhaps cost‐eﬀective alternative for monitoring mid‐ to large‐sized mammal species. © 2019 The Wildlife Society. KEY WORDS automated, baiting, capture, capture techniques, collaring, fawn, handling, mule deer, noninvasive, Odocoileus hemionus. State wildlife agencies in the western United States routinely capture and radiomark 6‐month‐old mule deer (Odocoileus hemionus) fawns to monitor status of populations and accomplish various research objectives. For example, Received: 12 November 2018; Accepted: 6 July 2019 Published: 28 November 2019 1 Email: chad.bishop@umontana.edu Current aﬃliation: Wildlife Biology Program, Department of Ecosystem and Conservation Sciences, University of Montana, 32 Campus Drive, Missoula, MT 59812, USA. 3 Current aﬃliation: National Wildlife Health Center, United States Geological Survey, 6060 Schroeder Road, Madison, WI 53711‐6223, USA. 2 Bishop et al. • Automated Collaring Device for Deer Colorado Parks and Wildlife, USA, (CPW) has captured 250−650 fawns annually during the past 20 years to monitor population survival rates and accomplish objectives of research projects. Most capture is accomplished with net‐ guns ﬁred from helicopters because it is typically eﬃcient (3–6 deer captured/hr) and allows speciﬁc age and sex classes of animals to be targeted for capture (e.g., fawns; Barrett et al. 1982, van Reenen 1982, Webb et al. 2008). However, helicopter net‐gunning is becoming increasingly expensive (i.e., >US$500/captured deer), requiring agency personnel to increase capture budgets or reduce sample sizes. Budget increases are often not feasible and sample size reductions inﬂate variance estimates and may compromise the study or project altogether. Also, helicopter‐based 717 �capture is not practical in residential areas, which is becoming a signiﬁcant limitation in some regions of the western United States as residential development expands into traditionally undeveloped areas. Failure to capture deer in these areas leads to biased data collection. Helicopter net‐ gunning is also inherently dangerous given the need to ﬂy aggressively at low altitude. Dozens of wildlife biologists and pilots have been killed in aviation accidents, which represent the leading job‐related cause of death for wildlife professionals (Sasse 2003). The authors of this manuscript alone have lost 9 colleagues in helicopter and ﬁxed‐wing aircraft crashes. Other colleagues have survived crashes but experienced life‐altering injuries. Presently, most agencies rely on contracted crews for net‐gun wildlife capture rather than accomplishing the work in‐house. Small markets and inherent risk limit availability of contractors, further constraining use of the technique. Helicopter net‐gunning remains the preferred capture technique for ungulates, but there is growing need for a safer, more cost‐eﬀective alternative capture technique, especially in areas where helicopter‐based capture is impractical. Helicopter drive nets (Beasom et al. 1980), baited drop nets (Ramsey 1968, Schmidt et al. 1978), baited clover traps (Clover 1954, 1956; VerCauteren et al. 1999), and darting (Wolfe et al. 2004) are used in the western United States to capture deer on occasion in lieu of net‐gunning, but these techniques can be time‐consuming and labor‐ intensive. Drive nets require use of a helicopter and ﬁeld crew and generally do not represent a cost savings over helicopter net‐gunning. Drop nets and clover traps rely on bait to attract animals, and therefore, are typically ineﬃcient and do not allow speciﬁc age classes of animals to be targeted for capture. These bait‐reliant techniques are weather‐dependent and considerably less eﬀective during mild conditions when animals are not as hard‐pressed to secure food. For comparison, one day of helicopter capture is roughly equivalent to 1–3 weeks of drop‐netting or clover‐trapping in terms of target animals captured (White and Bartmann 1994, DelGiudice et al. 2001, Webb et al. 2008). In our experience, ground‐based darting is also labor‐intensive and typically ineﬃcient when compared with helicopter capture. Many biologists lack time and resources given other job requirements to conduct ground‐based capture operations eﬀectively. In fact, when labor costs are considered, helicopter net‐ gunning has been found to be the most cost‐eﬀective capture option (White and Bartmann 1994). Wildlife professionals have need for an alternate capture technique that does not require use of helicopters or weeks of continuous ﬁeld work involving capture crews. Animal welfare is an increasingly important consideration relative to wildlife capture and research. Most wildlife studies involving animal observation, handling, or manipulation are now formally reviewed by an Institutional Animal Care and Use Committee (IACUC). Further, it is now commonplace for scientiﬁc journals, including the Wildlife Society Bulletin, to require documentation that vertebrate animals used in research were 718 treated ethically and humanely (Cox et al. 2018). The traditional capture techniques referenced above have been deemed acceptable by various IACUCs when implemented with procedural safeguards that minimize stress placed on animals (e.g., limiting helicopter pursuit times). Even so, it is not uncommon for a small percentage of animals (e.g., <5%) to die from capture‐related injuries or stresses when using these techniques (White and Bartmann 1994; DelGiudice et al. 2001, 2005; Webb et al. 2008; Jacques et al. 2009). An additional small percentage of animals experience injuries that do not result in immediate mortality but may negatively aﬀect their longer term ﬁtness (Webb et al. 2008). Traditional capture techniques have also been shown to alter animal movement and behavior during the initial days postcapture (Morellet et al. 2009, Northrup et al. 2014). An ideal capture technique would be cost‐eﬀective, eﬃcient, safe for humans, and minimally invasive for animals. This may sound unrealistic, but wildlife researchers have in eﬀect already accomplished this ideal through the development of alternative methods for ‘capturing’ and monitoring wildlife. Examples include the use of DNA (Luikart et al. 2010) and remote‐sensed cameras (O’Connell et al. 2011, Burton et al. 2015) to uniquely identify individuals and monitor them without capture or handling. However, there remains a clear need to attach transmitters (i.e., very‐high‐ frequency [VHF], Global Positioning Systems [GPS], satellite) to wildlife, such as mule deer in our case, to address research questions pertaining to ﬁne‐scale movements, survival, and cause‐speciﬁc mortality, as examples. Our study objective was to devise a technique for attaching a collar‐ type transmitter to a deer without physically restraining or handling the animal, with the ultimate goal of reducing costs, human risk, and animal stress associated with traditional collaring techniques. We conceived the idea of an automated collaring device that would attach a radiocollar to a deer of speciﬁed age class (i.e., fawn vs. adult) and record animal mass and sex without physical restraint or handling. Animal mass and sex are routinely used as covariates in survival studies and monitoring programs (Unsworth et al. 1999, Lukacs et al. 2009). Such a device could prove eﬀective as a supplemental technique to help achieve target sample sizes of collared deer and reduce capture‐ induced stress and mortality. We developed a conceptual model and collaborated with student and professional engineers to design, produce, and evaluate an automated device for collaring, weighing, and identifying sex of speciﬁed mule deer under free‐ ranging conditions. The collaring device is bait‐reliant and similar in size and shape to a clover trap. Our focus was on collaring ≥6 month old fawns to meet CPW deer management needs, although we designed the device such that any age or sex class (i.e., fawns, adult females, adult males [post–antler‐drop]) could be targeted for collaring. STUDY AREA We worked with captive deer at CPW’s Foothills Wildlife Research Facility (FWRF) in Fort Collins, Colorado, during 2008–2009 when designing the automated collaring Wildlife Society Bulletin • 43(4) �device and evaluating initial prototypes. During April–May 2009 and October 2010 to March 2011, we conducted ﬁeld evaluations of the collaring device with wild, free‐ranging deer at 5 sites along Colorado’s northern Front Range: 1) Horsetooth Reservoir, west of Fort Collins (1,690 m, private land); 2) Masonville, southwest of Fort Collins (1,645 m, private land); 3) Red Feather Lakes, northwest of Fort Collins (2,450 m, private land); 4) Hall Ranch, west of Lyons (1,890 m, Boulder County Parks and Open Space); and 5) Heil Valley Ranch, southwest of Lyons (1,950 m, Boulder County Parks and Open Space). During 2012, we conducted ﬁeld evaluations with free‐ranging deer at the Hall Ranch during January and in the Piceance Basin southwest of Meeker, Colorado, during February and March (1,950 m, public land). Vegetation varied among sites and consisted of grassland, mountain shrub, pinyon pine (Pinus edulis) and Utah juniper (Juniperus osteosperma), or conifer. Average minimum temperature during January was −9.0° C at sites along the northern Front Range and −14.5° C in Piceance Basin (Western Regional Climate Center [WRCC] 2016). The extreme minimum temperature during winter among sites when the collaring device was present was −29.0° C. Snow was often present from December through March at depths ≤10 cm. Snow depth occasionally exceeded 10 cm immediately following a snow storm. Mule deer were the most prevalent large mammals in the study area, with both migratory and nonmigratory deer present. Other mid‐ to large‐sized mammals among study areas included elk (Cervus canadensis), coyotes (Canis latrans), cougars (Puma concolor), bobcats (Lynx rufus), and black bears (Ursus americanus). METHODS We identiﬁed a set of speciﬁcations to inform the design of a collaring device prototype. According to our speciﬁcations, the operational prototype device would • remotely attach a radiocollar around the neck of a deer; • only attach a radiocollar to a deer in a speciﬁed mass range (i.e., to distinguish fawns from adults); • not attach a radiocollar to a nontarget animal; • prevent an individual deer from being collared more than once; • measure and record deer mass and sex when collar is deployed; • deter adult deer or other larger animals from entering and using the device (in order to target fawn deer); • have multiple exits to allow animals to easily exit the device at any time; • not cause injury to animals; • include a place for bait for luring deer to the device; • obtain photographs of collared deer; • ﬁt in the back of a typical full‐size pickup truck; and • be programmable and adjustable to allow other ages and species of ungulates to be collared with basic modiﬁcation. These speciﬁcations allowed us to identify a conceptual model of the collaring device. Bishop et al. • Automated Collaring Device for Deer We sought assistance from Colorado State University’s (CSU) Mechanical Engineering Department to reﬁne our conceptual model and develop a prototype device. Six CSU engineering students and a faculty mentor worked on the device during the 2008–2009 academic year as part of a senior design project. We met with the students weekly and provided them a materials budget of US$10,000 to produce the prototype. We conducted staged evaluations of device components throughout the year by working with captive deer at FWRF. The CSU student project team produced a prototype device that met our mechanical design speciﬁcations but did not meet many of the electrical design speciﬁcations, which were beyond the scope of a senior design project. Although the device was not fully functional, we conducted evaluations of the prototype with free‐ranging deer during spring 2009 at the Horsetooth Reservoir site. Field evaluations focused on how deer interacted with the device, which we used to inform reﬁnements to the design. We documented deer interactions with the device using direct observation and motion‐sensor digital cameras (Reconyx, Holmen, WI, USA). During this initial evaluation period, automated electrical components of the collaring device were disabled when we were not present at the site to prevent any potential harm to deer. Following preliminary ﬁeld evaluations, we reﬁned our design speciﬁcations (Appendix A; available online in Supporting Information) and contracted with Dynamic Group Circuit Design (DGCD; Fort Collins, CO, USA) to produce a fully functional prototype device. We met regularly with electrical engineers from DGCD and a subcontracted mechanical engineer during 2009–2010. These meetings ensured that our device speciﬁcations were satisfactorily met from both engineering and deer biology perspectives. Working with DGCD, we produced a fully functional prototype device in 2010 that met our various design speciﬁcations. We also designed an expandable radiocollar for use with the collaring device that accommodated neck circumferences from 28 to 41 cm and weighed <300 grams. The expandable collar comprised springs embedded in a stitched nylon casing that was attached to a VHF transmitter (Advanced Telemetry Systems, Isanti, MN, USA; Lotek Wireless, Newmarket, ON, Canada). We inserted surgical tubing between the VHF transmitter and one end of the nylon casing to make collar attachment temporary. We then partially cut the surgical tubing with a knife so collars would shed from deer approximately 4 weeks postcollaring. The fully functional prototype collaring device comprised an aluminum, rectangular cage attached to a bait compartment (Figs. 1 and 2). Dimensions of the collaring device were 236 cm (l) × 91 cm (w) × 114 cm (h), excluding wheels. The rectangular cage and bait compartment were detachable to facilitate transport. The collaring device included axles and wheels so one person could move the device. Deer entered the device through an adjustable opening at the front of the cage. The purpose of the adjustable opening was to deter entry of larger animals by adjusting both width and height. The sides of the cage comprised one‐way gates that prevented entry into the device, but allowed an animal to exit the device 719 �Figure 1. Design schematic of fully functional prototype device for remotely collaring and weighing mule deer using automation. at any time. The bait compartment was accessed by an animal through an opening positioned at the rear of the cage. We placed an expandable radiocollar in this opening by extending it around 4 rectangular, aluminum plates that held the collar in the fully expanded position (Fig. 3). Clear plexiglass separated the cage from the bait compartment to maximize visibility. To access the bait, a deer extended its head and neck through the expanded radiocollar positioned in the Figure 2. Fully functional prototype device for remotely collaring and weighing mule deer using automation, near Horsetooth Reservoir, Colorado, USA, during February 2011. 720 opening between the cage and bait compartment. The ﬂoor of the cage was a scale that continuously recorded mass and informed device operation. The scale was programmed to tare every 15 minutes (when mass was static) so that snow, bait, or debris on the scale did not aﬀect animal mass Figure 3. View of an expanded very‐high‐frequency (VHF) radiocollar within an automated collaring device for mule deer. The purpose of the device is to remotely place a radiocollar around the neck of a deer after it extends its head through the expanded collaring opening and only when all other requirements are satisﬁed (e.g., deer is in the correct mass range and is not already collared). When the device is triggered, the top 2 metal plates holding the collar in an expanded position collapse, causing the collar to contract around the deer’s neck. Wildlife Society Bulletin • 43(4) �readings. The scale sampled 4 mass readings/second and kept a running average of 8 samples to determine mass at any moment. Only animals in a speciﬁed mass range could be collared, which allowed us to target fawns and avoid collaring adult deer. When an animal entered the device that was heavier or lighter than the programmed mass range, the collar release mechanism was temporarily disabled. Additionally, an actuator moved a plexiglass plate into the opening between the cage and bait pan, preventing nontarget animals from accessing the bait. Shortly after a nontarget animal exited the device, the collar release mechanism was reactivated (i.e., set to release when triggered) and the actuator lowered the plexiglass plate so that the bait was accessible. To prevent an animal from being collared twice, we placed a loop antenna around the entrance to the cage and connected it to a radio frequency identiﬁcation (RFID) reader. All collars used with the device contained a small RFID transponder sewn into the nylon collar material. If a previously collared fawn entered the cage, the RFID transponder was detected, which in turn prevented the collar from being released and activated the actuator to block access to the bait. If a deer entered the cage that was in the speciﬁed mass range and had not been previously collared, the collar would release around the deer’s neck once it accessed the bait. The collar release was triggered when a deer’s head crossed an infrared (IR) beam positioned directly above the bait pan. Speciﬁcally, the broken IR beam activated a solenoid, which in turn released a lever that caused the upper 2 aluminum plates holding the expanded collar in place to collapse (Fig. 3). The collar then contracted around the deer’s neck. When the collar was released, 2 diﬀerent cameras were immediately activated to take a series of 3 photographs each. We positioned one camera in the back of the bait compartment and set it to take a close‐ up photo of the top of the deer’s head. We positioned the second camera in the ﬂoor of the cage and set it to take a photo of the deer’s abdomen and groin. These cameras were activated only when a collar was released and facilitated determination of deer sex. At the moment of collar release, the device also recorded and stored the mass of the deer based on the most recent running average of 8 mass readings collected during the 2 seconds prior to collar release. We manipulated program settings and device operation and uploaded mass data by plugging in an external computer loaded with appropriate software to the device. The device was powered by a 12‐V absorbent glass mat lead acid battery that had to be recharged every 2–3 days during continuous operation. The DGCD prepared a user’s manual that explained device operation and contained detailed schematics to allow future production. Readers interested in reviewing the detailed schematics or obtaining a device should contact the senior author or DGCD. We evaluated eﬀectiveness of the device in the ﬁeld during October–March 2010–2011 and January–March 2012. To attract deer, we baited evaluation sites with alfalfa hay, apple pulp, dried fruit, and cereal. We baited sites for several days, or sometimes weeks, prior to placing the collaring device at the site. Once the collaring device was present, we gradually reduced bait external to the device so that most bait was Bishop et al. • Automated Collaring Device for Deer contained within the collaring device. We always maintained some bait external to the device to encourage deer groups to keep attending the site. Exact baiting strategies were varied throughout the study as part of learning how to most eﬀectively attract deer to a site and encourage fawns to access the collaring device. During most of the study, we only set the device in an operational mode with a collar in place when we were on site and able to directly observe deer–device interactions. After collaring several animals in this manner and troubleshooting problems with the device, we set the device to operate remotely without an observer on site, which is how it was intended to be used. All evaluations of the collaring device with both captive and free‐ranging deer were approved by Colorado Parks and Wildlife’s Institutional Animal Care and Use Committee (Project 04‐2009). RESULTS Initial Field Evaluation We initiated baiting at Horsetooth Reservoir and Masonville on 21 October 2010, and deer promptly began accessing the bait. On 26 October, we placed the collaring device at Horsetooth and began encouraging deer to walk into the device by placing bait on the scale inside the cage. On 29 October, we observed a deer access the bait pan within the bait compartment (i.e., with head extended through collar opening). However, the collaring device malfunctioned and the deer was not collared. We observed additional deer enter the device and access the bait pan during the weeks that followed, but none were collared because the device was not working properly. Malfunctions were caused by electrical signal interference between the RFID reader and a camera and by inadequate structural support underneath the scale, which caused mass measurements to be inaccurate. Once diagnosed, the problems were promptly corrected and we remotely collared a female fawn on 17 November 2010 (Fig. 4). The fawn showed little reaction to the collaring event, calmly exiting the device shortly after receiving the collar. We successfully recorded the fawn’s mass and positively identiﬁed sex from a photograph of the fawn’s head. We continued to monitor the device at Horsetooth Reservoir because there were adequate numbers of uncollared fawns present. However, we continued to encounter various problems with the device that aﬀected functionality. Most notably, the collar release mechanism repeatedly failed to release the collar when a fawn was in the correct position. The source of the problem was a mechanical failing associated with the release mechanism itself, which we modiﬁed accordingly. We also noted that small fawns accessing the bait sometimes failed to break the infrared beam extending across the center of the bait pan, which we corrected by adjusting the position of the bait pan. We then successfully collared 2 more fawns (1 male and 1 female) on successive days, 13 and 14 December 2010. Also, the female fawn that was collared on 17 November shed its collar on 13 December and was successfully recollared on 20 December. On 21 December 2010, the actuator that opened and closed the bait door short‐circuited in response to cold, snowy weather and damaged the circuit board that 721 �Figure 4. First mule deer fawn remotely collared and weighed using an automated device near Horsetooth Reservoir, Colorado, USA, on 17 November 2010. The ﬁrst photo frame shows the female fawn extending her head through an expanded very‐high‐frequency (VHF) radiocollar prior to being collared. The second photo frame, captured 1 second after the ﬁrst, shows a side proﬁle of the fawn in the collaring device with her head extended through the expanded collar and accessing bait. The ﬁnal photo frame, captured 3 seconds after the second photo, shows the fawn standing in the collaring device with the collar around her neck. controlled operation of the device. We replaced the circuit board and eventually replaced the actuator with a heavier‐ duty model. However, until a new actuator could be researched, tested, and installed, we used the same actuator and positioned it diﬀerently so that it was less likely to take on moisture. Several weeks were required to make these 722 modiﬁcations, causing the device to be inoperable from 21 December 2010, through 15 January 2011. On 20 January, we recollared the female fawn that was initially collared on 14 December (it shed the ﬁrst collar on 13 January). After documenting 5 successful collaring events at Horsetooth Reservoir, we moved the device to the Masonville bait site on 21 January 2011. The Masonville bait site was regularly visited by 4 adult males, 3 adult females, and 2 fawns. The adult males aggressively chased the fawns in an attempt to keep them away from the bait, which kept fawns from entering the device. We resolved this behavioral challenge by creating a separate bait site for the adult males a short distance away. It took an additional week of acclimation before a fawn entered the collaring device and accessed the bait pan. After several missed collaring opportunities due to the malfunctioning actuator, we collared a male fawn on 4 February 2011. The other fawn on site showed no interest in extending its head through the expanded collar. After an additional week of observation, we stopped baiting the site on 12 February and moved the device to the Red Feather site. We collared a 7th fawn (female) on 27 February 2011, at the Red Feather site. Unfortunately, the RFID reader failed to detect this collared fawn the following day, allowing the fawn to receive a second collar on 28 February. We suspended collaring eﬀorts to evaluate the RFID failure. It became evident that if a collared fawn entered the device quickly, it could go undetected by the RFID reader. We resolved the issue by reprogramming the device and increasing sensitivity of the RFID reader and antenna. The fawn receiving 2 collars exhibited no ill eﬀects and was behaviorally similar to other fawns; therefore, we did not attempt to capture the fawn to remove a collar. Both collars shed from the deer within approximately 4 weeks as expected. On 17 March 2011, we moved the collaring device to the Heil Valley Ranch site. Deer regularly visiting the site included 4 adult males, 2 adult females, and 1 fawn. Unlike our experience at the Masonville site, we were unable to prevent the adult males from being aggressive toward the adult females and fawn around the collaring device, which prevented the fawn from entering the device. We moved the device to the Hall Ranch site on 24 March 2011, where 3–4 adult males, 2–3 adult females, and 1–3 fawns were using the site. Deer acclimated quickly to the collaring device and we collared an 8th fawn on 28 March, immediately after placing the collar in the device. A few days later we concluded the ﬁeld evaluation because temperatures increased, green forage was abundant, and bears were becoming active. Bears could potentially be attracted to a bait site and might pose a complication with the collaring device. The adjustable opening to the collaring device eﬀectively prevented entry of antlered adult males at all ﬁeld sites throughout the ﬁeld season. Adult females sometimes entered the collaring device but were never collared because they were outside of the programmed mass range and the device controls functioned eﬀectively. Programming Wildlife Society Bulletin • 43(4) �adjustments would have allowed adult females to be collared had that been our objective. Collaring Device Design Modiﬁcations During summer–autumn 2011, we made several modiﬁcations to the device to address issues that arose during the winter ﬁeld season. Notably, we incorporated an alternative release mechanism that used an archery caliper release instead of the existing metal, latch system. We also tested and installed a higher quality actuator that would be more likely to perform adequately under extreme ﬁeld conditions. Finally, we removed the ﬂoor‐mounted camera from the device because it commonly failed to provide useful images and was unnecessary for identifying fawn sex. Subsequent Field Evaluation We conducted a follow‐up ﬁeld evaluation during January– March 2012. We evaluated the collaring device during January 2012 at Hall Ranch near Lyons, Colorado. Bait sites were visited primarily by adult males, which inhibited adult female and fawn access to the bait and collaring device. Therefore, we did not collar any fawns. We moved the collaring device at the end of January to Piceance Basin, southwest of Meeker, Colorado, where a separate deer study was underway and could beneﬁt from additional collared fawns. We initiated our evaluation of the collaring device in Piceance Basin on 5 February 2012. Our ﬁrst monitoring period occurred during 5–10 February. We collared a male fawn on 7 February 2012 during early evening. During 11–18 February, we baited the collaring device but did not monitor deer activity on site or collar any fawns. During 19–24 February, we resumed direct ﬁeld monitoring of the device and consistently observed a mixture of adult females, fawns, and adult males on site. However, no uncollared fawn extended its head through the outstretched collar to access the bait pan, and therefore, we did not collar a fawn. The previously collared fawn was routinely on site and often entered the collaring device; the device controls successfully prevented the fawn from receiving a second collar. During 25 February–3 March, we again baited the collaring device but did not conduct ﬁeld observations or collar any fawns. We resumed ﬁeld monitoring during 4–9 March. Deer consistently accessed the bait site during this period and we collared a female fawn on 9 March 2012. We once again ceased direct ﬁeld monitoring during 10–18 March and completed our ﬁnal monitoring period during 19–21 March. Deer were active on site during this ﬁnal evaluation period, including the previously collared fawns, although we were unable to collar any new fawns. The device functioned eﬀectively, but uncollared fawns did not access bait in the bait pan and therefore were not collared. As during the initial ﬁeld evaluation, we did not detect entry of adult males into the collaring device and no adult females were collared. Our modiﬁcations to the collaring device in 2011 improved functionality of the device in 2012. However, the rate at which fawns were collared was slow. Our ﬁeld Bishop et al. • Automated Collaring Device for Deer observations indicated that deer were hesitant to extend their head and neck through the expanded collar opening to reach the bait pan. DISCUSSION We worked collaboratively with student and professional engineers and used modern technology to develop a technique for noninvasively radiocollaring deer and recording mass and sex. We incorporated electrical and mechanical controls that allowed speciﬁc sizes of animals to be targeted for collaring and prevented individual animals from being collared twice. Considerable troubleshooting was necessary to transform our conceptual model for an automated collaring device into a functional unit. We engineered the device using components that could withstand moisture and cold to allow device operation in various winter weather conditions and temperatures ranging to −30° C. The cost of manufacturing a single collaring device based on our ﬁnal design was approximately US$12,000 in 2012. We remotely radiocollared, weighed, and sexed 8 wild mule deer fawns (≥6 months old), 2 of which received a second collar after shedding their initial collar. Thus, we documented 10 successful collaring events. Our use of partially cut surgical tubing was an eﬀective strategy for causing collars to shed approximately 4 weeks postcollaring. Only one fawn received a second collar while still wearing the ﬁrst collar when the RFID reader failed. Aside from this single incident, collared fawns frequently interacted with the operable device and were not double‐collared. Our ﬁeld observations suggest we could have remotely collared adult female deer had we programmed the device accordingly. The collaring device would not be suitable, however, for remotely collaring neonatal fawns. We did not assess the exact age at which a fawn would be large enough to be collared using this technique. Deer receiving radiocollars in our study were not physically restrained and experienced negligible stress. Some deer showed essentially no response to the collaring event and did not immediately exit the collaring device afterward. Furthermore, no deer were restrained or harmed at any point in the study through interactions with the device. In contrast, traditional techniques cause stress to animals as they are pursued, captured, and handled. The rate at which deer were collared was slow, even when the device was fully functional. Our ﬁeld observations indicated that fawns sometimes required ≥1 weeks of exposure before they entered the device and accessed the bait in the bait pan. Some fawns were reluctant to enter the device even after days or weeks of exposure. We evaluated various baiting strategies. If bait was only placed in the bait pan inside the device, deer groups were not attracted and retained at the site, and therefore, no fawns were present. If too much bait was placed outside of the device, there was no incentive for fawns to enter the device and extend their head through the expanded collar to access bait in the bait pan. Generally, we learned that some bait should be placed outside of the device to attract deer groups to the site and that some bait should be placed on the ﬂoor or scale to lure 723 �fawns into the device. We also tried placing all bait external to the device in buckets to limit the number of animals accessing bait at any given time and make them more accustomed to placing their heads in an enclosed area. However, this technique was not eﬀective in improving fawn collaring rates. Winter weather conditions were overall mild during our study, particularly during winter 2011–2012, which may partly explain the slow rate at which fawns were collared. We would expect the collaring device to be more eﬀective during harsher winter conditions when less forage is available, similar to other bait‐reliant techniques (i.e., clover traps, drop nets). Our greatest challenge was how to eﬀectively bait sites when adult males were present. The problem of adult males dominating a bait site is not unique to this study and has been documented elsewhere when attempting to capture fawns where bait is used to attract animals to a trap (B. A. Banulis, C. J. Bishop, and B. E. Watkins, Colorado Parks and Wildlife, unpublished data). However, in this case, adult males were particularly problematic because they often prevented fawns from accessing the collaring device altogether. We had some success luring adult males away from the collaring device by establishing nearby bait sites, but generally, the presence of adult males inhibited fawn collaring success. Our results suggest potential for sample bias because some fawns were more reticent than others to enter the collaring device. Fawns in poorer body condition or with a stronger drive to access bait may have had greater probabilities of being collared, similar to other bait‐reliant techniques. Presence of adult males also appeared to inﬂuence fawn collaring probability, as noted above. Sample bias could be reduced if future design modiﬁcations made the device more appealing to deer, thereby increasing the likelihood that any target deer approaching the device would receive a collar. Although successful in many ways, the collaring device did not prove at this stage to be an eﬃcient alternative to traditional capture techniques. We had hoped initially that deer collaring rate would be roughly comparable to clover‐ trapping given the similarities of the devices and their reliance on bait. However, our device was less eﬀective than a clover trap because of the added requirement that a deer extend its head and neck through the expanded collar opening to access the bait. If future device modiﬁcations were to reduce deer acclimation time, then our collaring device would have clear advantages over clover traps by reducing personnel time associated with monitoring traps on a strict schedule and handling captured deer. The collaring device would have the added beneﬁt of not subjecting deer to capture‐related stress, myopathy, or mortality. Future work should focus on alternate baiting strategies and design modiﬁcations focused on the collar itself and the collaring apparatus within the device. Deer were often willing to enter the device shortly after it was placed on site but hesitant to extend their head and neck through the outstretched collar. Our past ﬁeld experiences suggested a deer would be willing to extend its head through a narrowed opening to gain access to food (bait) that otherwise was unavailable; however, deer were more reticent than we expected. Additionally, much of the 724 evaluation occurred during mild winter conditions when food was not overly limited. As a next step, we recommend focusing more attention on the design of the collar itself, such as using a more expansive collar material to increase the diameter of the expanded collar. We believe a fawn would more readily extend its head through the outstretched collar within the device if the diameter of the opening was larger. Collar design should accommodate various types of GPS and satellite transmitters given their common use. We also recommend placing >1 collaring devices and several cheap replica devices at a site and placing bait only within devices. This approach would likely compel multiple animals in a group to enter devices simultaneously, which might reduce competitive interactions between adults and fawns for access to any single device. Also, a fawn may be more willing to enter a device and extend its head through the outstretched collar if adults are likewise entering devices. One inherent limitation with this approach is that fawns (i.e., target animals) would only be collared if they entered actual collaring devices, and adults (i.e., non‐target animals) would only gain access to bait if they entered replica devices. It is possible adults would learn this pattern quickly and focus their attention on replica devices, thereby increasing fawn access to collaring devices. Generally speaking, collaring rate should improve signiﬁcantly with any modiﬁcation(s) that eliminates the need for a deer to extend its head and neck through a relatively narrow opening. Collaring rate might also improve with modiﬁcations that make the device look more natural. We also recommend incorporating a design feature that would remotely notify biologists when a collar has been deployed. If these device modiﬁcations proved eﬀective and collaring rate increased, the next phase of design research should focus on how to load a new collar for deployment, after the initial collar is deployed, without human intervention. We did not experience problems with nontarget species accessing the collaring device or interfering with mule deer accessing the device. Similar‐sized animals such as white‐tailed deer (Odocoileus virginianus) and pronghorn (Antilocapra americana) were uncommon or absent in our study areas. We do not believe this technique would work well in areas where mule and white‐tailed deer (or deer and pronghorn) are sympatric unless the study permitted >1 species to be collared. We expect that presence of abundant elk could also pose a problem by preventing deer from accessing the bait. Elk presence might provide another reason for placing multiple collaring or replica devices with bait at an individual site. The devices could be sized to prevent elk access while allowing and possibly incentivizing deer access (i.e., deer could outcompete elk for bait by entering devices). Our research was dependent on a collaborative, iterative design and evaluation process involving a team of ungulate biologists and engineers. Direct collaborations among engineers and biologists are not unique and have undoubtedly led to many of the modern‐day wildlife management techniques currently in use, particularly as it pertains to transmitters used to monitor wildlife (Rodgers and Anson 1994, Wildlife Society Bulletin • 43(4) �Rodgers et al. 1996). However, we believe there are opportunities to more rapidly advance techniques used to capture, mark, and measure wildlife by encouraging more biologist–engineer collaborations. Most of the wildlife capture techniques currently employed have changed very little from their initial development decades ago. This likely reﬂects the fact that wildlife capture is unique to our profession, and therefore, not ripe for large‐scale capital investment. Most technological advancements in the wildlife profession were made possible only because advances had been made in other business and governmental sectors for nonwildlife purposes. Examples include computers, satellites, data storage, remotely triggered cameras, unmanned aerial vehicles (UAVs), and genetics. In the case of animal capture and handling, wildlife professionals will need to take lead on incorporating modern technology to increase technique eﬃciency while minimizing animal handling stress and capture‐related mortality. A source of funding is therefore needed to support capture‐related technique research and development that allows for failures and associated troubleshooting. There is risk that some research will provide no return on the ﬁnancial investment. However, as history has shown, such research investments could lead to new techniques that ultimately save millions of dollars and transform the wildlife management profession. ACKNOWLEDGMENTS We thank D. J. Freddy for providing us the opportunity to pursue this research. S. W. Breck and K. C. VerCauteren provided initial design insights and guidance. We thank the Mechanical Engineering Department at Colorado State University for their work to produce an initial prototype of the collaring device. J. Vigil was instrumental in coordinating workﬂow. Multiple wildlife professionals and technicians with Colorado Parks and Wildlife provided guidance and helped conduct ﬁeld work. We thank D. B. Johnston and D. W. Tripp for reviewing an early draft of this manuscript and Associate Editor A. R. Rodgers and 3 anonymous reviewers for improving the manuscript during the review process. Funding was provided by Colorado Parks and Wildlife, Federal Aid in Wildlife Restoration Project W‐185‐R, and Mule Deer Foundation. LITERATURE CITED Barrett, M. W., J. W. Nolan, and L. D. Roy. 1982. Evaluation of a hand‐ held net‐gun to capture large mammals. Wildlife Society Bulletin 10: 108–114. Beasom, S. L., W. Evans, and L. Temple. 1980. The drive net for capturing western big game. Journal of Wildlife Management 44:478–480. Burton, A. C., E. Neilson, D. Moreira, A. Ladle, R. Steenweg, J. T. Fisher, E. Bayne, and S. Boutin. 2015. Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes. Journal of Applied Ecology 52:675–685. Clover, M. R. 1954. A portable deer trap and catch‐net. California Fish and Game 40:367–373. Clover, M. R. 1956. Single‐gate deer trap. California Fish and Game 42:199–201. Cox, A. S., A. S. C. Knipps, J. L. Wallace, T. E. Boal, P. R. Krausman, D. A. Haukos, and M. Ben‐David. 2018. Journal of Wildlife Management, Wildlife Society Bulletin, and Wildlife Monographs Author Guidelines. The Bishop et al. • Automated Collaring Device for Deer Wildlife Society, Bethesda, Maryland, USA. <http://wildlife.org/wp‐content/ uploads/2018/01/TWS‐Journal‐Guidelines‐2018.pdf>. Accessed 15 Jul 2018. DelGiudice, G. D., B. A. Mangipane, B. A. Sampson, and C. O. Kochanny. 2001. Chemical immobilization, body temperature, and post‐ release mortality of white‐tailed deer captured by Clover trap and net‐ gun. Wildlife Society Bulletin 29:1147–1157. DelGiudice, G. D., B. A. Sampson, D. W. Kuehn, M. Carstensen‐Powell, and J. Fieberg. 2005. Understanding margins of safe capture, chemical immobilization, and handling of free‐ranging white‐tailed deer. Wildlife Society Bulletin 33:677–687. 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. Luikart, G., N. Ryman, D. A. Tallmon, M. K. Schwartz, and F. W. Allendorf. 2010. Estimation of census and eﬀective population sizes: the increasing usefulness of DNA‐based approaches. Conservation Genetics 11:355–373. Lukacs, P. M., G. C. White, B. E. Watkins, R. H. Kahn, B. A. Banulis, D. J. Finley, A. A. Holland, J. A. Martens, and J. Vayhinger. 2009. Separating components of variation in survival of mule deer in Colorado. Journal of Wildlife Management 73:817–826. Morellet, N., H. Verheyden, J.‐M. Angibault, B. Cargnelutti, B. Lourtet, and M. A. J. Hewison. 2009. The eﬀect of capture on ranging behaviour and activity of the European roe deer Capreolus capreolus. Wildlife Biology 15:278–287. 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. O’Connell, A. F., J. D. Nichols, and K. U. Karanth, editors. 2011. Camera traps in animal ecology: methods and analyses. Springer‐Verlag, Tokyo, Japan. Ramsey, C. W. 1968. A drop‐net deer trap. Journal of Wildlife Management 32:187–190. Rodgers, A. R., and P. Anson. 1994. Animal‐borne GPS: tracking the habitat. GPS World 5:20–32. Rodgers, A. R., R. S. Rempel, and K. F. Abraham. 1996. A GPS‐based telemetry system. Wildlife Society Bulletin 24:559–566. Sasse, D. B. 2003. Job‐related mortality of wildlife workers in the United States, 1937–2000. Wildlife Society Bulletin 31:1000–1003. Schmidt, R. L., W. H. Rutherford, and F. M. Bodenham. 1978. Colorado bighorn sheep‐trapping techniques. Wildlife Society Bulletin 6:159–163. Unsworth, J. W., D. F. Pac, G. C. White, and R. M. Bartmann. 1999. Mule deer survival in Colorado, Idaho, and Montana. Journal of Wildlife Management 63:315–326. van Reenen, G. 1982. Field experience in the capture of red deer by helicopter in New Zealand with reference to post‐capture sequela and management. Pages 408–421 in L. Nielsen, J. C. Haigh, and M. E. Fowler, editors. Chemical immobilization of North American wildlife. Wisconsin Humane Society, Milwaukee, USA. VerCauteren, K., J. Beringer, and S. E. Hygnstrom. 1999. Use of netted cage traps for capturing white‐tailed deer. Pages 155–164 in G. Proulx, editor. Mammal trapping. Alpha Wildlife Research and Management, Sherwood Park, Alberta, Canada. 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 R. M. Bartmann. 1994. Drop nets versus helicopter net guns for capturing mule deer fawns. Wildlife Society Bulletin 22:248–252. Wolfe, L. L., W. R. Lance, and M. W. Miller. 2004. Immobilization of mule deer with thiafentanil (A‐3080) or thiafentanil plus xylazine. Journal of Wildlife Diseases 40:282–287. Associate Editor: Rodgers. SUPPORTING INFORMATION Additional supporting information may be found in the online version of this article. Appendix A. Automated collaring device speciﬁcations. 725 � https://cpw.cvlcollections.org/files/original/dfbfbebf248d208cee9610fa3871c1ee.pdf ebedf81f74e974bbe5e56e1ec9f595b8 PDF Text Text RH: Bishop et al. • Automated Collaring Device for Deer A Noninvasive Automated Device for Remotely Collaring and Weighing Mule Deer CHAD J. BISHOP,1,2 Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA MATHEW W. ALLDREDGE, Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA DANIEL P. WALSH,3 Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA ERIC J. BERGMAN, Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA CHARLES R. ANDERSON, JR., Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA DARLENE KILPATRICK, Colorado Parks and Wildlife, 317 W Prospect Road, Fort Collins, CO 80526, USA JOE BAKEL, Dynamic Group Circuit Design, Inc., 2629 Redwing Road, Suite 360, Fort Collins, CO 80525, USA CHRISTOPHE FEBVRE, Dynamic Group Circuit Design, Inc., 2629 Redwing Road, Suite 360, Fort Collins, CO 80525, USA 1 E-mail: chad.bishop@umontana.edu 2 Current affiliation: Wildlife Biology Program, Department of Ecosystem and Conservation Sciences, University of Montana, 32 Campus Drive, Missoula, MT 59812, USA 3 Current affiliation: National Wildlife Health Center, United States Geological Survey, 6060 Schroeder Road, Madison, WI 53711-6223, USA �SUPPORTING INFORMATION APPENDIX A. AUTOMATED COLLARING DEVICE SPECIFICATIONS These specifications were developed prior to design and manufacture of the fully functional collaring device. These were the exact specifications provided in advance to engineers. General Device Specifications • Place an expandable radiocollar around the neck of a deer remotely and noninvasively (i.e., without human presence or animal handling). Deer will be attracted into the device using bait. • Record and store the mass and sex of each deer that receives a radiocollar. • Prevent deer that have been collared previously from receiving a second collar. • Target specific deer by allowing only those deer that are within a specified mass range to receive a collar. • Function properly in various winter weather conditions and at temperatures as low as −32° C. • Allow deer to freely exit through the front entrance or sides of the device at any time. Mechanical Device Specifications • The collaring device frame will be manufactured from aluminum and be roughly 0.9 m wide, 2.4 m long, and 1.2 m high. The device will have 2 compartments: animal compartment (0.9 × 1.5 × 1.2 m) and bait compartment (0.9 × 0.9 × 1.2 m). The 2 compartments will be separated by the collaring mechanism. • The entrance to the animal compartment will be roughly 36 cm wide × 81 cm high and be adjustable. �• The sides of the animal compartment will comprise one-way gates that prevent entry but will allow an animal in the device to exit. • The floor of the animal compartment will comprise a scale so that an animal’s mass will be measured accurately regardless of where an animal stands within the compartment. • A deer in the animal compartment will only be able to access the bait compartment by extending its head through the collaring mechanism. • The collaring mechanism will comprise a rectangular, aluminum frame with machined (i.e., curved), aluminum plates extending from the sides of the rectangle. An expandable radiocollar in the fully expanded position will be placed around the aluminum plates. When a target animal is in the correct position and the collaring mechanism is triggered (see electrical specifications below), ≥1 aluminum plates will collapse causing the collar to be released around the animal’s neck. • When a nontarget deer enters the device, some type of blockade will extend into the opening of the collaring mechanism to physically block access to the bait compartment. The blockade will retract once the nontarget deer exits the device or after some preset time interval. Nontarget deer include those that already have radiocollars or are outside of a specified weight range. • The sides of the bait compartment will be constructed from a combination of steel mesh and netting. Netting will be consistent with that used to construct clover traps or dropnets. • The bait compartment will include an access door to facilitate efficient baiting of the device. �• The top (i.e., roof) of both the animal and bait compartments will comprise removable sheets of clear plexiglass and netting so that users of the device may switch between the two options. • A digital camera will be mounted in the bait compartment to photograph the top of an animal’s head and on the scale platform (i.e., device floor) to photograph an animal’s abdomen. A third digital camera will be mounted in the rear of the animal compartment to photograph the whole animal once it is in the collaring device. • All cameras will be mounted so that the SD card can be easily accessed. If this is not feasible (e.g., camera mounted in scale), then photos will be sent to and/or stored to an SD card externally mounted. • The entire collaring device, with minimal disassembly, will fit in the bed of a standard pickup truck to facilitate easy transport. • The collaring device will have axles and removable wheels to facilitate loading on or off a trailer or truck and to ease positioning of the device on site. Electrical Device Specifications • Continuously record mass of deer in the collaring device and use a specified mass range to determine whether a collar should be released on a deer (i.e., send signal from scale to inform device operation). • Prevent radiocollar from being released when a previously radiocollared deer enters the device. We intend this to be accomplished using an antenna placed around the entrance to the device that is attached to a radio frequency identification (RFID) reader. Each radiocollar will have a PIT tag that will be read by the RFID reader. Thus, the signal from the RFID reader needs to inform device operation. �• Allow a collar to be released only when a target deer (i.e., correct mass and not previously collared) has its head through the collaring mechanism and is accessing the bait. • Record and save mass measurements when a radiocollar is released using a data logger. • To obtain deer sex, trigger 2 digital cameras to take photo(s) when a radiocollar is released on a target deer. One camera will be positioned to take a photograph of the top of the deer’s head and the other camera positioned to take a picture of the deer’s abdomen (see above). A third digital camera will be positioned to record a series of photos of any deer entering the collaring device using standard infrared or motion-sensor technology. • Trigger a type of “blockade” (see mechanical specifications above) to block entry to the collaring mechanism when a nontarget deer is in the device. • Power the entire electrical system from a common battery source accessible from the outside of the device. • Any “reset” button associated with the circuitry that may need to be pressed during operation of the device should be easily accessible from outside of the device. • Design the electrical system so that wiring is contained in the aluminum frame, metal conduit, or other material that protects it from elements and from physical impacts associated with field work. � 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 A noninvasive automated device for remotely collaring and weighing mule deer Description An account of the resource <span>Wildlife biologists capture deer (</span><i>Odocoileus</i><span> spp.) annually to attach transmitters and collect basic information (e.g., animal mass and sex) as part of ongoing research and monitoring activities. Traditional capture techniques induce stress in animals and can be expensive, inefficient, and dangerous. They are also impractical for some urbanized settings. We designed and evaluated a device for mule deer (</span><i>O. hemionus</i><span>) that automatically attached an expandable radiocollar to a ≥6-month-old fawn and recorded the fawn's mass and sex, without physically restraining the animal. The device did not require on-site human presence to operate. Students and faculty in the Mechanical Engineering Department at Colorado State University produced a conceptual model and early prototype. Professional engineers at Dynamic Group Circuit Design, Inc. in Fort Collins, Colorado, USA, produced a fully functional prototype of the device. Using the device, we remotely collared, weighed, and identified sex of 8 free-ranging mule deer fawns during winters 2010–2011 and 2011–2012. Collars were modified to shed from deer approximately 1 month after the collaring event. Two fawns were successfully recollared after they shed the first collars they received. Thus, we observed 10 successful collaring events involving 8 unique fawns. Fawns demonstrated minimal response to collaring events, either remaining in the device or calmly exiting. A fawn typically required ≥1 weeks of daily exposure before fully entering the device and extending its head through the outstretched collar, which was necessary for a collaring event to occur. This slow acclimation period limited utility of the device when compared with traditional capture techniques. Future work should focus on device modifications and altered baiting strategies that decrease fawn acclimation period, and in turn, increase collaring rates, providing a noninvasive and perhaps cost-effective alternative for monitoring mid- to large-sized mammal species.</span> Bibliographic Citation A bibliographic reference for the resource. Recommended practice is to include sufficient bibliographic detail to identify the resource as unambiguously as possible. Bishop, C. J., M. W. Alldredge, D. P. Walsh, E. J. Bergman, C. R. Anderson Jr, D. Kilpatrick, J. Bakel, and C. Febvre. 2019. A noninvasive automated device for remotely collaring and weighing mule deer. Wildlife Society Bulletin 43:717–725. <a href="https://doi.org/10.1002/wsb.1034" target="_blank" rel="noreferrer noopener">https://doi.org/10.1002/wsb.1034</a> Creator An entity primarily responsible for making the resource Bishop, Chad J. Alldredge, Mathew W. Walsh, Daniel P. Bergman, Eric J. Anderson Jr, Charles R. Kilpatrick, Darlene Bakel, Joe Febvre, Christophe Date Created Date of creation of the resource. 2019-11-28 Rights Information about rights held in and over the resource <div class="element"> <div class="element-text"><a href="http://rightsstatements.org/vocab/InC-NC/1.0/" target="_blank" rel="noreferrer noopener">In Copyright - Non-Commercial Use Permitted</a></div> </div> Format The file format, physical medium, or dimensions of the resource application/pdf Language A language of the resource English Subject The topic of the resource Automated Baiting Capture Capture techniques Collaring Fawn Handling Mule deer Noninvasive <em>Odocoileus hemionus</em> Extent The size or duration of the resource. 9 pages Is Part Of A related resource in which the described resource is physically or logically included. Wildlife Society Bulletin Type The nature or genre of the resource Article