Overcoming field monitoring restraints in estimating marine turtle internesting period by modelling individual nesting behaviour using capture-mark-recapture data

Abstract

Marine turtles are intra-seasonal iteroparous animals; they nest from one to up to 14 times during the nesting season, laying up to 180 eggs each time. Their annual reproductive effort can therefore be estimated from clutch size, nesting frequency, and length of the nesting season. Moreover, the estimation of nesting frequency, usually obtained from the internesting period (i.e., the time in days between two nesting events) is essential for assessing the number of females in a population. However, the internesting period is strongly influenced by variation in individual behaviour of the nesting female, including abortion of nesting attempts. It is also affected by imprecise detection of females during beach monitoring, often related with a lack of fidelity to the nesting beach. Using an individual-focused model based on capture-mark-recapture data we were able to statistically characterize the nesting behaviour of the populations of green turtles (Chelonia mydas) and olive ridley turtles (Lepidochelys olivacea) in São Tomé and Príncipe (Eastern Atlantic). The developed model proposes a novel approach in estimating the internesting period, by including the different factors that lead to the heterogeneity observed in the duration of internesting periods across a single season, corrected for the probability of a female aborting a nesting process. The calculated lengths of the internesting periods for the two species are congruent with previous estimates, validating the model. Furthermore, the inference of the rank of a nest for an individual female is predicted by the model with high accuracy, even when the recapture rate is low and the time between observations is long. A limitation of the model is its inability to estimate the true clutch frequency at the scale of the population but it was not its purpose.