SCAT-DETECTION DOGS SURVEY LOW DENSITY MOOSE IN NEW YORK Heidi Kretser1, Michale Glennon1, Alice Whitelaw2, Aimee Hurt2, Kristine Pilgrim3, and Michael Schwartz3 1Wildlife Conservation Society, North America Program, 132 Bloomingdale Avenue, Saranac Lake, New York, USA 1298; 2Working Dogs for Conservation, 52 Eustis Road, Three Forks, Montana, USA 59752; 3United States Forest Service Rocky Mountain Research Station, National Genomics Center for Wildlife and Fish Conservation, 800 East Beckwith, Missoula, Montana 59801 ABSTRACT: The difficulty of collecting occurrence and population dynamics data in mammalian populations of low density poses challenges for making informed management decisions. We assessed the use of scat-detection dogs to search for fecal pellets in a low density moose (Alces alces) popula- tion in the Adirondack Park in New York State, and the success rate of DNA extraction from moose fecal pellets collected during the surveys. In May 2008, two scat-detection dog teams surveyed 20, 4-km transects and located 138 moose scats. In 2011 we successfully amplified DNA from 39 scats (28%) and were able to uniquely identify 25 individuals. Improved storage protocols and earlier lab analysis would increase the amplification success rate. Scat-detection dogs proved to be a reasonable, non-invasive method to collect useful data from the low density moose population in the Adiron- dack Park. ALCES VOL. 52: 55–66 (2016) Key Words: Adirondack Park, DNA, fecal pellets, moose, New York, scat-detection dog. Moose (Alces alces) were nearly extir- pated from the northeastern United States in the late 1800s, but have recently undergone natural recolonization in the region (Alexander 1993, Bontaites and Gustafson 1993, Wattles and DeStefano 2011) and the Adirondack Park in New York (Hicks 1993, Reeves and McCabe 1997, Jenkins and Keal 2004). Moose have no natural predators in New York other than possibly black bears (Ursus americanus) that prey upon neonatal calves, but concerns about over-browsing of regenerating forests, trampling of vacuum tubing in sugar maple (Acer saccharum) stands, and the potential for moose to pose roadway hazards have prompted calls for a hunting season. The recent population de- cline in Minnesota suggests that moose at the southern extent of their range may face thermoregulatory stress that could possibly translate to poor body condition, malnutri- tion, and energy loss making them more sus- ceptible to parasites (Lenarz et al. 2010). Although state wildlife biologists recognize the need to understand their population dy- namics and structure, moose in northern New York occur at low density in small, widely-scattered groups that challenge the collection of meaningful population data. Moose biologists from the region met in 2003 to discuss potential research and management methods to study the low density population in the Adirondack Park (Kretser et al. 2014). GPS radio-collaring of 10 females, aerial surveys, deer hunter surveys, and other non-invasive approaches were pre- sented as viable options for studying this low density population. At the time, cost of GPS Corresponding author: Heidi Kretser, Wildlife Conservation Society, 132 Bloomingdale Avenue, Saranac Lake, NY 12983, hkretser@wcs.org 55 mailto:hkretser@wcs.org radio-collars and the logistical difficulty in capturing moose were considered prohibi- tively expensive, especially in this heavily forested region dominated by dense conifer- ous and mixed forest with minimal road access. Deer hunter surveys began in 2005 with low participation rates, and flyovers occurred when helicopter availability and weather conditions aligned, albeit, not frequently enough to collect robust data. The Wild Center, a local natural history museum, offered pilot funding to test other non-invasive methods, specifically, using scat-detection dogs to collect population data. The initial objective of this study was to assess the use of scat-detection dogs to locate moose scat efficiently as a potential technique to estimate moose abundance in the dense forests of the Adirondack Park. Measuring occurrence and abundance of a wide-ranging mammal at low density poses challenges for biologists desiring to make informed management decisions (MacKay et al. 2008). At low abundance, the effort required to observe or capture individuals may exceed the resources available to obtain adequate and useful data. Methods such as camera trapping and track stations can supply presence/absence information, and in some cases, information about population structure (e.g., identifying males, females, and juve- niles in photographs); however, these meth- ods do not produce DNA samples. Non-invasive techniques such as hair snares and scat sampling are often good alter- natives for obtaining DNA samples. Hair snares work well in situations where bait and lure are used to attract an animal to the site (e.g., Woods et al. 1999), or when survey- ing areas such as feeding sites or habitat features where species congregate or visit regularly (Kendall and McKelvey 2008). Scat collection does not require luring a spe- cies to a specific site, rather, an efficient means of locating scat in a natural setting. Re- cent studies have used fecal DNA to identify individuals, evaluate kinship, and describe distributions and sex ratios in wild popula- tions (Taberlet et al. 1997, Lucchini et al. 2002, Eggert et al. 2003, Bellemain et al. 2005). Because human detection of scats is challenging in a low density population, scat-detection dogs are often used to increase efficiency (Smith et al. 2003, Long et al. 2008). Combining DNA analysis with the use of scat detection dogs eliminates the need to capture, handle, or observe individual animals and minimizes the field time required to collect samples (Kohn and Wayne 1997, Kohn et al.1999). Obtaining DNA from wild animals pro- vides for a variety of uses and approaches to extract relevant population data. Noninvasive genetic samples can be used in a population genetic framework to understand effective population size, gene flow, genetic diversity, and kinship across multiple populations (Schwartz and Monfort 2008). Mitochondrial DNA (mtDNA) can be used to identify indi- vidual species (Foran et al. 1997), nuclear DNA (often using microsatellites) can iden- tify individuals, and sex identification is possible by focusing on specific genes that determine gender (Schwartz and Monfort 2008). These data can be obtained from DNA extracted from blood, tissue, hair, or scat. Sampling of high quality template DNA samples (i.e., tissue and blood) is often invasive, requiring physically handling ani- mals which may entail high cost, physio- logical stress, and/or injury. To date, most research involving scat- detection dogs has focused on carnivores (Long et al. 2008). We sought to assess the feasibility of using these dogs to search for moose fecal pellets in the Adirondack Park and to determine whether DNA extrac- tion from moose fecal pellets was feasible. Several factors influence whether DNA can be extracted from a scat sample including diet, environmental conditions at collection, storage methods, and the specific extraction 56 SCAT-DETECTION DOGS SURVEY MOOSE – KRETSER ET AL. ALCES VOL. 52, 2016 method. This method has been used success- fully to empirically address a variety of ques- tions about carnivores (Smith et al. 2005, Beckmann 2006), and other organisms ran- ging from right whales (Eubalaena glacialis; Rolland et al. 2006) to invasive plants (Goodwin et al. 2010). Ungulate scat has been successfully amplified in a different ecosystem where scat remained frozen throughout the study period (Wasser et al. 2011). Our two primary objectives were to 1) evaluate if scat-detection dogs could effi- ciently locate moose scat in a low density population in the Adirondack Park, and 2) to determine the efficacy of extracting DNA from moose fecal pellets collected in this ecosystem. STUDY AREA The Adirondack Park (Park) in northern New York is a 24,000-km2 mountainous area with more than 3,000 lakes and ponds and 45,000 km of waterways (Fig. 1). Eleva- tion ranges from 305-1671 m and the domin- ant forest types are northern hardwood, conifer, and boreal upland forests. Northern hardwoods include American beech (Fagus grandifolia), yellow birch (Betula allegha- niensis), and sugar maple, with red spruce (Picea rubens) - balsam fir (Abies balsamea) forests at higher elevations and rare alpine vegetation above 1500 m. More than 280 bird, mammal, amphibian, and reptile species inhabit the landscape, alongside 130,000 full- time human residents in 103 rural communi- ties. Nearly half of the land within the Park boundary is privately owned and managed; the public land is permanently protected from development by the New York State (NYS) Constitution. The local economy is based on year-round tourism, commercial forest industry (private land), and govern- mental services (Jenkins and Keal 2004). The Adirondack Park Agency oversees and regulates activities on the privately-owned portions of the Park, and management of the wildlife resources on both public and private land rests with the New York State Depart- ment of Environmental Conservation (DEC), including hunting and responding to human- wildlife conflicts. METHODS Scat Detection The Wildlife Conservation Society (WCS) conducted a pilot test of scat-detection dogs in the northern Adirondacks in partner- ship with Working Dogs for Conservation, Inc. (WDC, Three Forks, Montana, USA). Initially, WCS staff worked with local indivi- duals to locate scat samples from multiple moose throughout the Park. Scats were col- lected on public and private lands including private parcels undergoing active or recent logging that provided contrast to the pro- tected and unlogged state lands. We sent scat samples to the WDC for training dogs on scents associated with scats that reflected the diet of moose in the Park; dogs were trained for 6 weeks followed by a 2-day in situ training. Maps and aerial photographs were used to establish 20 line transects of ~4 km at each site prior to deployment of the dog team. In the challenging terrain, a 4-km tran- sect was considered a reasonable distance for the team to traverse during a one-day session. Each team included one dog handler, one orienteer, and one dog. Each dog wore a GPS unit attached to a work vest to track their movements; likewise, each orienteer carried a GPS to track their movements. Handlers kept the dogs under voice command within 100 m of the transect, and we measured both human and dog tracks by summing the distance be- tween track points recorded every 15 sec. Each transect was labeled with a unique identifier code and described by date, start time, end time, duration, dog and handler, orienteer, temperature, weather at the start of the transect, human track (km), and dog track (km). We recorded the number of ALCES VOL. 52, 2016 KRETSER ET AL. – SCAT-DETECTION DOGS SURVEY MOOSE 57 moose scats, bear scats, and unknown scats; we included bear in the survey effort because both dogs were trained previously on black bear scat. Orienteers used latex gloves and plastic Zip-lock bags to collect and store scats in an attempt to maintain a sterile envir- onment and reduce the potential for cross- contamination of samples. Each scat was assigned a unique identifi- cation number and described by species, dog ! ! ! ! ! ! ! ! ! ! !! ! ! ! ! !! ! ! _̂ _̂ _̂ _̂ Tupper Lake Lake Placid Saranac Lake Lake George Ingraham Pond East Mountain Hardwood Hill Ragged Mountain 16 6 3 Scats located DNA amplified Individual moose identified 6 20 California Road Champion Land 14 1 1 9 Hatch Brook 1 11 Sporting Hill 9 6 5 Wolf Pond Mountain Catamount Mountain 15 12 6 Base of Whiteface VIC Moose Pond Grass Pond 1 1 2 Deer Pond Huntington Cherry Patch 12 1 1 South Meadow 15 1 1 Boreas I 15 9 7 Boreas II 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Fig. 1. Map of transects surveyed by working dog crews and the number of scats detected, amplified, and identified as individual moose at each transect. Dotted line denotes straight line path connecting three transects (Wolf Pond, Champion, and Boreas II) where the same individual moose was identified. Inset: Location of Adirondack Park within New York State. 58 SCAT-DETECTION DOGS SURVEY MOOSE – KRETSER ET AL. ALCES VOL. 52, 2016 or human found, date, time, zone, easting, northing, elevation in feet, canopy (i.e., open (>50%) or closed), forest type (i.e., hardwood, softwood, or mixed), and water characteristic of the site (i.e., wetland, within 100 m of a wetland, or upland). We assigned each scat to 1 of 3 condition categories: 1) excellent – pellets well-formed, moist or wet, and dark color (described as “fresh” in Smith et al. 2003, Mondol et al. 2009), 2) good – pellets becoming unformed and starting to have discoloration, or 3) poor- pellets not formed, dry, light in color, some- times moldy. Scats were stored in Ziploc bags in the field, subsequently transferred to 5 mL plastic vials and covered with etha- nol, and stored in a cool dark basement; duplicates were frozen in Ziploc bags. Ex- tremely moist samples were kept in open, brown paper bags (9×16 cm) to air dry (Franzen et al. 1998, Piggott and Taylor 2003) for 24–48 h prior to storage (Maudet et al. 2004). Scats in ethanol were submitted to the United States Forest Service Rocky Mountain Research Station Wildlife Genet- ics Laboratory (RMRS) in March 2010. DNA Extraction Using moose reference samples from the northeast, we optimized a panel of 9 variable microsatellites in an attempt to uniquely iden- tify individuals from the Park (Wilson et al. 2003, Schmidt et al. 2008, Wilson pers. com- mun.). Reference blood and tissue samples were obtained from collections and sampling of harvested and vehicular-killed moose from 4 northeastern states (New York, New Hamp- shire, Maine, and Vermont) and 4 Canadian provinces (New Brunswick, Ontario, Nova Scotia, and Quebec). The Nova Scotia sam- ples included moose from the mainland and Cape Breton Island. We performed an initial DNA extraction on 140 fecal pellet samples (stored in ethanol) using a standard protocol developed for ungulate fecal pellets (Maudet et al. 2004, Schwartz et al. 2007). We then performed a DNA extraction on these samples with a pellet swab to test the effi- cacy of this approach; after two failed attempts, we repeated this process on frozen duplicate samples (n = 42). DNA from reference samples was ampli- fied at the following 8 microsatellite loci: NVHRT21, BM1225, BM4516, FCB193, MAP2C, RT5, RT9, and RT30 (Wilson et al. 2003, Schmidt et al. 2008, Wilson pers. com- mun.). The reaction volume (10 μl) contained 1.0 μL DNA, 1× reaction buffer (Applied Biosystems), 2.0 mM MgCl2, 200 μM of each dNTP, 1 μM reverse primer, 1 μM dye- labeled forward primer, 1.5 mg/mL BSA, and 1U Taq polymerase (Applied Biosys- tems). The PCR profile was (94 °C/5 min, 94 °C/1 min, 55 °C/1 min, 72 °C/30 s) × 45 cycles. The resultant products were visua- lized on a LI-COR DNA analyzer (LI-COR Biotechnology). All non-invasive samples were initially amplified twice using the multi- tube approach (Eggert et al. 2003, Schwartz et al. 2004), and allele scores were entered only when consistent for both amplifications. Microsatellite data were checked for geno- typing errors (false alleles, allelic dropout and scoring errors) using the program Drop- out (McKelvey and Schwartz 2005, Schwartz et al. 2006). Microsatellite data was also error- checked with the program Micro-Checker (Van Oosterhout et al. 2004) to identify loci with possible genotyping errors leading to homozygote excess. We calculated the prob- ability of identity (PID) and probability of identify given siblings (Psib) from these samples. We used Chi-square with a Cochran’s test of linear trend to assess the relationship of scat condition, forest type, and wetland proximity to successful DNA extraction. We then evaluated all possible general linear model (GLM) combinations of these factors in SYSTAT with Akaike Information ALCES VOL. 52, 2016 KRETSER ET AL. – SCAT-DETECTION DOGS SURVEY MOOSE 59 Criterion (AIC) to assess their relative im- portance on our ability to successfully ex- tract DNA. RESULTS The 20 transects were sampled on 10 field days in May 2008. Dogs located 191 scats and 4 additional scats were located by the orienteers; 134 (69%) were moose, 56 (29%) bear, and 5 (2%) were unknown. The proportions of moose scat relative to condi- tion were 18% excellent, 45% good, and 36% poor; 63 and 37% of scats were col- lected on private and state lands, respectively. Dogs traveled 134 km and orienteers 114 km during the surveys. We genotyped 270 tissues and hair sam- ples from 10 locations at 8 variable microsat- ellite loci (Table 1). We obtained quality multi-locus genotypes from 28 pellet samples using the initial DNA extraction from fecal pellets. The swabbing method yielded quality DNA from 9 additional pellet samples, and the duplicate samples yielded 2 additional samples. DNA was successfully amplified from 28% (39 of 137) of collected scats using the 8 loci. Errors were identified in 3 samples at loci RT9 and NVHRT21. DNA from these samples was reanalyzed following the ap- proach of Schwartz et al. (2006) until no errors occurred. We identified 25 unique individuals. There was a 1 in 9,737 chance of identifying 2 individuals as identical (PID = 1.03×10−4), and a 1 in 64 chance of identifying 2 siblings as identical (Psib = 1.57×10−2). In 8 cases we identified the same moose from multiple scat, and one moose was identified on 3 different transects located >40 miles distant (Fig. 1). Successful amplification was achieved in 52% of excellent, 31% good, and 12% of poor scats; of the total amplified sample (n = 38), 34% were excellent, 50% good, and 16% were poor scats (Table 2). Scat con- dition (χ2 = 7.928, P < 0.001) and forest type (χ2 = 7.928, P < 0.05) affected our ability to successfully extract DNA; excellent scats and scats located in hardwoods had the high- est success rates. Location was not related to successful extraction (Table 2). The GLM Table 1. Origin and number of tissue samples received and analyzed to create markers for use in the DNA extraction of moose fecal scats, Adirondack Park, New York. Country Location Received Used US New York 16 16 US New Hampshire 30 30 US Maine 41 41 US Vermont 31 31 CA Nova Scotia – Mainland 31 29 CA Nova Scotia – Cape Breton Island 9 9 CA New Brunswick 46 46 CA Ontario 29 26 CA Quebec – RFPL 13 13 CA Quebec – PLC 29 29 Total 275 270 Table 2. The rate (%) of successful extraction of DNA (yes = 38, no = 99) from moose scat relative to scat condition, forest type, and microhabitat location in the Adirondack Park, New York. Sample sizes in parentheses. Variable Yes No Scat condition* Excellent (25) 52% 48% Good (62) 31% 69% Poor (50) 12% 88% Forest type† Hardwood (74) 36% 50% Softwood (45) 22% 78% Mixed (18) 6% 94% Location‡ Immersed in Water (2) 50% 50% Near water (24) 21% 79% Upland (111) 29% 71% *χ2= 13.782, P < 0.001. †χ2= 7.928, P = 0.019. ‡χ2= 1.131, P = 0.568. 60 SCAT-DETECTION DOGS SURVEY MOOSE – KRETSER ET AL. ALCES VOL. 52, 2016 underscored the relative importance of scat condition and forest types. The top four GLMs included scat condition, with the top model indicating that 69.7% of model weight was associated with scat condition and forest type (Table 3). DISCUSSION We demonstrated that scat-detection dogs were effective at locating moose scat in a low density moose population in the dense forests of the Adirondack Park. Scat- detection dogs are more frequently used in carnivore research because of their obvious advantage in sampling wide-ranging and low abundance populations (MacKay et al. 2008); their use in ungulate research is less common. Wasser et al. (2011) used these dogs to locate a variety of species including moose and woodland caribou (Rangifer tarandus caribou) in areas proximal to the Alberta tar sands, and successfully extracted DNA from scats of both. Similarly, dogs located scats of a variety of deer species (Mazama spp.) in Brazil and outperformed human searchers; humans located zero scats whereas dogs located 0.21 scats/km (de Olivera et al. 2012). The success rate of our dogs was ~1.4 samples/km, confirming our supposition that dogs could efficiently ‘sample’ a low density moose population which cannot be easily observed/sampled. Although we successfully amplified DNA from moose fecal pellets, our success rate was relatively low (<30%) but was explained by pellet condition and location. Age and environmental factors (e.g., precipi- tation, temperature) affect the quality of col- lected scats (Brinkman et al. 2010), and these factors also affect detection rate (Reed et al. 2011). These factors undoubtedly affected the quality of our samples as we col- lected scat in the spring, relatively soon after snowmelt. Environmental conditions are dif- ferent within the 3 forest types, with mixed and softwood stands moister at the forest floor which would presumably degrade scat faster; in fact, extraction rates were higher in scats collected in hardwood forest (Table 2). Collection of fresh scats across seasons and repeat sampling would improve our sampling protocol. For example, in an area of winter concentration of moose, repeat sampling would minimize exposure of fecal pellets to the elements (see Brinkman et al. 2010). Future work may also take advantage of in situ photographs of scats to compare con- dition across sites, and relate condition to amplification success more objectively. Storage methods and storage time had strong influence on our success rate of DNA amplification. We stored scats in ethanol based on the best information available at the time, and although a common storage method, it was not ideal in our study. Researchers exam- ining relative success rates associated with vari- ous storage media hesitate to provide overall Table 3. Model selection results using AIC for General Linear Models to predict successful DNA extraction from moose scat collected in the Adirondack Park, New York. Model Rank Variables AIC ΔAIC AIC weights 1 Scat Condition + Forest Type 157.53 0.00 0.697 2 Scat Condition + Forest Type + Water 159.98 2.45 0.204 3 Scat Condition 162.07 4.54 0.072 4 Scat Condition + Water 164.51 6.98 0.021 5 Forest Type 168.43 10.90 0.003 6 Forest Type +Water 168.65 11.12 0.003 7 Water 175.46 17.93 0.000 ALCES VOL. 52, 2016 KRETSER ET AL. – SCAT-DETECTION DOGS SURVEY MOOSE 61 recommendations because of inconsistency among studies (Schwartz and Monfort 2008); however, ethanol was considered the worst storage medium among 3 alternatives (Soto-Calderón et al. 2009). The relatively long storage time of our samples (2 years) was probably the major reason for our low success rate with amplification; however, we still recovered DNA from ~25% of the highly variable sample and >50% of excel- lent pellets. Schwartz and Monfort (2008) suggest processing samples immediately because DNA is more stable in laboratory buffers than fecal material, but the logistical issues of fieldwork would often preclude this approach. However, DNA swabbed from moose fecal pellets and processed within a few weeks of collection yielded high amplifi- cation rates (90%; K. Pilgrim, RMRS, unpub- lished data). Field swabbing of fresh samples also results in higher amplification rates com- pared to swabbing frozen samples (Rutledge et al. 2009). Scat-detection dogs offer many advan- tages but with certain considerations. Cost is a concern for any field-based project and sub- stantial costs are associated with dog and handler selection and training, as well as field time for executing transects (MacKay et al. 2008). In this study we spent $25,000 to hire WDC to find 191 scats, at a per-unit cost of ~$130 per sample. Numerous scats are required to address population studies and depending upon animal density, may require substantial field time and cost. Researchers must understand and work within the physical limitations of the dog and recog- nize that detection rates often vary among dog/handler teams. The handler must ensure that the dog focuses on the desired scat and is not inadvertently trained to non-target scat; this is particularly salient for handlers when target and non-target species have mor- phologically similar fecal pellets. Lastly, this method may result in real or perceived potential conflict with wildlife (MacKay et al. 2008), and the presence of dogs in a given en- vironment may result in unforeseen conflicts with local wildlife. Despite certain limitations, there are nu- merous and obvious advantages in using scat-detection dogs. In comparison to human searchers, dogs are highly efficient and ef- fective at locating scats (de Olivera et al. 2012), and in this study, covered ~20% more ground than humans and collected all but 4 scats. Dogs create minimal sampling bias (MacKay et al. 2008), allowing for quick confirmation of occupancy of the study area by target species. Collection of scats ul- timately provides for discrimination between species and individuals, and has proven ap- plicable to a wide variety of species and habi- tat types. Collection of scat not only allows for subsequent assessment of population structure, it also provides opportunities to ex- plore additional factors such as stress levels and diet. The charismatic and broad public appeal of using dogs should not be dis- counted as an opportunity for public outreach and engagement (MacKay et al. 2008, Woollett et al. 2014). One of the lasting impacts of using scat-detection dogs was the creation of two high definition videos de- scribing the project and highlighting the dogs in the field. According to The Wild Cen- ter staff, these two films continue to capture audiences largely due to the appeal of the dogs performing in the field. Our pilot research in the Adirondack Park of New York State is one of a limited number of studies in which scat-detection dogs have been used in ungulate research, and these dogs provided a viable method for sampling a low density moose population. We also found that forest type, the condition of fecal pellets, storage method, and storage time influenced the efficacy of DNA amplifi- cation. The impact of these factors can be controlled through improved study design that addresses temporal sampling, field swab- bing, shorter storage time, and performing 62 SCAT-DETECTION DOGS SURVEY MOOSE – KRETSER ET AL. ALCES VOL. 52, 2016 extractions soon after sample collection. In particular, we recommend swabbing the scats using synthetic swabs (e.g., Dacron Swabs), at least two swabs per sample, and storing them dry in envelopes or vials. Fecal pellets can be stored in vials of 95% ethanol at room temperature, frozen in secure plastic bags, or air dried and kept at room tempera- ture. Ideally, the swabs and scats would be submitted for DNA extraction and analyses within a few days of collection. This ap- proach works for a variety of carnivores and would be an improved protocol for moose re- search (Reed et al. 2004, McKelvey et al. 2006, Rutledge et al. 2009, Anwar et al. 2011). Our data and that collected in subsequent and future surveys provide an important foundation to understand habitat use and population dynamics of moose in the Park, and conduct genetic research to determine relationships among Park and regional moose. Given the increased interest and fund- ing available for moose research in New York State, we encourage continued use of scat- detection dogs, in concert with other techni- ques, to monitor and study the low density moose population in the Adirondack Park. ACKNOWLEDGEMENTS Funding for this project was provided by The Wild Center, the Northeastern States Re- search Cooperative, and the New York State Department of Environmental Conservation. We thank E. Reed, New York Department of Environmental Conservation, for providing transect site maps, our orienteers G. Lee and B. Kitchen, and many individuals who pro- vided access to private lands and assisted with the data collection and general field work details: Upland Forestry, P. Bogdanovich, The Nature Conservancy, M. Carr, J. Fogarty, and S. Moody and A. Brown. Finally, to moose biologists in the northeast who assisted with collection of tissue samples for DNA markers: K. Hynes, New York Department of Environmental Conservation; C. Alexander, Vermont Department of Fish and Wildlife; K. Rines, New Hampshire Fish and Game Department; L. Kantar, Maine Department of Inland Fish and Wild- life; A. R. Rodgers – Ontario Ministry of Nat- ural Resources; P. Wilson – Trent University, Ontario; C. Dussault, D. Jean, S. Lefort, and B. Beaudoin – Quebec Ministère des Ressources Naturelles et de la Faune; K. Craig and D. Sabine - New Brunswick De- partment of Natural Resources; S. McBurney and T. Nette - Nova Scotia DNR and the Canadian Cooperative Wildlife Health Unit. REFERENCES ALEXANDER, C. E. 1993. The status and man- agement of moose in Vermont. Alces 29: 187–195. ANWAR, M. B., R. JACKSON, M. S. NADEEM, J. E. JANEČKA, S. HUSSAIN, M. A. BEG, G. MUHAMMAD, and M. QAYYUM. 2011. Food habits of the snow leopard Panthera uncia (Schreber, 1775) in Baltistan, Northern Pakistan. European Journal of Wildlife Research 57: 1077–1083. BECKMANN, J. P. 2006. Carnivore conserva- tion and search dogs: the value of a novel, non-invasive technique in the Greater Yellowstone Ecosystem. Pages 28–34 in A. Wondrak Biel, editor. Greater Yellow- stone Public Lands: A Century of Dis- covery, Hard Lessons, and Bright Prospects. Proceedings of the 8th Biennial Scientific Conference on the Greater Yellowstone Ecosystem. Yellowstone Center for Resources, Yellowstone National Park, Wyoming, USA. BELLEMAIN, E., J. E. SWENSON, D. TALLMON, S. BRUNBERG, and P. TABERLET. 2005. Estimating population size of elusive animals with DNA from hunter- collected feces: four methods for brown bears. Conservation Biology 19: 150–161. ALCES VOL. 52, 2016 KRETSER ET AL. – SCAT-DETECTION DOGS SURVEY MOOSE 63 BONTAITES, K., and K. GUSTAFSON. 1993. The history and status of moose and moose management in New Hampshire. Alces 29: 163–167. BRINKMAN, T. J., M. K. SCHWARTZ, D. K. PERSON, K. L. PILGRIM, and K. J. HUNDERTMARK. 2010. Effects of time and rainfall on PCR success using DNA extracted from deer fecal pellets. Conser- vation Genetics 11: 1547–1552. DE OLIVEIRA, M. L., D. NORRIS, J. F. M. RAMÍREZ, P. H. DE PEREZ, M. GALETTI, and J. M. B. DUARTE. 2012. Dogs can de- tect scat samples more efficiently than humans: an experiment in a continuous Atlantic Forest remnant. Zoologica 29: 183–186. EGGERT, L. S., J. A. EGGERT, and D. S. WOODRUFF. 2003. Estimating population sizes for elusive animals: the forest ele- phants of Kakum National Park, Ghana. Molecular Ecology 12: 1389–1402. FORAN, D. R., K. R. CROOKS, and S. C. MINTA. 1997. Species identification from scat: an unambiguous genetic method. Wildlife Society Bulletin 25: 835–839. FRANTZEN, M. A. J., J. B. SILK, J. W. H. FERGUSON, R. K. WAYNE, and M. H. KOHN. 1998. Empirical evaluation of preservation methods for faecal DNA. Molecular Ecology 7: 1423–1428. GOODWIN, K. M., R. E. ENGEL, and D. K. WEAVER. 2010. Trained dogs outperform human surveyors in the detection of rare spotted knapweed (Centaurea stoebe). Invasive Plant Science Management 3: 113–121. HICKS, A. C. 1993. Using road-kills as an index to moose population change. Alces 29: 243–247. JENKINS, J. C., and A. KEAL. 2004. The Adi- rondack Atlas. Syracuse University Press, Syracuse, New York, USA. KENDALL, K. C., and K. S. MCKELVEY. 2008. Hair collection. Pages 141–182 in R. A. Long, P. Mackay, W. J. Zielinski, and J. C. Ray, editors. Non-invasive Survey Methods for Carnivores. Island Press, Washington D. C., USA. KOHN, M. H., and R. K. WAYNE. 1997. Facts from feces revisited. Trends Ecological Evolution 12: 223–227. ———, E. C. YORK, D. A. KAMRADT, G. HAUGHT, R. M. SAUVAJOT, and R. K. WAYNE. 1999. Estimating population size by genotyping faeces. Proceedings of the Royal Society B: Biological Sciences 266: 657–663. KRESTER, H. E., M. G. GLENNON and Z. P. SMITH. 2014. NGOs enhance state wild- life agencies’ capacity to meet public trust doctrine obligations. Human Dimensions of Wildlife 19: 437–447. LENARZ, M. S., J. FIEBERG, M. W. SCHIRAGE, and A. J. EDWARDS. 2010. Living on the edge: viability of moose in northeastern Minnesota. Journal of Wildlife Manage- ment 74: 1013–1023. LONG, R. A., P. MACKAY, W. J. ZIELINSKI, and J. C. RAY. 2008. Non-invasive Sur- vey Methods for Carnivores. Island Press, Washington D. C., USA. LUCCHINI, V., E. FABBRI, F. MARUCCO, S. RICCI, L. BOITANI, and E. RANDI. 2002. Noninvasive molecular tracking of col- onizing wolf (Canis lupus) packs in the western Italian Alps. Molecular Ecology 11: 857–868. MACKAY, P., D. A. SMITH, R. A. LONG, and M. PARKER. 2008. Scat detection dogs. Pages 182–222 in R. A. Long, P. Mackay, W. J. Zielinski, and J. C. Ray, editors. Non-invasive Survey Methods for Carnivores. Island Press, Washington D. C., USA. MAUDET, C., G. LUIKET, D. DUBRAY, A. VON HARDENBERG, and P. TABERLET. 2004. Low genotyping error rates in wild ungulate faeces sampled in winter. Molecular Ecology Notes 4: 772–775. MCKELVEY, K. S., and M. K. SCHWARTZ. 2005. DROPOUT: a program to identify problem loci and samples for non- invasive genetic samples in a capture- 64 SCAT-DETECTION DOGS SURVEY MOOSE – KRETSER ET AL. ALCES VOL. 52, 2016 mark-recapture framework. Molecular Ecology Notes 5: 716–718. ———, J. VON KIENAST, K. B. AUBRY, G. M. KOEHLER, B. T. MALETZKE, J. R. SQUIRES, E. L. LINDQUIST, S. LOCH, and M. K. SCHWARTZ. 2006. DNA ana- lysis of hair and scat collected along snow tracks to document the presence of Canada lynx. Wildlife Society Bulletin 34: 451–455. MONDOL, S., K. U. KARANTH, N. S. KUMAR, A. M. GOPALASWAMY, A. ANDHERIA, and U. RAMAKRISHNAN. 2009. Evalu- ation of non-invasive genetic sampling methods for estimating tiger population size. Biological Conservation 142: 2350–2360. PIGGOTT, M. P., and A. C. TAYLOR. 2003. Ex- tensive evaluation of faecal preservation and DNA extraction methods in Austra- lian native and introduced species. Aus- tralian Journal of Zoology 51: 341–355. REED, J. E., R. J. BAKER, W. B. BALLARD, and B. T. KELLY. 2004. Differentiating Mexican gray wolf and coyote scats using DNA analysis. Wildlife Society Bulletin 32: 685–692. ———, A. L. BIDLACK, A. HURT, and W. M. GETZ. 2011. Detection distance and en- vironmental factors in conservation de- tection dog surveys. Journal of Wildlife Management 75: 243–251. REEVES, H. M., and R. E. MCCABE. 1997. Of moose and man. Pages 1–75 in A. W. Franzmann and C. C. Schwartz, editors. Ecology and Management of North America Moose. Smithsonian Institution Press, Washington D. C., USA. ROLLAND, R. M., P. K. HAMILTON, S. D. KRAUS, B. DAVENPORT, R. M. GILLETT, and S. K. WASSER. 2006. Faecal sam- pling using detection dogs to study repro- duction and health in North Atlantic right whales (Eubalaena glacialis). Journal of Cetacean Research Management 8: 121–125. RUTLEDGE, L. Y., J. J. HOLLOWAY, B. R. PATTERSON, and B. N. WHITE. 2009. An improved field method to obtain DNA for individual identification from wolf scat. Journal of Wildlife Management 73: 1430–1435. SCHMIDT, J. I., K. J. HUNDERTMARK, R. T. BOWYER, and K. G. MCCRACKEN. 2008. Population Structure and Genetic Diversity of Moose in Alaska. Journal of Heredity 100: 170–180. SCHWARTZ, M. K., S. A. CUSHMAN, K. S. MCKELVEY, J. HAYDEN and C. ENGKJER. 2006. Detecting genotyping errors and describing American black bear movement in northern Idaho. Ursus 17: 138–148. ———, G. LUIKART, R. S. WAPLES. 2007. Genetic monitoring as a promising tool for conservation and management. Trends in Ecological Evolution 22: 25–33. ———, and S. L. MONFORT. 2008. Genetic and endocrine tools for carnivore sur- veys. Pages 238–262 in R. A. Long, P. Mackay, W. J. Zielinski, and J. C. Ray, editors. Non-invasive Survey Methods for Carnivores. Island Press, Washington D. C., USA. ———, K. L. PILGRIM, K. S. MCKELVEY, E. L. LINDQUIST, J. J. CLAAR, S. LOCH, and L. F. RUGGIERO. 2004. Hybridization be- tween Canada lynx and bobcats: genetic results and management implications. Conservation Genetics 5: 349–355. SMITH, D. A., K. RALLS, B. L. CYPHER, and J. E. MALDONADO. 2005. Assessment of scat-detection dog surveys to determine kit fox distribution. Wildlife Society Bul- letin 33: 897–904. ———, ———, A. HURT, B. ADAMS, M. PARKER, B. DAVENPORT, and J. E. MALDONADO. 2003. Detection and ac- curacy rates of dogs trained to find scats of San Joaquin kit foxes (Vulpes macrotis mutica). Animal Conservation 6: 339– 346. SOTO-CALDERÓN, I. D., S. NTIE, P. MIKALA, F. MAISELS, E. J. WICKINGS, and N. M. ANTHONY. 2009. Effects of storage type and time on DNA amplification success ALCES VOL. 52, 2016 KRETSER ET AL. – SCAT-DETECTION DOGS SURVEY MOOSE 65 in tropical ungulate faeces. Molecular Ecological Resources 9: 471–479. TABERLET, P., J. J. CAMARRA, S. GRIFFIN, E. UHRES, O. HANOTTE, L. P. WAITS, C. DUBOIS-PAGANON, T. BURKE, and J. BOUVET. 1997. Noninvasive genetic tracking of the endangered Pyrenean brown bear population. Molecular Ecol- ogy 6: 869–876. VAN OOSTERHOUT, C., W. F. HUTCHINSON, D. P. WILLS, and P. SHIPLEY. 2004. Micro-Checker: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes 4: 535–538. WASSER, S. K., J. L. KEIM, M. L. TAPER, and S. R. LELE. 2011. The influences of wolf predation, habitat loss, and human activity on caribou and moose in the Al- berta oil sands. Frontiers of Ecological Environment 9: 546–551. WATTLES, D.W., and S. DESTEFANO. 2011. Status and management of moose in the northeastern United States. Alces 47: 53–68. WILSON, P. J., S. GREWAL, A. RODGERS, R. REMPEL, J. SAQUET, H. HRISTIENKO, F. BURROWS, R. PETERSON, and B. N. WHITE. 2003. Genetic variation and population structure of moose (Alces alces) at neutral and functional DNA loci. Canadian Journal of Zoology 81: 670–683. WOODS, J. G., D. PAETKAU, D. LEWIS, B. N. MCLELLAN, M. PROCTOR, and C. STROBECK. 1999. Genetic tagging free- ranging black and brown bears. Wildlife Society Bulletin 27: 616–627. WOOLLETT (SMITH), D. A., A. HURT, and N. L. RICHARDS. 2014. The current and fu- ture roles of free-ranging detection dogs in conservation efforts. Pages 239–264 in M. E. Gompper, editor. Free-Ranging Dogs and Wildlife Conservation. Oxford University Press, Cambridge, England. 66 SCAT-DETECTION DOGS SURVEY MOOSE – KRETSER ET AL. ALCES VOL. 52, 2016 SCAT-ETECTION DOGS SURVEY LOW DENSITY MOOSE IN NEW YORK STUDY AREA METHODS Scat Detection DNA Extraction RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES