Individual recognition performances of photo-identification and pelvic belt marking methods in a camouflaged and diurnal Neotropical frog

Field studies requiring individual identification of amphibians have led to the development of new marking techniques, particularly noninvasive ones. Here, we evaluated the application of two noninvasive techniques, pelvic belt and photographic identification of natural marks (photoID), for individual identification of Hylodes nasus, a territorial, diurnal, and camouflaged species. Data were collected from a population over one year, in Rio de Janeiro, Brazil. A total of 135 individuals were marked with belts. Of these, 41 individuals were recaptured at least once, resulting in 202 encounters and 606 photographs (ventral, lateral, and dorsal views). PhotoID comparisons revealed significant differences across the views (p < 0.05), with the best performance for the ventral view. False rejection rates varied from 0.08 to 0.39 and 0.12 to 0.20 for photoID and belt, respectively. We found differences between these rates of both methods only for the dorsal view (p < 0.05), in which false rejection rates were 0.08 and 0.20 for photoID and belt, respectively. PhotoID was particularly efficient in identifying H. nasus individuals using the ventral region, which may have been less affected by evolutionary pressures imposed by predation risk. Therefore, we recommend the use of both methods combined for accurate data collection.