Tag loss is a minor limiting factor in sea turtle tagging programs relying on distant tag returns: the case of Mediterranean loggerhead sea turtles

As in many other species, tagging has been routinely conducted for decades in over a hundred sea turtle capture-mark-recapture (CMR) programs worldwide. Tag loss is a key limiting factor because it violates the main assumption in CMR models; however, very few estimates of tag loss exist, and we provide here a review. No published estimations of tag loss are available for the Mediterranean, in spite of intensive tagging since the 1980s. This study aims to provide an estimation of tag loss in loggerhead turtles tagged in the Mediterranean. We modeled 64 tag returns out of ca. 2200 loggerhead turtles tagged at Mediterranean foraging grounds, with mark-recapture intervals up to 11.5 years, in order to estimate a daily tag loss probability of the most used tag applied to the most common turtle species in the region. Five models were evaluated through maximum likelihood estimation. The model with the best fit described a tag loss initially high and then decreasing to a lower asymptote, which is probably due to some defective tag applications. The resulting tag loss (0.15 in the first year and 0.31 after 5 years) was comparable or even lower than those from other areas and/or species and predictions indicate that double tagging can make a turtle identifiable for a long period. Hence, in our tagging program and probably in similar ones as well, tag loss appears to be the less important of the factors affecting tag returns, and efforts in other directions are more likely to improve CMR results.