Male-Biased Primary Sex Ratio of Leatherback Turtles (Dermochelys coriacea) at the Huon Coast, Papua New Guinea
Abstract
This study investigated the sex ratio of leatherback turtle hatchlings along the Huon Coast, Papua New Guinea, from January to March 2007. Results show that this population predominantly produces male hatchlings throughout the peak of the nesting season. The long-term decline of leatherback nesting along the Huon Coast, in conjunction with the highly male-biased sex ratio, presents serious challenges for the management and recovery of the western Pacific leatherback meta-population.
The leatherback turtle (Dermochelys coriacea) nesting population at the Huon Coast, Morobe Province, Papua New Guinea (PNG; Fig. 1) is one of the largest known nesting populations in PNG and possibly the second largest in the western Pacific; only West Papua, Indonesia records more individuals (Hitipeuw et al. 2007). Therefore, it is important to gain insight about the dynamics of sex ratios in order to develop an appropriate management plan for this population.
Citation: Chelonian Conservation and Biology 9, 1; 10.2744/CCB-0763.1
Many reptiles, including all crocodilians, many chelonians, some squamates (Godfrey et al. 2003), and all species of sea turtles exhibit temperature-dependent sex determination (TSD) in which primary sex ratio is influenced by the temperature experienced by eggs during incubation (FAO 1990; Janzen 1994). Even slight, sustained alterations to incubation temperatures during this period may radically change the offspring sex ratio, although short-term fluctuations in temperature seem not to affect it (Rimblot-Baly et al. 1987; Chan and Liew 1995). In sea turtles, TSD results in the production of females at higher incubation temperatures (Godfrey et al. 2003).
The sex of leatherback hatchlings is determined during the second trimester of development, which is called the “thermosensitive period” (Mrosovsky and Pieau 1991). The pivotal temperature, at which a hatchling sex ratio of 1 : 1 is produced, lies between 29.2°C and 30.4°C for leatherback turtles in Malaysia (Chan and Liew 1995). Dutton et al. (1985) report a pivotal temperature range between 28°C and 30.5°C in Suriname, and Rimblot-Baly et al. (1987) report 29.5°C for leatherback turtles in French Guiana. In comparison, a transitional range of temperatures produces a mixed sex ratio (Godfrey et al. 2003). Temporal variations in sand temperatures also need to be taken into consideration because certain sections of the beach may be thermally different over the course of the nesting season (Godfrey and Mrosovsky 1999).
The aim of this study was to evaluate the sex ratio of leatherback hatchlings along the Huon Coast during the 2006–2007 nesting season and correlate this with meteorological changes during the nesting season. The knowledge gained from the data collected on this population will contribute to further understanding of their life history.
Methods
The Huon Coast ( lat 7°S, long 147°E) is located along the southeastern section of the Huon District of the Morobe Province, which occupies a part of the northeastern coast of PNG (Fig. 1). The project site of Busama lies approximately 20 km southeast of Lae City (lat 6°S, long 146°E; Fig. 1). The climate along the Huon Coast can be described as tropical equatorial. It is distinguished by a wet season from April to October with an average precipitation of 441.7 mm per month and a dryer season from November to March with an average precipitation of 297.4 mm per month. The most precipitation occurs during August with 535.4 mm, and the lowest is during February with 236.0 mm. The total annual rainfall amounts to 4581.4 mm (Hoare 2007). The air temperature in the region ranges from an average of 22.0°C in July to 31.1°C in February with an annual average of 26.3°C (Hoare 2007).
We collected data on temperature regimes during the last third of a 6-month peak nesting period from October 2006 to March 2007. The entire nesting period could not be sampled due to logistical reasons (i.e., the temperature data loggers were not available at the beginning of the season). The temperature regime differences, and thus the emerging sex ratios, were determined by deployed temperature data loggers (Log Tag TRIX-8) with a temperature range from −40°C to 85°C and an accuracy of ± 0.5°C, which recorded the surrounding temperature every 30 minutes. We placed 4 loggers inside nests during oviposition so that the loggers were located approximately in the middle of the clutch. Another 6 loggers were placed in habitats that represented a range of nesting sites and positions on the beach at a depth of approximately 70 cm (e.g., Eckert and Eckert 1990). We recorded the serial number of each logger along with its GPS location and the date of insertion. They were later retrieved at the end of the incubation period and at the end of the season to determine the differences in temperature regimes and to estimate the sex ratio of hatchlings. We followed the procedure in Booth and Astill (2001) and assumed that all hatchlings were male at temperatures ≤ 29.2°C, all hatchlings were female at temperatures ≥30.5°C, and that the proportion of females increased linearly between 29.2°C and 30.5°C. These assumptions were made to calculate the sex ratio of the hatchlings.
Results and Discussion
Of the 10 loggers deployed, data from only 6 loggers were used in this analysis because 2 loggers could not be relocated and another 2 failed whilst deployed. A number of variables other than average temperature can influence the sex ratio of sea turtle hatchlings. Short-lived but large temperature variations within the thermosensitive period, differences in clutch size and metabolic warming, as well as the degree of compaction of the sand are examples of these variables (Mrosovsky et al. 1999). However, daily variation in temperature was not taken into account in the analysis of this study because the depth of nests provided an effective thermal buffer (Godfrey et al. 1996; Mrosovsky et al. 1999). Daily fluctuations in temperature averaged 1.1°C, which made virtually no difference to the constant temperature equivalent (Georges et al. 1994). Therefore, an arithmetic mean daily temperature was calculated for each nest and location in the sand by averaging the 30-minute intervals over each 24-hour period. This was done for nests during the entire incubation period and for locations in the sand for the months of February and March.
The fluctuation of nest temperature of 3 nests laid at Busama during the 2006–2007 nesting season was recorded (Fig. 2). Nest 1 and Nest 2 were laid in the beginning of February 2007, and Nest 3 was laid in mid-January 2007. The average temperatures in nests and the predicted sex ratios are summarized in Table 1. The results show that the predicted sex ratio of hatchlings in the 3 monitored nests is highly skewed towards males with only 24.2% ± 35.2% females in Nest 1, 27.6% ± 38.0% females in Nest 2, and 5.4% ± 18.4% females in Nest 3 (Table 1).
Citation: Chelonian Conservation and Biology 9, 1; 10.2744/CCB-0763.1
The fluctuation of sand temperature of 3 different locations in the sand at Busama during the last third of the 2006–2007 nesting season was recorded (Fig. 3). The results show that the predicted sex ratio of hatchlings varies considerably with the location, but when considering the combined values of the 3 locations the sex ratio is relatively balanced between males and females (Table 2).
Citation: Chelonian Conservation and Biology 9, 1; 10.2744/CCB-0763.1
The average air temperature and precipitation in the study area (Hoare 2007) is summarized in Figure 4. It shows that the average air temperature is lowest and the average precipitation is highest in the beginning of the nesting season. Thus, it can be concluded that the percentage of female hatchlings during the beginning of a nesting season is even less than towards the end (i.e., February and March; Table 2). This is also supported by decreasing incubation durations during the nesting season with 61.8 ± 4.2 days in November and 55.8 ± 3.4 days in February (A.S., unpublished data, nesting season 2006/2007). The temporal distribution of nests (n = 171; A.S., unpublished data) across the entire nesting season is taken into account to estimate the overall sex ratio of this nesting season with the highest number of nests in December and January. It is estimated that during the 2006–2007 nesting season the overall sex ratio is highly skewed towards males, with only 7.7% of hatchlings being female.
Citation: Chelonian Conservation and Biology 9, 1; 10.2744/CCB-0763.1
Sex of hatchlings is confirmed through histological investigation of the gonads, which requires sacrificing the hatchlings. To bypass this requirement, estimates of sex ratios can be made based upon incubation temperature. The thermal tolerance range for sea turtle hatchlings in general was estimated to be between 25°C and 35°C (see Ackerman 1997) in comparison to the more conservative estimation of 24°C to 32°C made by Yntema and Mrosovsky (1982). Data from the 6 loggers used in this analysis showed a range of temperatures between 19.3°C and 31.9°C, which generally exceeded the thermal tolerance of leatherback turtle embryos, at least at the lower range.
The potentially high male bias in sex ratio (based on the mean incubation temperature for leatherback turtles) at the Huon Coast may be problematic for the recovery of western Pacific leatherback meta-population (A.S., unpublished data, nesting season 2006/2007) and the Pacific population as a whole. However, additional long-term data are necessary to further substantiate this finding. Previous studies elsewhere around the world have estimated a strong female bias in hatchling production: 53.6% females for leatherback turtles in Suriname (Godfrey et al. 1996) and 100% females for those in Malaysia (Chan and Liew 1995). In Papua, Indonesia, high average sand temperatures suggest a female-biased population (see Tapilatu and Tiwari 2007). However, their recorded mean incubation periods of 61.5 ± 4.7 days and 61 ± 5.1 days (Tapilatu and Tiwari 2007) are similar to the length of incubation recorded in November, which suggests a male-biased sex ratio. In contrast, a more balanced sex ratio approaching 1 : 1 in leatherback turtle hatchlings has been observed in French Guiana (Rimblot-Baly et al. 1987). Thus, the production of predominantly male hatchlings in this study could benefit the declining meta-population in the western Pacific because it may produce significantly more females in most other areas that contribute to the genetic pool of this meta-population.
Other sea turtle species exhibit the same diverse pattern, but it seems that many populations around the world predominantly produce females under natural incubation conditions. Loggerheads (Caretta caretta) produce more than 90% female hatchlings in Florida (Mrosovsky and Provancha 1992), and Standora and Spotila (1985) estimated female-biased sex ratios of hatchlings for Kemp's ridleys (Lepidochelys kempii) in Mexico and green turtles (Chelonia mydas) in Costa Rica, the Seychelles Islands, and Yemen.
The foregoing results and discussion include only data collected during the monitoring period of January to March 2007. Nesting primarily takes place during the “dry” season of the austral summer (i.e., October to March), with the highest numbers of nesting activities occurring during December and January (Pilcher 2006). Nevertheless, nesting does also occur in other areas outside this hypothetical time frame and may be due to local “dry” seasons (Benson et al. 2007). Furthermore, all the beaches are prone to erosion and accretion to some extent. These events occur throughout the year but are more frequent during the wetter months from May to August. The relocation of these threatened nests has to be taken into consideration in the future to further increase the overall hatching success. It is possible that even more males could be produced during these cooler months; July being the coldest month with mean air temperatures of 24.9°C (Hoare 2007). Although the percentage of hatchlings produced during that time would probably be negligible due to low levels of nesting and concerns of ongoing egg harvest outside the monitoring period.
Research area at the Huon Coast, Morobe Province, Papua New Guinea.
Nest temperature profiles of 3 different nests at Busama during the nesting season 2006–2007; shaded area indicates pivotal temperature for leatherback turtles (see text).
Sand temperature profiles of 3 different locations at Busama during the nesting season 2006–2007; shaded area indicates pivotal temperature for leatherback turtles.
Average 24-h air temperature (1949–1990) and precipitation (1945–1990) in the study area during the peak nesting season (see Hoare 2007).
