First assay of photo-identification in marine turtles’ nesting population
Photo-ID can be one of the most efficient methods for long-tern monitoring of a population, despite it not being rerror-free. Different indiciduals can present similar mark patterns that can lead us to make errors in an indicidual’s identification (false posi-tive), or multiple indentifications of the same individual (false negative). Low quality of the photos also diminishes the efficiency of this method (…) Using the numeric cord of the scales pro-posed by jean et al. (2010) improve the identification progress with regards the qualitative method used previously (…) photo-ID is still in development and new approaches to improve the keeps coming out, such as a newly developed method for automated recognition of photographs (…)
The photo-ID system that we used has been employed previously by other authors such as jean et al. (2010) and Reisser et al. (2008) but only for underwater identification. Photo-ID proved to be useful for ecological characterization of feeding aggregation (…) and for management decisions such as the limits of a marine Protected Area (…). In all studies the authors noted the advantages of the method and were successful in identifying different individual animals. Similar results were obtained in our study with only one identification error. Nevertheless 37 pictures had to be discarded because the visibility of the scales were compromised due to sand or head retraction. In those cases, identification would be erroneous and would lead to mistakes in the database. Application of photo-ID in nocturnal studies could have also the disadvantage of stressing the animals when are flashed. Nevertheless, this process is always performed after the female finished laying the eggs and, so far, no sign of disturbance had been noticed. In any case, we concur with Reisser et al. (2008) with the fact that the possible disturbance that the flash may cause to nesters needs a follow up since Campbell (1994) in Costa Rica, recorded a reduction in the time spent by females covering the nest when they were photo-graphed.
In this study, working during the night and at a nesting beach presents certain difficulties in detecting facial scale patterns. For one, photos taken during the night are often not clear enough to identify and count all scales.