Austral Ecology · 2021 · 12 citations · 30 references
Life SciencesEcologyZoologyHerpetologyOrnithologyPopulation EcologyAmphibian BiologyAnimal IdentificationSpecies IdentificationThreatened Species MonitoringWildlife ConservationWildlife EcologyAmphibian Population MonitoringDorsal PatterningMarkingsNatural Feature MarkersPhotographic IdentificationSpecies-level Visual IdentificationThreatened SpeciesVisual Animal Identification
Abstract Individuals within amphibian populations are commonly identified using artificial marking techniques, such as toe clipping and microchipping. However, many species in this group may be strong candidates for visual identification from photographs given intraspecific variability in skin features. We investigated the potential for dorsal skin patterns to be used as natural markers for the photo‐identification of both juveniles and adults of the green and golden bell frog ( Litoria aurea ). This is a threatened species that has come under intense population monitoring using capture‐mark‐recapture procedures primarily involving the use of artificial markers, with no apparent investigation of the potential for natural markers to be used instead. We collected photographs of marked individuals to determine the level of intraspecific variability in dorsal patterning within a population. This photo‐database was subsequently used to create an online survey in which participants were asked to match separate images of query frogs from small image pools by comparing dorsal patterns. Photographs were taken on a smartphone device under field conditions to test whether this technique could be applied to the study of wild populations with little cost or expertise required. We showed that dorsal patterns are clear and distinct among L . aurea individuals and easily visualised from field‐acquired images to detect recapture events by eye with a low error rate. While an overwhelming majority of adults possessed dorsal patterning that can be easily distinguished by eye, juveniles often showed a complete absence of patterning, suggesting that photo‐identification may be more effective for adult stages. Nevertheless, we highlight the feasibility of collecting visual information on the natural markings of a threatened anuran, providing evidence that it may be used as a supplementary form of identification alongside more traditional techniques, highlighting a potential direction for the future monitoring of this species.
30
R: A Language and Environment for Statistical Computing
R Core Team · 2000 · 353K citations · Full text
Natural Sciences, Computer Science, Statistical Computing +10
Object recognition from local scale-invariant features
David Lowe · 1999 · 16.2K citations
ESTIMATION OF TIGER DENSITIES IN INDIA USING PHOTOGRAPHIC CAPTURES AND RECAPTURES
K. Ullas Karanth, James D. Nichols · Ecology · 1998 · 1.1K citations
Life Sciences, Ecology, Zoology +17
Pose tracking from natural features on mobile phones
Daniel Wägner, Gerhard Reitmayr, Alessandro Mulloni et al. · 2008 · 427 citations