The individual color pattern on the back of Bufotes viridis balearicus (Boettger, 1880) allows individual photo-identification recognition for population studies

• Published in: Canadian journal of zoology Vol. 102; no. 4; pp. 393 - 402
• Main Authors: Lassnig, Nil, Guasch-Martínez, Sergi, Pinya, Samuel
• Format: Journal Article
• Language: English
• Published: Ottawa Canadian Science Publishing NRC Research Press 01.04.2024
• Subjects: Adaptability; Bufotes viridis; Identification; Identification methods; Image quality; Learning; Performance measurement; Population studies; Quality control; Rejection rate; Software; Standardization; Wildlife
• ISSN: 0008-4301; 1480-3283; 1480-3283; 0008-4301
• DOI: 10.1139/cjz-2023-0019

This study explores the potential of photo-identification methods (PIMs) as a viable, non-invasive, and ethical tool for wildlife studies, with a specific focus on anuran species such as Bufotes viridis balearicus (Boettger, 1880). Although the automatic photo-identification suite (APHIS) software was initially designed for lizard identification, our research shows its adaptability for anuran species, achieving a high detection accuracy rate of 95.28%, thus obtaining outstanding and higher values in compared to previous studies on this species. Crucially, our findings indicate that the success of PIM and the efficacy of image identification software like APHIS is dependent on the quality and standardization of the images collected. The study also underscores the importance of practical experience and continuous learning for the optimal utilization of software like APHIS. Despite occasional false rejected matches, the overall strong performance metrics with low false rejection rate demonstrate that these instances do not significantly impact the reliability of the technique. Thus, this research highlights the importance of careful implementation, continuous learning, and image quality control in leveraging the full potential of image identification software in wildlife studies.