CLPEX Chat Board

Using Patterns in Track-Plate Footprints to Identify Individual Fishers

CARL J. HERZOG, New York Department of Environmental Conservation, 625 Broadway, Albany, NY 12233-4750, USA
ROLAND W. KAYS,1 New York State Museum, 3140 CEC, Albany, NY 12230, USA
JUSTINA C. RAY, Wildlife Conservation Society Canada, 720 Spadina Avenue #600, Toronto, ON M5S 2T9, Canada
MATTHEW E. GOMPPER, Department of Fisheries and Wildlife Sciences, University of Missouri, Columbia, MO 65211-7240, USA
WILLIAM J. ZIELINSKI, United States Department of Agriculture Forest Service, Pacific Southwest Research Station, Redwood Sciences Laboratory,
Arcata, CA 95521, USA
RICHARD HIGGINS, New York Division of Criminal Justice Services, 4 Tower Place, Albany, NY 12203, USA
MIKE TYMESON, New York Division of Criminal Justice Services, 4 Tower Place, Albany, NY 12203, USA

ABSTRACT If individuals can be identified from patterns in their footprints, noninvasive survey methods can be used to estimate abundance. Track plates capture fine detail in the footprints of fishers (Martes pennanti), recording rows of dots corresponding to tiny papillae on the animal’s metacarpal pad. We show that the pattern of these dots can be used to identify individual fishers, similar to human fingerprints. A probabilistic model of uniqueness based on variation in spacing between 1,400 pairs of dots that we measured in prints of 14 different fisher feet suggests the probability of encountering a similar pattern in the print of a different foot by chance alone is ,0.35n, where n¼the number of dot pairs examined. This predicts a 0.00003 probability that a match made using 10 pairs of dots is false. Dot spacing from footprints made by the same foot was remarkably consistent (r¼0.02 mm, n¼24 dot pairs). Combined, these results suggest dot patterns in fisher footprints were unique to individuals and were consistently reproduced on track plates. Empirical tests of matching accuracy were best with good-quality prints, highlighting the need for experience judging when prints are usable. We applied print matching to fisher detections collected on track plates deployed at 500-m intervals along 10 3.5-km transects in the Adirondack region of New York, USA. Of 62 fisher detections, 85% had 1
footprint of suitable quality to compare with other high-quality prints. We found that most detections from a transect were from the same individual fisher suggesting nonindependence of detections. Thus, data from traditional track-plate deployments over small time periods cannot
be used as a measure of abundance, but new study designs using print matching could obtain robust noninvasive, mark–recapture density
estimates. ( JOURNAL OF WILDLIFE MANAGEMENT 71(3):955–963; 2007)
DOI: 10.2193/2006-408
KEY WORDS census, fingerprint, fisher, footprint, identification, Martes pennanti, track plate.
Establishing and monitoring abundance of rare species remains a challenge, particularly for forest mammals. Abundance data can be collected through live-trapping but this is invasive and can be logistically difficult at scales relevant for wide-ranging species. As a result, there has been
increased attention paid to methods of obtaining abundance information over large areas that are noninvasive and require less manual labor (Zielinski and Kucera 1995, Harrison
et al. 2002, Gompper et al. 2006).
Accurate methods to estimate abundance require unambiguous recognition of individuals (Otis et al. 1978). Identifying individuals through genetic fingerprinting of hair or feces is increasingly common but can be expensive, laborious, and error prone (Woods et al. 1999, Creel et al.
2003). Camera traps are increasingly favored for carnivore surveys but, when applied to density estimation, generally are limited in usefulness to species with individually unique pelage patterns (Karanth 1995, Karanth and Nichols 1998, Trolle and Ke´ry 2003) or populations that have been
otherwise captured and tagged with large unique markers (e.g., ear tags; Fuller et al. 2001). Foot prints from natural substrates and track-boxes have long been used to identify species presence. Some efforts have been made to identify individuals of certain large species from gross footprint morphology (Panwar 1979, Riordan 1998, Jewell et al. 2001,
Sharma et al. 2005), but these methods have not been widely applied and their validity has been questioned (Hamm et al. 2003, Karanth et al. 2003).Although observational studies routinely use unique animal pelage patterns or facial markings to identify individuals (e.g., Goddard 1966, van Lawick-Goodall 1968, Peterson 1972, Scott 1978) most do not address the possibility of pattern duplication across individuals and variability is assumed to result in combinations that are
unique or at least practically unique. A few studies have usedrudimentary information-theory techniques (Shannon 1948) to estimate the likelihood of a duplicate occurring given the
overall frequency of character occurrence (Pennycuick and Rudnai 1970, Miththapala et al. 1989). The most rigorous analyses of individual marking uniqueness have come from the study of human fingerprints (e.g., Roddy and Stosz 1997, Pakanti et al. 2002). These include creation of
statistical pattern-models to estimate the probability of false duplication, and empirical comparisons of a large number of prints using a defined matching technique. The purpose of our study is to show that individual fishers (Martes pennanti) can be recognized from footprints made at enclosed track-plates and to apply this method to field data obtained from free-ranging fishers. 1 E-mail: rkays@mail.nysed.gov
Herzog et al. Using Footprints to Identify Fishers 955