Climate data
Thermal profiles showed considerable diel variation at each field location (Fig. 3). Soil temperatures on the ground surface fluctuated the most, reaching maximum temperature (Mt Ginini, 45.5 °C; Piccadilly Circus first season, 46.5 °C; Piccadilly Circus second season, 46.0 °C; Cooma, 42.1 °C; Dartmouth, 47.5 °C) during the day (1330–1730 h) and dropping to low level (Mt Ginini, 4.0 °C; Piccadilly Circus first season, 8.5 °C; Piccadilly Circus second season, 6.5 °C; Cooma, 7 °C; Dartmouth 12.5 °C) at night (2330–0400 h) (ANOVA, F4, 6521 = 279.4, P < 0.0001). The soil temperature at 20 cm, the deepest we monitored, showed the least fluctuation at all locations (see Fig. 3). Nests intermediate in depth between these two extremes (surface to 20 cm) showed intermediate diel fluctuations. At all the field locations, lower mean and lower minimum temperatures and less diel thermal variation occurred at 20 cm depth than the soil surface. Monthly mean air temperatures were averaged over the skink active months (i.e. early November to late February, a 16 week period) in each year to reveal a warming trend between 1889 to 2019 (Figure S2-Mt Ginini: F1,517 = 523.9, P < 0.0001, R2 = 0.50; Piccadilly Circus: F1,517 = 539.6, P < 0.0001, R2 = 0.51; Cooma: F1,517 = 537.9, P < 0.0001, R2 = 0.50; Dartmouth: F1,517 = 627.5, P < 0.0001, R2 = 0.54). Air temperatures were consistently cooler at higher elevational locations than the lower elevations, which inversely correlated with elevation when B. duperreyi eggs were incubation (F1,22 = 4.795, P < 0.039, R2 = 0.17). The weekly mean Tmax and Tmin showed that temperatures fluctuated substantially during the 9-week eggs incubation period (Fig. 4). The highest mean rainfall events (17.0 ± 23.87 mm) were recorded at the Piccadilly Circus in the first week of egg incubation period in the first season. The highest mean rainfall events (5 ± 11.85 mm) were recorded at Mt Ginini during the third week of B. duperreyi egg incubation. The highest rainfalls were recorded during the fourth week of the egg incubation period at Cooma (5 ± 6.3 mm) and Dartmouth (1 ± 0.99 mm). Rainfall among the four sample locations were statistically different (Kruskal–Wallis; H = 16.83, P < 0.05).
The diurnal variation of hourly soil temperature at the Picadilly circus field station. LOWESS curve fitted (solid lines) with a smooth curve to aid visual interpretation, (red): soil surface, green: at the depth of 10 cm and blue: at depth of 20 cm (blue). The minimum amplitude of the soil temperature variation approached in 20 cm depth. The daily variation of soil temperature showed a sinusoidal pattern, and the soil temperature decreased with the increase of the soil depth at all field stations. The order of the measured soil temperatures from high to low is: T0 cm > T10 cm > T20 cm.
Weekly mean temperatures in the core of Bassiana duperreyi nests in each field locations (green) and the mean nest depth in each field location (right bottom). Red denotes Tmax and blue denotes Tmin.
Timing of breeding season and nest search protocols
In Piccadilly Circus during first season, female B. duperreyi laid their first eggs in the first week of December and 39 nests were found (only 35 monitored). In the second season (2018/19) females laid their eggs in early January and 26 nests were found (only 14 monitored). In the second season, in early January we found 11 nests at Mt Ginini (only nine were monitored) and nine nests at Cooma (only eight monitored). In the second week of December four nests were found (all monitored) at Dartmouth. Overall, we observed 1335 eggs in 89 nests at the four locations (Figure S3). A total of 84 nests survived the natural incubation period; five nests (3 from Piccadilly Circus and 2 from Mt Ginini) loss their eggs for unknown reasons.
Nest characteristics and nest temperature
Nests were typically constructed beneath rocks (98%), though some were found associated with logs (2%). All nests were deposited in open grassland and received direct sunlight at the surface for a large proportion of each day. Once the rock was removed, nests were typically partially buried in the soil (a few eggs were visible without disturbance) (91%) with a vertical nest chamber; some nests were completely buried (eggs well covered by soil) (7%) and few were found on top of the soil (not buried) (2%). Lengths of the rocks used for nesting averaged 28.8 ± 9.5 cm for Mt Ginini, 32.8 ± 20.7 cm for Piccadilly Circus, 38.6 ± 16.8 cm for Cooma, and 32.00 ± 9 cm for Dartmouth. Rock length differed significantly with location (F3,87 = 5.71, P < 0.0001), decreasing with elevation (F1,87 = 11.94, P < 0.001, R2 = 0.12). Nests were shallower in the first season compared to the second season in Piccadilly Circus (t = 3.7, df = 61, P < 0.05). At Piccadilly Circus, females constructed progressively deeper nests over past 22 years when combining our data (2017 to 2019) an those of Telemeco et al.30 (1997 to 2007) (F1,8 = 11.36, P < 0.05, R2 = 0.58; Fig. 5). Of the 89 nests we located, 69 (71.91%) nests were communal i.e., contained more than nine eggs which according to Radder and Shine36 is the maximum clutch size, Radder and Shine 2007). Nest depth varied from 1 to 85 mm. The deepest nests were recorded at the lowest elevational location of Dartmouth (79 ± 6.1 mm). The shallowest nests (i.e., 15 mm) were found at the highest elevational of Mt Ginini (28.8 ± 11.7 mm) (Fig. 4). Nest depth was inversely related to elevation (F1,86 = 39.80, P < 0.0001, R2 = 0.32; Figure S4).
Long-term variation of mean nest depth of Bassiana duperreyi. The data from 1997–1998 through 2006– 2007 and 2005–2006 seasons nest depth data (blue) (Telemeco et al.30) and current study (green) (2017/18 and 2018/2019).
The lowest (19.67 ± 0.63 °C) and the highest (25.59 ± 1.25 °C) mean nest temperatures were recorded at Mt Ginini and Dartmouth, respectively. During nest incubation at Piccadilly Circus, mean nest temperature was lower (21.6 ± 1.41 °C) in the first season than in the second (22.2 ± 0.22 °C) (F8, 410 = 88.59, df = 8, P < 0.0001). The highest nest temperature (47.5 °C) and the highest mean daily range temperature (19.1 ± 2.30 °C) were recorded at the highest elevational location, Mt Ginini.
Most nests experienced high mean temperatures and a considerable diel range of temperatures in all field locations. Mean daily temperatures experienced by the eggs differed among nests in highest elevation, Mt Ginini (18.76–20.57 °C) to lowest elevation Dartmouth (23.89–27.39 °C), as did mean maxima in Mt Ginini (35.8–27.8 °C) and minima (10.6–13.5 °C), and mean maxima in Dartmouth (34.67–38.67 °C) and minima (15.25–17.75 °C). The nests showed significant differences in mean nest temperature maxima in all locations except Dartmouth, but mean nest temperature minima shows a significant difference in all locations (Table 1).
Mean weekly nest temperature was correlated with mean weekly air Tmax (R2 = 0.42–0.74, P < 0.05) and Tmin (R2 = 0.70–0.93, P < 0001) at all field locations. The significant warming trend was recorded during the incubation weeks in first season at the Piccadilly Circus (F1,300 = 31.74, P < 0.001, R2 = 0.09), but a significant cooling trend was recorded in the second season (F1,124 = 52.63, P < 0.001, R2 = 0.29). In the second season a significant cooling trend was also recorded at Mt Ginini (F1,79 = 132.3, P < 0.001, R2 = 0.62). Whereas at Cooma and Dartmouth showed no significant trend as the season progressed.
Mean nest temperatures during the incubation period were inversely correlated with elevation (F1,2 = 41.71, P < 0.05, R2 = 0.95) (Fig. 6a). The highest and lowest mean daily CTE were recorded at Dartmouth (30.35 ± 0.12 °C) and Mt Ginini (26.2 ± 1.98 °C), respectively. The mean daily CTE was significantly inversely correlated with elevation (F1,74 = 11.39, P < 0.001, R2 = 0.17). When the CTE dropped below the 20 °C (the threshold for reversal, Shine et al.15) during the thermosensitive period, for even for a short time during the incubation period, sex reversal was observed (Fig. 7 and Figure S1 B-F).
The frequency of sex reversal in nests Bassiana duperreyi. (a) Linear regression of mean nest temperature in each field location (F1,2 = 41.71, P = 0.023; R2 = 0.95). (b). The trend in sex reversal frequency of B. duperreyi with the elevation (F1,3 = 41.71, P < 0.05; R2 = 0.95). The number indicates field locations, as indicated in Fig. 1. Grey circles indicate first season data (2017/18) and black circles indicate second season data (2018/19). Broken lines denote the 95% confidence interval, and significance was assumed if P < 0.05.
Temperature trace for the core of a nest of Bassiana duperreyi showing traces for the mean and the constant-temperature equivalent (CTE) for the Dallwitz-Higgins model34,35. (a). The nest produced only XYmale and XXfemale offspring, nest location: Dartmouth. (b). The nest produced XYmale, XXmale and XXfemale offspring, nest location: Mt Ginini. Shaded area: expecting to sex reversal happening during the incubation period. The threshold for sex determination (20 °C; Shine et al.15) and the thermosensitive period lies between lower and upper limit of development. Note that the thermosensitive period does not correspond to the middle third of incubation, either in position or duration, owing to the nonstationary trend in temperatures with season.
Nest depth was not significantly associated with mean nest temperature (F1,84 = 0.081, P = 0.77) at any location. Examination of mean nest temperatures suggested a slight warming trend at Piccadilly Circus during last 22 years, but this was not statistically significant (F1,8 = 3.49, P = 0.09). Nests that produced sex-reversed hatchlings had mean daily nest temperatures cooler than nests that did not produce sex reversed hatchlings. However, mean weekly temperature regimes for these two categories of nests were not significantly different at any field location (Table S1).
Genotypic sex identification and the frequency of sex reversal
During the study period, we collected a total of 415 eggs from the natural nests. The hatching success rate was 95.2%. Therefore, a total of 395 hatchlings were able to be phenotypically identified. Of them, 262 (66.3%) were phenotypic males and 133 (33.6%) were phenotypic females. The sex ratio (Phenotypic male: female) of each location is as follows; Mt Ginini (2018/19) = 2.3 : 1; Picadilly Circus (2017/18 and 2018/19) = 2.27 : 1; Cooma (2018/19) = 1.38 : 1 and Dartmouth (2018/19) = 1.14 : 1, yielding a male-biased sex ratio in the high elevation sites, where adult sex reversal has been previously identified by Dissanayake et al.21 (chi-squared test with Yates correction, χ2 = 4.05, df = 1, P < 0.05). A total of 59 (7.03%) phenotypically male hatchlings were sex-reversed. The highest frequency of sex reversal in hatchlings was recorded at the highest elevation site (28.6%, Mt Ginini, 1,640 m a.s.l.), and zero sex reversal was observed at the lowest elevation (Dartmouth, 380 m a.s.l) (Fig. 1a). This observation concorded with the previous study has been contacted for adult individuals (see Dissanayake et al.21). The frequency of sex reversal was positively correlated with elevation (F1,3 = 41.71, P < 0.05; R2 = 0.95) (Fig. 6b) and each population showed a negative correlation with mean nest temperature (Pearson’s correlation coefficient r = − 0.95, P < 0.05); Tmax (R2 = 0.99, P < 0.05) was significantly negatively correlated with sex reversal frequency (Table S2). When we compared the current study with Dissanayake et al.21 for adult sex reversal frequency, the frequency of sex reversal in the nest was higher than the sex-reversed adult in the same field locations, but not significantly (P = 0.36). However, both hatchlings and adults (Dissanayake et al.21) rates of sex reversal frequency positively correlate with their respective elevation (F1,2 = 15.17, P < 0.005; R2 = 0.77) (Figure S5) (see also Dissanayake et al.21).
Source: Ecology - nature.com
