Emergence Timing of Leatherback Hatchlings (Dermochelys coriacea) At Sandy Point National Wildlife Refuge, 2010–2014
Abstract.
Leatherback sea turtles (Dermochelys coriacea) are listed as Vulnerable globally by the International Union for Conservation of Nature. To preserve this species, we must protect them during several life stages, including protecting hatchlings emerging from nests. We studied the timing of emergence of leatherback hatchlings at Sandy Point National Wildlife Refuge, an area created for the protection of this species. We surveyed 524 nests in 2010–2014 and evaluated the time that hatchlings began emerging from the surface of the sand, the time that the last hatchling in the group finished emerging from the sand, the incubation duration of each nest (number of days between deposition and first emergence), and the number of hatchlings in the initial pulse of emergence. In situ clutches (those occurring naturally and not manipulated by location) had an incubation period of 61.4 ± 21.1 d (mean ± SD; n = 332) and emerged from 1825 to 1906 hrs. The incubation duration varied significantly from relocated clutches (those that had been moved at deposition because of erosion threats), had an incubation period of 60.0 ± 2.7 days (n = 192), and emerged from 1827 to 1903 hrs (on average, for all years combined; p < 0.0001). These nests yielded 36.6 ± 21.1 first-emergence hatchlings from in situ clutches, which was different from relocated clutches (14.9 ± 13.6 first-emergence; p < 0.0001). Other studies have found that hard-shelled sea turtles all begin emerging once the sun sets and continue emerging throughout the night. In contrast, we found that leatherback sea turtle hatchlings emerge earlier in the evening, at least at Sandy Point. During the seasons sampled, emergence often began before sunset (between 1834 and 1859 hrs) and hatchlings usually did not continue to emerge through the night. To allocate resources most effectively, beach management programs at Sandy Point National Wildlife Refuge should be maximized between 1700 and 2000 hrs, the period when most hatchlings are emerging and require the most protection from predators. This work provides important information about the timing of leatherback emergence and may be applicable to other leatherback nesting beaches where predation may be a problem. These findings can help ensure that nest protection programs maintain efficiency and protect hatchlings during their vulnerable crawl to the sea.
Many marine species exhibit some level of parental care, with neonates being protected or fed until they are large enough to fend for themselves. However, investment by parents ranges widely (Kendeigh 1952; Wesolowski 1994) from extreme neonate protection, as seen in the brooding behavior of cichlid fish (Taborsky 1984), to little more than laying eggs and leaving them to incubate in the sand as with sea turtles (Shine 1988; Iverson 1990).
The level of parental care in sea turtles is minimal. In general, female sea turtles emerge from the ocean at night to stay within their thermal tolerance limits (Reina et al. 2002; Law et al. 2010). Individuals then nest on sandy beaches before returning to the water. Therefore, apart from nest site selection and nest-digging abilities, female turtles have little parental care and no role in the development of their young (Wallace et al. 2007). After incubating for a period of 7–13 wks, the hatchlings produced will emerge from the sand and make their way to the ocean on their own (Miller 1997). All species of sea turtles go through roughly the same nesting process, with only minor differences such as relative location on the beach, how deeply eggs are buried, and how much sand covers the nest (Miller 1997).
Sea turtle hatchlings must be prepared upon emerging to find their way to the sea and survive on their own. However, many threats may block their path. If ambient temperatures are too high, leatherback (Dermochelys coriacea) hatchlings could suffer from limited mobility or even heat stroke (Mrosovsky 1968; Drake and Spotila 2002). Predation is also a serious risk (Santidrián Tomillo et al. 2010). However, hatchlings have several strategies for minimizing exposure to both heat and predators. To avoid overheating, hatchlings that are ascending from the egg chamber stop moving upward when sand temperatures are too hot, moving upward for the final push to the surface only once the top layer of sand has cooled (Drake and Spotila 2002). Cooler sand temperatures usually happen near nightfall, when predators are fewer and there is lesser risk of hatchlings overheating on hot sand on their way to the water (Miller et al. 2003). Species differ in the timing of, and optimal temperatures for, emergence. Drake and Spotila (2002) analyzed the critical temperature maxima (CTM) of sea turtle hatchlings at Santa Rosa National Park in Costa Rica to determine how sand temperature acted as a cue for hatchlings to emerge from the nest. Leatherback hatchlings were observed to emerge when the sand dropped below 36°C and olive ridley hatchlings (Lepidochelys olivacea) emerged when the sand dropped below 34°C. These temperatures were most often observed at or near sunset or on rainy days (Drake and Spotila 2002). For green turtle hatchlings (Chelonia mydas) in Suriname, Mrosovsky (1968) found that heat inhibited activity below the sand surface but that hatchlings became active once the sand temperature dipped below 28.5°C. This temperature threshold is usually reached at nightfall (Mrosovsky 1968; Moran et al. 1999).
Nightfall also provides cover for hatchlings as they emerge from the nest. They then crawl quickly to the water and swim rapidly away from shore (Wyneken and Salmon 1992). However, predation of hatchlings on their way to the water may be high in some locations. For example, at Playa Grande, Costa Rica, 11.7% of leatherback hatchlings were preyed on before reaching the ocean (Santidrián Tomillo et al. 2010). Thus, it is important for hatchlings to move quickly to the ocean once they emerge (Santidrián Tomillo et al. 2010). Once newly hatched sea turtles are in the water, yolk stores within emerged hatchlings help power their offshore swim frenzy (Carr 1962; Wyneken and Salmon 1992), ideally sustaining the hatchling until it is able to find food.
Here, we studied leatherback hatchling emergence at Sandy Point National Wildlife Refuge (SPNWR) at the southwestern point of St. Croix, US Virgin Islands. This refuge was established in 1984 for the protection of leatherbacks. Although 2 other federally protected sea turtle species also nest there, hawksbills (Eretmochelys imbricata) and green turtles (Boulon et al. 1996), SPNWR is best known as a leatherback nesting beach (Dutton et al. 2005). While this species is Vulnerable globally (International Union for Conservation of Nature [IUCN] 2013), leatherbacks have been studied at SPNWR since 1977 and conservation efforts, including relocating clutches that are in danger of inundation or erosion, have resulted in an increase in nesting and in the number of individuals recorded at this beach (Dutton et al. 2005; Turtle Expert Working Group 2007), although these increases have been recently lost (Northwest Atlantic Leatherback Working Group 2018). Leatherback hatchlings emerging from nests at Sandy Point mainly face predatory birds such as frigatebirds (Fregata magnificens) during the day and night herons (Nyctanassa violacea) at night. Sharks and other large fish are predators in the shallow waters immediately offshore (Nellis 2000).
The purpose of our study was to study leatherback emergence tendencies to determine when they may be most vulnerable to predators on this beach. We analyzed data collected from 2010 to 2014, including incubation duration, the time that hatchlings were first spotted at the sand surface, and the time that the last hatchling in the group completed emergence and began crawling. We considered 2 main groupings of clutches by comparing the timing and number of first emergence turtles from both relocated and in situ clutches. An improved understanding of the emergence tendencies of leatherback hatchlings will help refine current management practices and objectives of the refuge's conservation plan regarding the timing of nest protection.
METHODS
We studied a 3-km stretch of beach at SPNWR, which is managed by the US Fish and Wildlife Service (Fig. 1). The north beach of SPNWR faces north and west and is on the leeward side of St. Croix. The west beach connects with the north beach and runs west to Sandy Point. The south beach of SPNWR faces south and east and has more vegetation and dried seagrass cover than does the north beach (Garrett et al. 2010; Conrad et al. 2011). The beaches along SPNWR often change seasonally when sand near the point on the west beach is eroded by longshore currents and redeposited to the north. This process may be important for nesting habitat quality because it cleans sediment and removes vegetation seasonally (Conrad et al. 2011). Within the refuge, owing to the seasonal pattern that sets up each year, the west beach is referred to as the erosion zone (Garrett et al. 2010; Conrad et al. 2011) within the Comprehensive Conservation Plan for the refuge (Evans 2010).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 18, 2; 10.2744/CCB-1362.1
A consistent leatherback monitoring program has been conducted at SPNWR since 1981 (Boulon et al. 1996; Dutton et al. 2005). From late March until mid-July each year, the nesting beach is patrolled every 45 min, from 2000 to 0500 hrs or until the last nesting turtle leaves the beach (Dutton et al. 2005; Garrett et al. 2010). When a nesting turtle is observed, the monitoring team records the location of the nest by triangulation to stakes at the vegetation line along with various characteristics of the nesting female, including the presence of flipper tags or passive integrated transponder (PIT) tags as a means of identification.
If the nest location is within the erosion zone on the west beach, the clutch may erode before producing hatchlings. To mitigate this problem, eggs laid in the erosion zone are collected and relocated to a safe region of the beach, usually on the north beach (Boulon et al. 1996; Dutton et al. 2005; Garrett et al. 2010). The width, depth, and general shape of the original nest cavity are measured when the female forms it and these dimensions are recreated in the new location. Clutches may also be relocated if they are laid below the high tide line or are otherwise at risk of not producing hatchlings (e.g., within thick vegetation). Clutches that do not need to be relocated will incubate in their original position and are therefore called in situ clutches.
Since 2009, a genetic tagging research program at SPNWR has targeted leatherback hatchlings (Stewart and Dutton 2011, 2014; Dutton and Stewart 2013). Each evening from mid-June to mid-August, as part of the nest management component of the monitoring program for leatherbacks, a team patrolled SPNWR to look for emerging leatherback hatchlings (Stewart and Dutton 2011, 2014). Nests that were found were monitored and protected from the beginning of emergence until the last turtle left the sand. Hatchlings emerged independently from both relocated and in situ clutches and were not dug out of the nest chamber. We considered the time when we first spotted signs of hatchling emergence (noses or heads protruding from the sand; Fig. 2) as the start time for the emergence and the end was when all hatchlings were fully emerged and on top of the sand surface. The number of hatchlings at first emergence was recorded, and all that emerged were collected and sampled for genetics; they were released with their clutchmates on the south beach of the refuge as soon as possible thereafter (Stewart and Dutton, 2011, 2014).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 18, 2; 10.2744/CCB-1362.1
We compiled leatherback nest and emergence data from 5 nesting seasons (2010–2014). To calculate the incubation period for each nest we subtracted the date that the clutch was laid from the date that hatchlings emerged. The times that emergence began and ended were used to determine the total emergence time for each clutch. As we were interested in the initial emergence date as well as the number of first emerging hatchlings for this study, we focused our efforts on recording variables related to the first emergence only. Nests tend to produce a few additional hatchlings after first emergence on subsequent evenings, but the most vigorous turtles usually belong to the first emergence. In all of the analyses, a distinction was made between in situ and relocated clutches. We used unpaired Student's t-tests to identify differences in variable means between relocated and in situ clutches.
Finally, we analyzed the time of day of first emergence of all clutches (2010–2014) with respect to the date of emergence. We then calculated the leastsquares linear regression and coefficient of determination (R2) to see whether there was a relationship between the time of year and the time of emergence.
RESULTS
We began monitoring nests by mid-June (range = 17–27 June) and ended monitoring in mid-August (range = 8–26 August). In 5 yrs (2010–2014) we recorded the precise emergence times for 332 in situ clutches and 192 relocated clutches. These 524 study clutches had a complete first emergence that was watched from start to finish. As many nests as possible were protected, but not all were included in the analysis. Nests may not have been evaluated for emergence date or number of first emergence hatchlings because they were washed out, their eggs did not hatch, they were located out of the area where we patrolled most often, or hatchlings emerged during daytime rain showers or, occasionally, after we had left the beach for the night.
Numbers of in situ and relocated clutches varied annually (Table 1) depending on the erosion pattern along the west beach. In situ clutches on SPNWR incubated significantly longer than did relocated clutches (1.4 d; t = 3.2; p < 0.0001). The average incubation period for all in situ clutches was 61.4 ± 2.7 d (n = 332; values are mean ± SD throughout) whereas relocated clutches incubated for 60.0 ± 2.7 d (n = 192).
Over the course of the sampling period, the sun began setting between 1834 and 1859 hrs and twilight occurred between 1857 and 1923 hrs (Edwards 2018). On average, we observed the beginning of emergence for in situ clutches from as early as 1804 hrs (in 2014) to as late as 1855 hrs (in 2011; Table 1). The average time at which in situ clutches were completely emerged ranged from 1847 hrs (in 2013) to 1924 hrs (in 2011; Table 1). The peak of in situ emergence across all years occurred between 1700 and 1800 hrs (Fig. 2A). Spotting times (the time that the emergence was first observed) for relocated clutches ranged from 1813 hrs (in 2013) to 1833 hrs (in 2011). The average time at which the relocated clutches finished emerging ranged from 1850 hrs (in 2013) to 1914 hrs (in 2011; Table 1). The peak of relocated clutch emergence occurred at 1730 hrs (Fig. 2B). The average total emergence time during first emergence for all years was under an hour (40 ± 35 min for in situ clutches and 36 ± 37 min for relocated clutches), with the shortest emergence time averaging 26 min (relocated clutches in 2010) and the longest emergence time averaging 57 min (in situ clutches in 2012) (Table 1).
In situ clutches yielded significantly more hatchlings on first emergence than did relocated clutches across all years (t = 13.2; p < 0.001) as well as within years (Table 1). The average number of first emergence hatchlings was greatest in 2013 (43.6 from in situ clutches and 18.9 from relocated clutches) and lowest in 2012 (29.3 from in situ clutches and 10.1 from relocated clutches) (Table 1). Furthermore, for all years combined the number of in situ hatchlings (36.6 ± 21.1) produced was more than double the number of relocated hatchlings (14.9 ± 13.6) produced at first emergence (Table 1).
Finally, the date of hatchling emergence was plotted against the time of emergence for all clutches. Despite daylight hours declining as the season progressed, there was no relationship between time of emergence and time of year (R2 = 0.0342, p > 0.0001; Fig. 3; Edwards 2018).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 18, 2; 10.2744/CCB-1362.1
DISCUSSION
While previous work has connected hatchling emergence with temperature cues (Mrosovsky 1968; Moran et al. 1999; Drake and Spotila 2002), little research has been done on the time of day during which this temperature threshold is met and hatchlings emerge. The aim of this study was to determine the timing of leatherback hatchling emergence at SPNWR to ensure that hatchlings are protected at the time of evening when they are most vulnerable. In other studies, green, hawksbill, and loggerhead (Caretta caretta) turtles have all been observed to emerge throughout the night (between 1800 and 0800 hrs, Glen et al. 2006; between 1700 and 0700 hrs, Salmon and Reising 2014; between 2000 and 0730 hrs, Adam et al. 2007; respectively). Although we did not measure sand temperatures at emergence, we determined that leatherback hatchlings begin emerging in the early evening (average 1804–1855 hrs), even before sunset, and they do not continually emerge for the rest of the night, at least for this beach at St. Croix.
Emergences generally began after 1630 hrs and were completed before 2000 hrs each evening (Fig. 3). These results differ from the time of emergence data collected for hawksbill, green, and loggerhead sea turtles by Salmon and Reising (2014). In that study, hawksbill turtles emerged throughout most of the night, from 1700 to 0400 hrs, peaking between 1800 and 1900 hrs (Reising 2014; Salmon and Reising 2014). Similarly, green turtles from various beaches emerged between 1800 and 0600 hrs (modified from Glen et al. 2006; Gyuris 1993) and loggerhead sea turtles emerged from 1800 hrs until after midnight (Neville et al. 1988; Witherington et al. 1990; Salmon and Reising 2014). Also, Adam et al. (2007) reported that only 3% of loggerhead hatchlings emerged prior to 2000 hrs and Glen et al. (2006) concluded that green turtle nests emerge throughout the night, mostly after sunset and before sunrise, from 1700 to 0700 hrs.
We observed that the majority of leatherback hatchlings at Sandy Point emerge earlier than those of other species of sea turtles, with the earliest hatchlings beginning to emerge at 1600 hrs (i.e., before sunset; Edwards 2018) and the latest at 2230 hrs (Fig. 3). Over the emergence season there was no appreciable change in the time of emergence (Fig. 3). Therefore, the approximate time leatherbacks emerge remains fairly consistent throughout the season, despite the length of daylight varying by 30 min (Edwards 2018). This consistency in emergence time is most likely due to the relative consistency of the time of sunset. If there were to be a significant change in day length throughout the season, we would expect to see this trend mirrored in hatchling emergence time.
Generally, the hatching success of relocated clutches is lower than that of in situ clutches (Eckert and Eckert 1990). Our study supports this trend, as the in situ nests surveyed over the course of this project produced more than twice the hatchlings that relocated clutches did at first emergence (Table 1). This lower hatching success within relocated clutches is most likely due to trauma to the eggs in the transferal process or introduction of microbial pathogens in the reburial process (Eckert and Eckert 1990). Also, while little research has been conducted on the incubation duration of relocated clutches, previous studies have shown that incubation periods vary between in situ and relocated clutches (Mrosovsky and Yntema 1980, Abella et al. 2007). At SPNWR, we observed in situ clutches having a significantly longer incubation period than did relocated clutches (p < 0.0001; Table 1). This project yielded results that agree with previous studies and contributes to the growing base of knowledge regarding leatherback hatching tendencies.
All hatchlings that we collected were released on the south beach of SPNWR, where the water is less clear and has fewer predatory fish than are found off the north beach. Although we may have missed emergences after leaving the beach (~25 per season), we are confident that while we were on the beach, we observed the majority of emergences during the time the nest protection program was in place. The time of emergence was recorded as the time when emerging hatchlings were spotted, as this is when the nests should begin being protected. However, beach patrols to locate emerging hatchlings should plan to begin prior to this time in order to ensure that all emergence activity is monitored.
We found that the majority of leatherback hatchlings begin emerging well before dark and that hatchlings are completing emergence and beginning their sea-finding path around dusk each evening. Future studies could measure how the time of emergence and incubation duration affects fitness in hatchlings (crawl speed, swim speed, etc.). The new information presented here should allow beach management at SPNWR to better anticipate when leatherback hatchlings will emerge, allowing managers to protect them from both predators and the elements. Furthermore, our methods should prove useful for other leatherback nesting programs so that managers will be able to estimate when most hatchlings will emerge and need the most protection. To adequately protect the leatherback sea turtle, effective beach management programs must be able to predict when and where hatchings will emerge from their nests so that they can best protect them.
Map of SPNWR with respect to the greater Caribbean. This 3-km section of beach is managed by the USFWS and is located in the southwest corner of St. Croix, US Virgin Islands (inset). The north beach of SPNWR faces north and west and is on the leeward side of the island. The west beach connects with the north beach and runs west to Sandy Point. The south beach of SPNWR faces south and east and has more vegetation and dried seagrass cover than does the north beach.
Timing of emergence from in situ (A) and relocated (B) clutches on SPNWR between 2010 and 2014. The dashed line at ~1850 hrs represents the average sunset time during this study. These results reflect that the peak of hatchling emergences occurred before sunset for both in situ and relocated nests.
Relationship between date of emergence and timing of emergence at SPNWR between 2010 and 2014, showing the leastsquares linear regression relationship and its coefficient of determination. These results reflect that although the time of the sunsets will vary throughout the year, there is no significant difference in the time hatchlings begin emerging throughout the season.
Contributor Notes
Handling Editor: Jeffrey A. Seminoff
