98 differences were found between average temperature at both nesting areas (ANOVA test on GAM fixed “nesting season” term, df = 2, F = 0.234, p-value = 0.791). The nests of the GLS were however significantly cooler than the nests of the RP on average (ANOVA test on GAM fixed “nesting area” term, df = 1, F = 6.438, p-value = 0.011*). This temperature difference was evident in all three seasons and all along the incubation period, with the temperature profiles fitted in the GLS being on average 2.66°C (± 0.43) below RP in 2019-2020, 2.56°C (± 0.37) in 2021-2022, and 3.91°C (± 0.92) in 2022-2023 (Figure 21). Across the three considered seasons, the average fitted temperature difference between the areas was 2.97°C (± 0.83). Furthermore, based on the 2021-2022 season dataset (n = 8 monitored nests for both the RP and the GLS areas), we observed a greater dispersion of temperature values around the mean value in the GLS than in the RP area (F test: F =0.547, num df = 421, p-value < 0.001***), indicating greater variance across nests and/or through time in the GLS. Figure 21. Fitted trends of nest temperature at the RP (red) and GLS (blue) nesting areas across three nesting seasons (November-March). Solid (RP) and dashed (GLS) lines represent the mean predicted nest temperature by day. Shaded areas represent approximate 95% confidence intervals. Observed daily temperatures are represented in light shades in the background. All four sex ratio predictive models, based on observed nest temperature and observed incubation duration, suggest that the RP area has produced a highly female-skewed primary sex ratio over the study period, with the most conservative estimate at 97.4% of females (Tableau 7). In the GLS, balanced 1:1 primary sex ratios were predicted by two models (Girondot 1999; Chu et al. 2008), while both Mrosovsky et al. (2002) models indicate a male-
RkJQdWJsaXNoZXIy MjE1NDI=