113 Every loggerhead turtle nesting activity observed over the total perimeter of the islets were recorded, along with date, site, GPS location, and its nature based on the available cues. The presence of a body pit, a depression made by a female turtle following oviposition, reveals the presence of a successful nest. Its absence otherwise display the recording of an aborted attempt (SWOT Scientific Advisory Board 2011; Barbier et al. 2023). This dataset allowed us to estimate the nesting success on field observations only. However, as the visual determination of the marks left in the sand by reproductive females can be biased by the age of the nesting activity: the older the activity, the higher the chance to mistake a successful nest with an aborted one, and the other way around. Persistence over time may vary according to weather conditions and to the exposure to wind (Metcalfe et al. 2015). Considering that up and down crawls are the first one to disappear, the nesting success per islet and per season was assessed using a data subset only including activities for which locomotion crawls were clearly distinguished, as they are considered to be fresh enough to limit the observer’s bias. It is given by the equation: = number of successful nests number of successful + aborted nests (1) Then, following the Girondot phenological model (Girondot 2010, 2017), we assessed the predicted seasonal number of nests per islet (‘nest_islet_median’) with a 95% Credible Interval through the modelling of the nesting phenology parameters over multiple seasons (Bourgogne et al. 2024c). A Chi-squared test was run to test the homogeneity of the predicted distribution of the number of nests over the distinct islets and nesting seasons within the GLS. Kruskal-Wallis tests were used to compare the mean predicted number of nests across seasons and islets respectively. Visualization of the spatial distribution of the predicted seasonal number of nests over the GLS area was carried out by season, using a Kernel density estimate, applied through the ‘Heat map’ tool of QGIS (version 3.34.3). Predicted values were extrapolated to islets x season combinations in which no nest count data was effectively collected. In this case, the islet has been highlighted by a black triangle in the heat maps of predicted number of nests. Moreover, 95% confidence interval values representative of the uncertainty surrounding predictions per islet and per season were mapped with a similar approach, and presented in Table 13 SI Appendix.
RkJQdWJsaXNoZXIy MjE1NDI=