Female sea turtles lay eggs on tropical and subtropical beaches around the world (Miller 1997). After 50–70 days of incubation (Ackerman 1997), hatchlings emerge from the nest and move rapidly to the ocean (Carr and Ogren 1960). Most hatchlings emerge at night because emergence from the nest is inhibited by high temperature (Mrosovsky 1968; Drake and Spotila 2002). Hatchlings typically emerge in several groups (Carr and Hirth 1961; Bustard 1967) and after emergence, crawl rapidly to the water displaying a frenzy motile behavior that allows them to reduce the time exposed to predators (Dial 1987). Once in the water, hatchlings also exhibit a frenzy swimming behavior that allows them to leave the beach rapidly (Carr 1962). The swimming frenzy lasts for approximately 24 hours in green turtles (Chelonia mydas) and loggerhead turtles (Caretta caretta; Wyneken and Salmon 1992) and may constitute a mechanism to reduce inshore predation (Salmon and Wyneken 1987).

Predation on hatchling sea turtles is assumed high, but the impact of predators on emerging hatchlings has not been documented extensively. Likewise, predator species and predation rates vary among species and geographical locations. Among the predator species reported to eat sea turtle hatchlings are ghost crabs (Fowler 1979; Stancyk 1979; Limpus et al. 1983), foxes, dingos, and rats (Limpus 1971), dogs, raccoons, and pigs (Leslie et al. 1996), night herons (Limpus et al. 1983), storks (Whiting and Guinea 1999), varanid lizards (Blamires et al. 2003), and crocodiles (Sutherland and Sutherland 2003).

Synchronous hatching and emergence in groups by some reptiles dilutes predation risk (Spencer et al. 2001). The lowered probability of predation of an individual when it is part of a group is known as the dilution effect and has been extensively described in ecology in different taxa (Foster and Treherne 1981; Dehn 1990; Wrona and Dixon 1991). On the other hand, grouping also increases the probability of detection by predators (Dehn 1990). For instance, Steward and Wyneken (2004) observed that predatory fish were more abundant in waters near beach hatcheries where a high number of hatchlings were released than at natural locations. Hatchling sea turtles may have efficiently evolved mechanisms to reduce predation on the beach by 1) hatching and emerging synchronously in groups, which dilutes the predation risk, 2) emerging from the nest at night, and 3) rapidly crawling down the beach.

Hatchlings must efficiently locate the water to leave the beach rapidly. If hatchlings encounter difficulties during “seafinding”, they extend the time on the beach and the probability of predation increases. For example, artificial lighting from development behind or near nesting beaches has a negative effect on hatchling orientation (Witherington and Bjorndal 1991). Hatchlings become disoriented by lights and often die from dehydration or are depredated (Peters and Verhoeven 1994; Lorne and Salmon 2007). Therefore, orientation to the sea plays an important role during the crawl to the water because it affects exposure time to predators.

The focus of this study was to characterize predation on the beach and to describe the process of departure from the nest to when hatchling leatherback turtles (Dermochelys coriacea) reach the water at Playa Grande, Costa Rica. Quantifying predation and understanding threats to hatchlings on their way to the water will help in developing stage-based demographic models for leatherbacks in the Eastern Pacific Basin.

METHODS

Playa Grande is the main nesting beach at Parque Nacional Marino Las Baulas (10°20′N, 85°51′W) on the Pacific coast of Costa Rica. The beach is 3.6 km long and extends from Playa Ventanas to the north, which is a dark and small beach suitable for nesting, to the very developed town of Tamarindo to the south, which was once a nesting beach but where no nesting has occurred in the last 20 years. We conducted the study during the 2004–2005 and 2005–2006 nesting seasons. We also made observations of predation during morning surveys in 2005–2006. We divided the beach into 50-m sections with painted posts and marked the location of nests by inserting thermocouples. We measured distance from each nest to the closest marker to the north and to the south. We placed the thermocouple into the nest as turtles laid eggs to record temperatures during incubation. When nests were expected to hatch, we checked them daily at sunrise, before sunset, and several times during the night.

We used a night-vision scope (SU-87/PVS-4 80063) in order to see hatchlings without a light. Nests that were expected to hatch were monitored during the night starting shortly after sunset. Nocturnal observations were made by one observer located 15–20 m away from the nest to avoid interaction with predators. We only observed hatchlings in the first emergent group from each nest, which consisted on average of 87% (2004–2005) and 78% (2005–2006) of the total number of hatchlings that emerged from each nest. We recorded observations on predation and times of activity. Finally, we counted and measured the tracks coming from the nest after hatchlings left the beach.

Duration, Distance Covered, and Rate of Movement

We identified the following stages of activity during the process of departure from the nest and recorded the times at which they occurred: 1) started when sand depression was present over the nest and ended when hatchlings were first visible, 2) began when hatchlings were first visible and ended when hatchlings first began crawling, and 3) started when hatchlings began crawling and ended when they reached the water. Once all hatchlings had left, we measured the distance from the nest to the water line. Because of the difficulty of following several hatchlings individually, we calculated the rate of movement for the first hatchling to reach the water from each nest by dividing the distance from the nest to the water by the time elapsed.

Predation

We recorded every time a hatchling was depredated, the time at which it happened, and the predator. Once all hatchlings were gone, we checked for unrecorded predation. We recorded instances of potential mortality when hatchlings were found upside down and could not right themselves and when they were trapped in vegetation or by other obstacles. In addition to night observations, we recorded predation events during the day when they were observed during the beach surveys.

Patterns of Movement

After all hatchlings entered the ocean we measured the straightness of their tracks and dispersion. We measured tracks during morning surveys when nests that hatched the previous night had been missed by the night observer. Straightness indices of tracks were calculated at 2 locations: 1) within the first meter of the emergence location, and 2) within the last meter of the high-tide line. We measured a meter in a straight line from the nest toward the ocean and then made 2 parallel lines to the shore, first through the center of the emergence location (nest) and second, 1 m away from the nest (Fig. 1). We measured the total distance covered by each of 5 hatchlings between the 2 lines. Then, we calculated the straightness index of each track by dividing 1 m by the actual distance covered by each hatchling between the 2 lines. We only measured 5 tracks per nest because of the difficulty in discriminating tracks near the nest. We estimated straightness indices for the 2 outermost tracks and 3 centric tracks that were distinguishable and covered the distance between the 2 lines. Finally, we calculated the straightness index per nest at the emergence location by averaging the straightness index of the 5 tracks measured. We took the same measurements between the high-tide line and 1 m away inland (Fig. 1), estimated straightness indices of 5 tracks, and calculated the mean straightness index at the entering location.

View Figure

• Download as Powerpoint
• Download Figure

We measured hatchling dispersion (m) by the nest and by the high-tide line. To measure dispersion, we used the lines parallel to the shore 1 m away from each location and measured the distance between the outermost tracks along each line (Fig. 1).

We used SPSS to conduct all statistical analysis. Distributions of variables were tested for normality using the Kolmogorov-Smirnov test. We used t-test to compare differences between the 2 nesting seasons and the Mann-Whitney U-test when assumptions of normality were not met. We used linear regression to quantify effects of predation and on movement patterns. Straightness indices were arcsin-transformed prior to analysis to correct for nonnormality.

RESULTS

Duration, Distance Covered, and Rate of Movement

Leatherback hatchlings took a mean of 34 minutes to reach the ocean after they were first visible at the nest; they were visible on top of the nest a mean of 20 minutes before they started crawling toward the water. Hatchlings spent a longer time visible in 2004–2005 (range  =  19–73 minutes) than in 2005–2006 (range  =  8–51 minutes), but the difference was not significant (independent t-test: t₂₁  =  −1.966; p  =  0.063). Likewise, the time elapsed in each stage of activity was always longer in 2004–2005 (Table 1) but in no case significant. The distance covered by the hatchlings on their way to the water ranged from 16 to 107 m (mean ± SD  =  46.8 ± 30.2 m). The rate of movement from the nest to the water ranged between 1.80 m/min and 6.00 m/min (mean ± SD  =  3.11 ± 1.29 m/min).

Table 1 Time elapsed in stages of activity of leatherback turtle hatchlings on Playa Grande, Costa Rica, during the process of departure from the nest to the water. The stages of activity were: 1) started when sand depression first noticed and ended when hatchlings were first visible, 2) started when hatchlings were first visible and ended when they started crawling, and 3) started when hatchling started crawling and ended when they first reached the water. N is the mean number of hatchlings seen per nest.

View Fullscreen

Predation

On average, 83% of all hatchlings seen after emergence reached the ocean, 12% were depredated, and 5% were counted as potential mortalities because the hatchlings were trapped in vegetation or debris, or were turned upside down and were incapable of righting themselves (Table 2). There were also hatchlings that were attacked by predators and survived, but we could not determine the extent of the injuries. Yellow-crowned night herons (Nycticorax violaceus) often took hatchlings that were moving and dropped them to take other hatchlings. In 2 instances, hatchlings escaped from the predator and reached the water, but we could not check for injuries. Overall, 74% of nests had hatchlings depredated in the first emergent group. The number of hatchlings depredated per nest in 2004–2005 and 2005–2006 was (mean ± SD) 2.5 ± 2.9 and 1.4 ± 1.1, respectively, and the number of hatchlings per nest that reached the water was 11.4 ± 12.5 and 19.3 ± 17.8, respectively. There were no significant differences between 2004–2005 and 2005–2006 in the number of hatchlings that reached the water (Mann-Whitney U-test: U  =  164.5, n₁  =  23, n₂  =  20, p  =  0.11), were depredated (U  =  198.5, n₁  =  23, n₂  =  20, p  =  0.43), or were determined as potential mortalities (U  =  212, n₁  =  23, n₂  =  20, p  =  0.61).

Table 2 Number and percentage of leatherback hatchlings at Playa Grande, Costa Rica, that were depredated, were determined as potential mortalities, and reached the water in 2004–2005 and 2005–2006.

View Fullscreen

The most common predator of hatchling leatherbacks was the ghost crab (Ocypode occidentalis; Table 3). Nocturnal bird predators included yellow-crowned night herons and great blue herons (Ardea herodias). Crested caracaras (Caracara plancus) ate hatchlings early in the morning near sunrise. Additionally, there were unidentified diurnal predators that fed on hatchlings when these emerged late in the day before sunset. When this occurred, we found hatchlings partially eaten. We were able to estimate the number of hatchlings killed, but we could not identify the predator. Ghost crabs took 1–3 hatchlings per nest when they were present and night birds and diurnal predators took 1–9 hatchlings. Number of hatchlings eaten by an individual bird ranged between 1–7 hatchings for great blue herons and 1–2 hatchlings for yellow-crowned night herons. Yellow-crowned night herons attacked as many as 5 hatchlings from the same nest but never ate more than two. We could not differentiate individual ghost crabs. Ghost crabs, night birds, and diurnal predators accounted for 48.3%, 26.4%, and 25.3% of depredated hatchlings, respectively. However, crabs depredated a much higher number of nests (27) than night birds (8) and diurnal predators (8). Thus, night birds and diurnal predators depredated fewer nests but had a larger impact per nest than ghost crabs. The number of hatchlings depredated by crabs increased with the total number of hatchlings seen during the crawl to the water (R²  =  0.11, p < 0.05). There were no significant differences between seasons in the number of hatchlings depredated by crabs (Mann-Whitney U-test: U  =  181.5, n₁  =  23, n₂  =  20, p  =  0.21) and by night birds (U  =  189.5, n₁  =  23, n₂  =  20, p  =  0.15). No mammals were seen to depredate hatchlings during the night observations.

Table 3 Predation of leatherback turtle hatchlings on Playa Grande, Costa Rica, in 2004–2005 and 2005–2006.

View Fullscreen

The number of hatchlings seen that reached the water (linear regression: R²  =  0.11, p < 0.05) decreased toward the south end of the beach. However, the total number of hatchlings depredated was not correlated to the location of the nest on the beach (Spearman correlation: r_s  =  0.007, p  =  0.965), with the exception of the number of hatchlings depredated by yellow-crowned night herons; for this species, predation was significantly higher toward the south end of the beach (linear regression: R²  =  0.13, p < 0.05).

We found 250 nests that had emerged the previous night during 49 morning surveys in 2005–2006. There was at least 1 hatchling still on the beach from 33 nests (13%). Crested caracaras flew along the beach between 0500 hours and 0600 hours nearly every morning, although predation could not always be verified if we were too far from the nest. Frigatebirds (Fregata magnificens) also took hatchlings from the water on 2 occasions. In addition to natural predators, domestic animals frequently disturbed nests on the beach. Nineteen nests were excavated by dogs (7.6%) after hatchlings emerged or were about to emerge, and 2 domestic cats took at least 1 hatchling from a nest located close to one of the beach entrances. Fifty-six hatchlings from 4 different nests desiccated in 2005–2006 due to daytime emergence.

Patterns of Movement

Tracks were significantly straighter within the last meter of departure (by the high-tide line) than within the first meter (by the nest; paired t-test: t₃₇  =  −6.631, p < 0.01). The straightness index by the nest was (mean ± SD) 0.71 ± 0.19 and by the high-tide line was 0.90 ± 0.09. Therefore, hatchling tracks became straighter and less variable as they neared the water.

There was a significant positive relationship between the number of tracks and dispersion, both by the nest (linear regression: R²  =  0.54, p < 0.001) and by the high-tide line (linear regression: R²  =  0.35, p < 0.01). Therefore, dispersion was greater when there was a large number of hatchlings coming from the nest (Fig. 2). However, the number of tracks was not significantly related to the straightness indices either by the nest (R²  =  0.16, p  =  0.085) or by the high-tide line (R²  =  0.012, p  =  0.598). Additionally, the straightness index was negatively correlated with the dispersion by the nest (Spearman correlation: r_s  =  −0.681, p < 0.0001) and by the high-tide line (r_s  =  −0.475, p < 0.005).

View Figure

• Download as Powerpoint
• Download Figure

Finally, the straightness index of nests varied along the beach. Tracks were straighter toward the south than toward the north both by the nest (linear regression: R²  =  0.14, p < 0.05) and by the high-tide line (linear regression: R²  =  0.14, p < 0.05; Fig. 3). Dispersion by the high-tide line also was related to the location on the beach, being greater toward the north (linear regression: R²  =  0.14, p < 0.05). However, dispersion by the nest was not significantly affected by location on the beach (Fig. 3).

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

Predation rates of hatchling leatherback turtles on the beach were estimated and predator species were identified for the first time at Playa Grande, Costa Rica. Mortality rates of early life stages are important traits in life history. Their quantification will contribute to improve demographic models of the leatherback turtle population that nest at Playa Grande.

Hatchling leatherback turtles spent on average 34 minutes on the beach after emerging from the nest, which constitutes a very small percentage of the life of a turtle. However, 12% of the hatchlings die on the way to the ocean. Therefore, adaptations to reduce the probability of mortality on the beach are highly valuable because of the high risk of being depredated on the crawl to the water.

Because mortality on the beach is high, hatchlings must reduce the time exposed to predation. We found that leatherback hatchlings at Playa Grande spent about 20 minutes visible on the beach before they moved to the water. Motionless hatchlings could become easy targets to predators. Great blue herons and yellow-crowned night herons are visual predators with good nocturnal vision (Rojas et al. 1999). However, detecting hatchlings on top of the nest at dark may be difficult for any visual predators if hatchlings are not moving. In fact, we found very few predation events before hatchlings moved and these occurred during the day, with the exception of 1 nest that was depredated by a night heron at night. Consequently, once hatchlings move, they must do it rapidly because the probability of detection and the predation risk increase.

Leatherback turtles emerge from the nest and crawl to the ocean in groups diluting the predation risk. The dilution effect in groups reduces the per capita predation risk on species such as marine insects (Foster and Treherne 1981), mormon crickets (Sword et al. 2005), ungulates (Kie 1999), and aphids (Turchin and Kareiva 1989). On the other hand, the detection probability increases with group size. Birds at night frequently fly and rest on the beach. A bird flying will more easily detect a large group of hatchlings running to the ocean than a small group. Likewise, a ghost crab that detects prey by noise and vibration (Horch 1971), will more easily sense a large group of hatchlings. Our results show a positive relationship between the total number of hatchlings coming from a nest and the number depredated by crabs, although the relationship was not significant for other predators.

The time spent on the beach and the distance covered by hatchlings was highly variable. The distance between high and low tides at Playa Grande ranges from 50 to > 100 m (Reina et al. 2002). Hatchlings that emerged around the time of high tide benefited from a short distance to the water and spent shorter times on the beach. On the other hand, hatchlings that emerged when the tide was low spent longer time on the beach and covered a greater distance. However, hatchlings emerging at low tide could potentially benefit from a receding tide that would take them away from the beach faster and reduce inshore predation. We made observations on nests that emerged both at high and low tides. However, because we did not monitor nests over the 12 hours of darkness every night, we could not accurately characterize emergence patterns at Playa Grande.

Leatherback hatchlings at Playa Grande had slower rates of movement (x ¯  =  3.1 m/min; maximum  =  6.0 m/min) than those reported for loggerhead hatchlings in Florida (5.2–7.0 m/min; Dial 1987), which could reflect species-specific differences. However, Dial (1987) only included hatchlings that maintained constant hatchling frenzy locomotion and traveled directly to the water, which could also explain the differences between the 2 studies. Additionally, we measured the rate of movement between 2 points (nest and water line) regardless of the straightness of the route followed. Therefore, our values do not reflect the actual speed at which the animal moved but the rate of movement between the nest and the water. The rate of movement of both leatherback and loggerhead hatchlings were below the reported speeds for ghost crabs. Moving crabs reach speeds of about 4.8 m/min, but under stress can reach speeds as high as 49.8 m/min (Weinstein 1995). Therefore, they can easily outrun their prey.

Ghost crabs were the most common predators of hatchling leatherback turtles, followed by night herons and diurnal predators. Surprisingly, mammals did not depredate hatchlings during the night observations. The presence of the observer may have prevented mammals from approaching the nest. However, this is unlikely because raccoons frequently jump over the fence of the hatchery at Playa Grande when people are inside, and generally act fearlessly. A second possibility is that raccoons are more interested in eggs than in hatchlings (Stancyk 1979), although raccoons have been observed eating hatchlings at Playa Grande in the past. Additionally, none of the marked leatherback nests were depredated during incubation. However, olive ridley (Lepidochelys olivacea) eggs are frequently eaten by raccoons at Playa Grande. A third explanation comes from the opportunistic nature of raccoons as predators (Garmestani and Percival 2005). Since the number of female leatherbacks has declined by 95% at Playa Grande (Spotila et al. 2000; Santidrián Tomillo et al. 2007), the production of hatchlings also has declined. Thus, the probability of being on the beach when hatchlings emerge from a nest is very low and raccoons may have reduced their visits to the beach as a consequence.

Domestic animals, mainly dogs, disturbed 8% of the nests (during the day, at dawn, or dusk) in 2005–2006. The presence of domestic animals on the beach is increasing at Playa Grande because of increasing development. The impact of domestic animals increased in 2007–2008 and 2008–2009 when more than 50% of nests showed some sort of disturbance by dogs. This is higher than the rate of 33% reported for leatherback nests at Tortuguero, Costa Rica (Leslie et al. 1996) and similar to the impact of sand goannas (Varanaus panoptes) on flatback hatchlings (52%–67%) at Fog Bay in Australia (Blamires et al. 2003). Predation on hatchlings by unnatural predators and other human-related activities (i.e., lights, sand compaction, trash, etc.) is turning tourist development into the greatest threat to the survival of the leatherback turtle population at Las Baulas.

Despite the higher concentration of nests that are deposited every year around the central part of the beach at Playa Grande, predation in general was not correlated with the location along the beach. Predation by yellow-crowned night herons, however, was significantly greater toward the south end. This could be related to the shortest distance between the southern end and the estuary of Tamarindo, where yellow-crowned herons frequently are seen during the day.

We found that dispersion of hatchlings on the crawl to the water increased with the number of tracks both near the nest and by the high-tide line. Hatchlings could benefit from dispersing on the crawl to the water in several ways: 1) they could be more efficient at finding and reaching the water if no other hatchlings are in their way; 2) if the probability of being depredated is higher at the core of the group, some hatchlings would disperse away from the group to increase their chances of surviving; and 3) dispersing and increasing the distance from each other could make the capture of multiple hatchlings by a single predator difficult. We could not test the above hypotheses because of small sample sizes. Further studies are needed to investigate the role of dispersion on the predation risk.

Hatchlings followed paths consistently straighter once they were near the high-tide line than when they were around the nest. Therefore, their ability to find the ocean and follow the shortest possible way improves as they get closer to the water. However, the straightness index is affected by both the straightness of the track and the direction the hatchling is following. A hatchling crawling in a straight line but in the wrong direction results in a low straightness index, even if the track was straight. We measured straightness to the water, so hatchlings with low straightness indices were those that left a wavy track and those that were oriented to a source other than the water. We also found straighter tracks coming from nests deposited toward the southern part of the beach. Because of the shape of the beach, hatchlings at the north end have a higher exposure to the lights coming from the town of Tamarindo, although no measurements on light intensity and exposure were taken. Despite the conservation efforts of the Las Baulas National Park to protect leatherback turtles from development, light pollution at Tamarindo has increased considerably in the last 10 years and continues to rise annually. The town of Tamarindo has become a developed island surrounded by Las Baulas National Park, and there are no regulations on night illumination.

The population of leatherback turtles that nests at Las Baulas Park has been intensively studied for more than 15 years (Steyermark et al. 1996; Reina et al. 2002; Santidrián Tomillo et al. 2007). However, the effects of hatchling predation had not been previously assessed at Playa Grande. This study constitutes the first estimation of predation rates of hatchlings on the beach and the first description of the process of departure from the nest to the water. Future studies should explore the effect of dilution, detection, and dispersion on predation risk. Additionally, the extent of inshore predation remains unknown and needs to be quantified.