96 minimal interval of 75 days after clutch events to make sure the natural emergence of hatchlings had occurred, based on the 51.1-71.2 days incubation period reported by UsateguiMartín et al. (2019). The emergence date was estimated a posteriori, as it is characterized by a rapid and permanent temperature drop, corresponding to the hatchlings' emergence to the beach surface (Neville et al. 1988; Drake and Spotila 2002). A calibration control step was conducted by simultaneously deploying both an iButton and a HOBO in four nests over the full incubation period during the 2021-2022 nesting season. A linear regression was used to compare the time series of temperature from the iButtons and the HOBOs devices (Table 9 SI Appendix). Measurements issued by the two devices were highly consistent (R = 0.99, p-value < 0.001 ***) and the linear regression revealed an intercept of + 0.64°C in the HOBOs devices, which was subsequently corrected by adding 0.64°C to all iButtons values. 3.3 Variability across nesting areas Generalized Additive Mixed Models (GAMMs) were used to compare temperature profiles between the GLS and RP areas, while accounting for within and across season variability, as well as individual nest variability. The average temperature recorded per day per nest was used as the response variable following a Gaussian distribution. The GAMM was fitted with a restricted maximum likelihood method in the mgcv R package (Wood 2011). Additive explanatory terms were composed of an interactive fixed effect between season and nesting area, and of a smooth effect of day of year by season and nesting area (with basis size limited to 5). Nest identification was included as a random effect. The effect of the explanatory variables on the overall model fit was assessed by examining the percent of deviance explained in comparison to a null model (Derville et al. 2023). The average effect of season and nesting area on temperature was assessed through an ANOVA. The difference in smooth relations with day of year by season and nesting area was evaluated with partial dependence plots while excluding random effects. 3.4 Sex ratio assessments Multiple predictive models were used to estimate hatchling sex ratios from temperatures observed in the GLS and RP nests (Jribi and Bradai 2014; Laloë et al. 2023; Staines et al. 2023; Fuentes et al. 2024). In this study, the proportion of female and male hatchlings were estimated for each monitored nest using four different models to provide a comparative perspective. The Chu et al. (2008) model:
RkJQdWJsaXNoZXIy MjE1NDI=