Nearshore Turtle Hatchling Distribution and Predation

 Nicolas J. Pitcher, Simon Enderby, Thomas Stringell, Lynne Bateman

 Abstract

Hatchling green turtles (Chelonia mydas) emerge from their nests at night and crawl over the beach to the sea using light as the primary cue for seafinding. Once they enter the sea they are usually guided by wave and magnetic cues and head straight off shore. Predation rates decrease as hatchlings distance themselves from their natal nests.

In Sabah the turtles' open sea finding mechanism must be altered slightly to account for the numerous small islands and reefs that prevent a clear off shore migration, and hatchlings face higher rates of predation than elsewhere in the world..To test nearshore orientation and rates of predation, hatchlings were fitted with harnesses and lighted floats and were followed as they left the natal beach. Swim speed and off shore headings were determined from repeated position data gathered by DGPS and triangulation, and predation rate and location were notedfor each trial.

Hatchlings suffered 40-60 % mortality within the first two hours at sea, the majority before they crossed the 10 m depth contour. Once hatchlings made it past the reef and into deeper water predation rates dropped by two thirds. Hatchlings were found to orient in the nearshore in the absence of waves, and to turn toward the open sea long after leaving the beach, after picking up orientation cues from waves and currents.

Introduction

Hatchling sea turtles emerge from their nests and crawl down the beach to the sea. Once they enter the surf line they begin a general offshore migration during which they face a number of obstacles, in particular currents, waves and predators. A hatchling's ability to move offshore quickly greatly increases its chances of survival, as predation rate decreases with depth and distance from land (Gyuris 1994). In general, hatchlings are known to orient directly offshore using a combination of wave, light and magnetic cues (Lohmann & Lohmann 1996). In particular, hatchlings orient into oncoming waves as they leave the shoreline, a cue which helps take them directly offshore (Goff et al. 1998, Lohmann & Lohmann 1992). One currently unresolved issue is how hatchlings determine a 'direct offshore' bearing in the absence of waves, such as when swimming through sheltered lagoons and embayments. During studies by Salmon & Lohmann (1989), hatchlings oriented randomly in the absence of waves. It is known that hatchlings can acquire a swim bearing during the beach crawl (Lohmann et al. 1995a), and it may be possible that this bearing is used until another directional cue is picked up by the hatchling. Most studies to date have been carried out in locations where natal beaches fronted open oceans, such as those in South Florida, where thousands of miles of open Atlantic Ocean await the hatchlings, and where beach waves generally approach the shore perpendicularly. In Sabah the situation differs in having numerous small islands and reefs in the path of what would normally be a 'directly offshore' heading. In many cases beaches face other islands, and even the mainland coastline. If hatchlings were to simply swim in a straight line away from their natal beach, and in the absence of other cues, they would end up running into land once again, suggesting that a delayed acquisition of directional cues contributes to their offshore movements. It was postulated that hatchlings are able to set an initial offshore heading in the absence of waves, and do not simply swim directly offshore, rather, they seek directions which would carry them into open sea situations after picking up directional cues from waves after they exit from sheltered (wave-less) waters. In addition, it was believed that hatchlings might rely on cues acquired during the crawl to initially maintain an offshore bearing in the sheltered waters.

Predation on hatchlings is also a concern in Sabah for two reasons. First, the hatchlings are normally released en masse after having spent several hours in a hatchery enclosure. Each night up to ten or twenty nests might be released together, and it was believed that predation rates could be higher under these circumstances than if hatchlings were released in smaller groups, approximating what would occur under natural emergence conditions, where a significant rate of predation upon hatchlings already exists (Central East Florida: 6.8 %, Witherington & Salmon 1992; Bermuda: 8.3 %, Frick 1976, Hirth 1971). Second, hatchlings face the highest chance of predation when close to reef structures that support a greater density of predators than does the open sea (see Stancyck 1982; Witherington & Salmon 1992, Witzell 1981, Wyneken et al. 1994), and those hatchlings leaving the Sabah islands must pass by a number of reef structures / islands before reaching the open sea..

This study attempted to determine initial offshore headings for hatching green turtles (Chelonia mydas), distances they may be able to put between themselves and the shoreline after hatching, and relative predation in deep (> 10 m, away from reef structures) versus shallow water in more complex oceanographic conditions than those previously studied.

Methods and Materials

Swimming trials were carried out off Selingaan island (6^°10'30" N, 1 l8^°03'30"E) in Sabah, Malaysia (Fig. 1) during March and April and again in December 1998. Green turtle hatchlings were collected from nests in the island's hatchery immediately after emerging, between 1700 and 2000h, and fitted with a light-bearing float. Hatchlings were placed in a Lycra harness (as per Salmon & Wyneken 1987) to which a Styrofoam float and light were attached with 80 cm of nylon fishing line (Fig. 2). The Styrofoam float measured approximately 4x6 cm and the light source consisted of three small chemical lights (Starlight SL-5, Luminous Arts Japan Ltd.) placed in a 2.5 ml plastic tube glued to the float. The float was painted black on the underside to minimise detection by underwater predators. The total weight of the light/float averaged less than 2 g (x = 1.92, SD = 0.064, n = 21). Hatchlings were released 5 to 10 m from the shoreline and allowed to crawl down the beach and enter the sea naturally. As hatchlings crawled down the beach, the floats were carried by team members to minimise drag and chances of entanglement. Tracking was accomplished using two different methods. During the first set of trials in March/April, in order to determine precise locational data, 21 hatchlings were tracked individually by following 10- 15 m behind in a small inflatable boat, and their position and water depth were recorded every five minutes using a Differential GPS (Furuno F-80 + 10 m) and Hummingbird depth sounder (± 10 cm). Notes were also kept on swimming behaviour, direction and water conditions. Hatchlings were collected and released if they had headed into the current and remained geographically motionless for 60 mins or more (is was assumed this represented a point at which they had picked up directional cues from waves or magnetic fields), or because wave and wind conditions exceeded those under which normal swimming patterns could be maintained, and on one occasion due to personnel safety. In December the primary objective of the second set of trials was to determine rates of predation and only approximate location and depth, and 62 hatchlings were tracked in groups of four. During these trials, individual hatchlings towed the same type of light source, but four hatchlings were released simultaneously from the beach, and were tracked by a single boat. The hatchlings' location at point of depredation was determined by triangulation of three bearings on fixed locations using a hand-held compass (± 5°) and water depth was determined using a Scubapro PDS Sonar (± 10 cm). Due to natural dispersal patterns, these hatchlings could not be reliably tracked for more than 60 min before being lost to sight. Hatchlings were collected and released if they were still swimming after 60 min, in keeping with reliability period of tracking data. Mean trajectory for these hatchlings was calculated from point of release to approximate loss/release point. From this trajectory a mean net heading inclusive of current displacement was calculated, along with mean net swimming speed and distance.

Fig. 1: General location of study area.

Surface water current, from a combination of sea- and wind-driven forces, was determined by allowing an floating object to be carried over a one minute period, where start and end positions were recorded by DGPS and net current speed was the function of distance over time. Currents, along with wave and hatchling orientation, were noted on a marine line-of-sight compass (±5°).

Circular statistical procedures (Zar 1984) were used to determine individual mean initial and final offshore bearings based on the first and last three displacements, and a second order overall mean bearing for both initial and final displacement before hatchlings were released or lost to predation. In the second trials, offshore orientation was based on mean bearing to last known location. The Watson one-sample U test was used to determine if bearings were significantly oriented in any particular direction. A modified Rayleigh test, the V test, was used to test closeness to expected offshore orientation and waves. A geometric breakdown of individual position records was used to determine individual over-the-ground distances covered.

Figure 2: Float used in tracking trials, a. Schematic of complete set-up. b. Top view of float, measuring 6 x4 cm. c. Lycra harness, cutfrom a single piece and folded over, then point-stitched on the distant corners.

Results

Dispersion - Precise positional data (± 5 m) was calculated for the 19 hatchtings tracked with DGPS. All of these hatchlings were released from the W-facing beach, where hatchlings were released en masse by Park staff. Initial swimming orientation for these hatchlings, calculated as the net heading inclusive of current displacement over the first 15 min, was significantly grouped with an average angle a = 208.2° and vector magnitude r = 0.868 (Fig. 3a). However, this orientation was SW and significantly different from a 255° direct offshore heading (V-test: u = -5.091, P < 0.005). No waves were present in the nearshore waters during the nearshore portion of the trials that might have had an influence on the hatchlings' orientation. Hatchlings tended to swim in a SW direction initially, after which they turned north, or continued further south, until out of the shelter of the island (Fig. 4). Final orientations, calculated as the mean net heading over the last 15 min prior to depredation or release, were found to be more randomly distributed with a weak overall vector r = 0.484. The mean overall final bearing of a = 163.2 (Fig. 3b) is indicative of a gradual shift toward the NW away from the 'direct offshore' heading which hatchlings were expected to have picked up on the beach crawl.

All of the six hatchlings that swam south (represented by the six lowest tracks, Fig. 4) turned and oriented into small waves (<20 cm) having a mean heading of 269° and a current of 1.5 to 2.2 km.hr-¹. However, the hatchlings' swimming efforts only marginally maintained headway against the current. In all but one case the hatchlings were pushed sideways (S) or backwards (SW to W). After swimming on the same heading for 60 min or more the hatchlings were collected and released, as they were considered to have acquired and settled into an offshore orientation opposite that of oncoming waves and current.

Four of the six hatchlings that turned north (the upper six tracks in Fig. 4) were collected and released after 60min or more, again after it was assumed a general offshore course had been established based on waves and current (mean heading = 210°). Though these hatchlings were also subjected to currents from the NE once past the shelter of the island, the severity of the effect was reduced due to a deeper and more open expanse of water, where currents were not found to exceed 1.5 km.hr^-1. Overall, these hatchlings appeared to try to round the island and head in a N to NW to W direction, away from the direct offshore bearing from the beach, and out to the open sea.

The second set of trials, which only determined an approximate final position (± 25 m), included hatchlings released at the W- and the E-facing beaches, at the same time as when single nests of hatchlings were released by Park staff. Although these trials were aimed mainly at determining rates of predation, and not as detailed and precise as the earlier trials, these data provide an indication of general movements away from the beach, and include hatchlings released into the prevailing currents from the E-facing beach. Statistical analysis of the orientations of hatchlings swimming away from the E-facing beach produced a mean vector strength r = 0.911, indicating a strong tendency to orient with an average overall heading a = 112.1° (Fig. 3c). Mean wave direction during these swims was 87.8°, and not uniformly distributed (r = 0.200). Swim direction was not overly dependant on wave direction, and average swim direction and wave direction were weakly correlated (r² = 0.093), suggesting that around the Sabah islands waves are not primarily responsible for initial hatchling orientation. Analysis of the orientations of hatchlings swimming away from the W-facing beach produced a mean vector strength r = 0.949, also indicating a strong tendency to orient with an average overall heading a = 257.7° (Fig. 3d). During these swims, average wave direction was 206.4°, and was slightly more grouped than in previous trials (r = 0.378). A correlation between swim speed and direction resulted in a coefficient r² = 0.103.

Swim Distance - The most accurate swim data was derived through the first set of trials, with positional data collected through DGPS. During these trials, total distance covered by hatchlings ranged from 5 m to 6,687 m. However, due to the variable time each was followed, the overall average swim speed (1.2 km.hr^-1) was considered a more representative expression of swim performance (Table I). The overall mean swim speed under natural conditions ranged from 0.000 to 0.965 m·sec-¹. The lowest speed (0.000) was posted by those hatchlings which were taken almost immediately by predators, and when hatchlings paused to rest between swimming bouts. The average time it took hatchlings to cross the 10 m depth contour was 39.5 minutes (after covering roughly 786 m). Although during the present study the distances were calculated from positional data measured every five minutes, they do not always reflect the exact swimming movements as in many cases hatchlings retraced their swims several times within the five minute period, resulting in shorter net distance covered.

During the second set of trials, in which hatchlings were released in groups of four, and final position was only determined through bearing triangulation, the accuracy of positional data was only ± 11.8 m for each 5° change in angle averaged over all trials. Given the compass-reading accuracy, positional data were believed to be accurate only within a 22 m x 22 m square. In these trials, the average swim speed away from the W-facing beach was 0.206 m.sec-¹ and that from the E-facing beach was 0.428 m.sec^-1, within the range found with the more accurate tracking.

Figure 3: Dispersal patterns for hatchlings leaving P. Selingaan. A - Initial orientation of hatchlings departing W beach during DGPS-tracked trials. B - Final orientation of hatchlings after dispersal. C - Initial orientation of hatchlings departing W beach using triangulation method during predation trials. D - Initial orientation of hatchlings departing E beach using triangulation method during predation finals.

Figure 4: Tracks of hatchling movements as they left Selingaan Is. (shaded, lower right), and across the 10 m depth contour. Check marks indicate hatchlings that were released after monitoring, crosses indicate hatchlings that were lost to predators.

Table I: Summary of hatchling swim data. Initial bearings computed from first three position changes, final bearings computed from last three position changes. C - Collected and released E - Eaten by predator

Predation - Of the 21 hatchlings tracked during the March/April study, thirteen (61.9 %) were lost to predation, and the remainder were collected and released. Predation was determined after the floats/lights were noted to dip suddenly and then disappear or stop moving, and upon closer inspection the line between the float and hatchling was found to be severed. In all but one case the floats were recovered. It is not believed the hatchlings could have broken the 2.5 kg line unaided. These hatchlings were released individually at the same time as multiple nests were released en masse by Park staff .

In the December study, 8 hatchlings were lost from sight, and discarded from overall calculations, as it could not be determined if they were lost to predation or simply out of sight. Of the remaining 54, 22 (40.7 %) were lost to predators (all floats were collected) and the balance were released. These hatchlings were released at the same time as single nests were released by Park staff, except for on two occasions, when two nests of hatchlings were released simultaneously. The predation rate when hatchlings were released en-masse (61.9 %) was nearly 50 % higher than when hatchlings were released in smaller groups (40.7 %). Overall, 46.7 % of all hatchlings were lost to predators. The average water depth at which hatchlings were taken by predators (x = 6.3 in) was found to be significantly shallower (z Test: z = 5.747, P = 0) than that at which they were collected and released (x = 15.0 m). Only 3 of the 40 hatchlings that were collected were in shallow water (< 10 m). It is suggesting that those hatchlings that manage to swim past the reef (past a drop-off and into >l0m-deep water) have a much greater chance of survival within the first hour of swimming. Of the 43 hatchlings that made it to deep water, only 9 were lost to predators (20.9 %). Of the 30 hatchlings that remained in shallow water, 23 were lost to predators (76.7 %), substantially higher than the rate of predation in deep water. These hatchlings may have been depredated at a later time, but within the first 1- 2 hours after entering the sea, those hatchlings that made it to deep water had a higher change of survival. Within the scope of these trials, the probability of a hatchling surviving the first few hours at sea once it reached deep water was 0.46, whereas the probability of survival within the first swimming hours in shallower reef waters was 0.01.

Discussion

Orientation - These trials have shown that hatchlings can and do navigate around small landmasses to head toward the open sea, rather than just swimming directly away from shore, and that they can establish offshore orientations in the absence of waves. The orientation pattern hatchlings adopt as they leave the W shores of P. Selingaan suggest that in the absence of waves hatchlings will swim until they pick up other stimuli (summarised by Wyneken, 1996), in particular as their initial offshore orientation differs from the expected offshore heading and their subsequent change of direction to orient into oncoming waves.

Hatchlings are known to be attracted to light sources while on the beach, (Mrosovsky & Carr 1967, Witherington & Bjorndal 1990) but once they enter the sea the visual cues are less important in determining an offshore heading (Lohmann & Lohmann 1996, Salmon & Wyneken 1990). In these trials hatchlings were allowed to crawl down the beach to acquire an initial offshore heading (as per Lohmann et al. 1990), during which they oriented in a heading approximating 255°. Once in the water, they were found to orient differently, toward 208° (initial trials of the W beach), 112° (second trials, E beach) and 257° (second trials, W beach). Although hatchlings are known to orient into waves (Goff et al. 1998, Lohmann & Lohmann 1992) this is normally done as they enter the sea and cross the surf line. Hatchlings released from the E beach reflected this pattern heading offshore on an average course heading of 112°. However, hatchlings released on the W beach had been swimming for at least 45 minutes through sheltered waters before encountering the waves as they rounded the northern or southern end of the reef/island.

The W beach from which the hatchlings were released was sheltered from the predominant water currents and waves at the time the trials were conducted. In contrast to findings by Salmon & Lohmann (1989), these hatchlings assumed a particular course of travel rather than random orientation in the absence of waves. This suggests that not all hatchlings need waves to establish an initial offshore orientation. It is believed that the beach crawl similarly did not influence the initial offshore orientation as hatchlings navigated away from the island on a range of headings, from 160° to 300° but in general further SW than the 255°- oriented beach crawl. Once away from the island shelter, however, hatchlings oriented into the waves and currents as found at other locations (Lohmann & Lohmann 1992, Lohmann et al. 1995b, Salmon & Lohmann 1989). Although Lohmann et al. (1995a) discusses the need for a suitable beach crawl to establish a subsequent magnetic compass heading, this mechanism could not be the primary cue used by hatchlings leaving the W beach as they swam a significant distance offshore in a SW direction before assuming the offshore course that would take them to the open sea, in contrast to the manner in which hatchling released from the E beach oriented in a E direction.

If hatchlings were to leave the beach at Selingaan and head directly offshore on a bearing of 255° (grey arrow, Fig. 5), they would end up in a semi- enclosed bay on the E coast of Sabah (N of Sandakan, Fig. 5). Heading NW or S they would also encounter the Sabah coast. To reach the nearest 'open' sea they would need to head toward the Philippines in a NE direction (arrows, lower map, Fig. 5), consistent with the present findings. However, these findings are representative only of the first few hours of swimming and do not address long- term dispersal patterns. For this, trials of longer duration are needed, for which the present methodology is unsuitable. Given the logistical constraints of long- term hatchling tracking (see Liew & Chan 1995), it is suggested that only with the development of miniature transmitters that can be towed by hatchlings will long-term tracking be feasible, and always at the risk of hatchling and transmitter loss through predation.

Predation - The trials also indicate that a significant amount of predation occurs within the first hour at sea (40-60 %). This raises an important management consideration for conservation programmes in which the release of numerous hatchlings is a prime consideration. These rates of predation are six to ten times greater than that reported for other locations (see Frick 1976, Witherington & Salmon, 1992) and is believed to be, at least partially, a result of the density at which hatchlings enter the sea.

Under normal emergence conditions, a clutch of 60 to 100 hatchlings reach the sea over a period of several minutes, and disperse after they enter the surf zone (Gyuris 1993). In Sabah, due to hatchery practices, several hundred (sometimes even several thousand) hatchlings may all be released within a five-metre strip of beach at the same time. Even assuming natural dispersion once they enter the sea, the density of hatchlings per unit area in these cases is many times greater than that which would occur naturally. The 60 % predation rate found in Sabah during trials when multiple nests were released simultaneously, and the 40 % predation rate found during the second set of trials when hatchlings were released among single nests, far exceeds the <10 % predation at other locations, although it unlikely that the density of predators is significantly higher in Sabah than elsewhere.

It is possible that the float/harness combination influenced the rate of predation, though this is not believed to be significant, or the cause of a 300 % in crease in predation over other localities. Hatchlings with floats were observed to swim in a similar manner to those without floats, adopting the 'powerstroke' style described by Wyneken & Salmon (1990). Although the out-of-water weight of the floats was 2 g, the in-water weight, after accounting for the buoyant forces of seawater, was negligible. In addition, the dark colouration of the float, although presenting a silhouette shape from below, was believed to minimise observation by predators, and the lights were embedded (nearly flush) with the upper surface of the float and did not shine on the water. Hatchlings have been tracked in the open sea in a number of studies using a similar protocol (Gyuris 1994, Ireland et al. 1978, Liew & Chan 1995, Witherington 1991) and it is believed that although predation may increase, it represents only a marginal increase over the overall possibility of a hatchling being depredated. Irrespective of this, hatchlings in both sets of trials were similarly equipped, while predation among en masse-released hatchlings was higher than among single nest releases.

Part of the unnaturally high predation in Sabah is a result of the manner in which hatchlings are released night after night in the same location, in dense (multiple nest) groups. In addition, although this study did not address the issue of repetitive releases at the same location, in contrast to natural emergence cases whereby hatchlings enter the sea at random locations, the continuous release of hatchlings from the same 5 m stretch of beach may be responsible for a certain amount of 'learning' by the local predator population. To optimise the Park's management procedures though, and those of other locations where similar release practices occur, it is suggested that hatchlings are released fewer at a time, and that the release point be randomly changed to avoid congregations of predators.

Figure 5: Selingaan Is. and its geographic relation with nearby islands and the Sabah mainland. Upper grey arrow indicates a 'direct offshore' heading for hatchlings, the upper clear arrows represent the true headings found in this study. Lower arrows indicate a theoretical 'shortest' distribution pattern that would enable hatchlings to reach the open sea.

Acknowledgements

We gratefully acknowledge the assistance of Sabah Parks directors and personnel for logistics support during the project, in particular Datuk Lamri Ali, Paul Basintal, Asdari Belout and Karim Kassim. This work was partially supported by MacArthur Foundation Grant No. 44416-0 (to Sabah Parks) and UNIMAS Grant No. 90/96(9).

References

Frick, 1., 1976. Orientation and behaviour of hatchling green turtles (Chelonia mydas) in the sea. Anim. Behav., 24: 849-857.

Goff, M., Salmon, M. & Lohmann, K., 1998. Hatchling sea turtles use surface waves to establish a magnetic compass direction. Anim. Behav. 55: 66-77.

Gyuris, E., 1993. Factors that control the emergence of green turtle hatchlings from the nest. Wildl. Res., 20: 345.

Gyuris, E., 1994. The rate of predation by fishes on hatchlings of the green turtle. Coral Reefs 13: 137.

Hirth, H.F., 1971. Synopsis of the Biological Data on the Green Tuttle, Chelonia mydas (Linnaeus) 1758, FAO, Rome.

Ireland, L.C., J.A Frick & D.B Wingate, 1978. Nighttime orientation of hatchling green turtles (Chelonia mydas) in the open ocean. In: Animal Migration, Navigation, and Homing, (K. Schmidt-Koenig & W.J. Keeton, eds.): Springer-Verlag, New York: 420-429.

Liew, H.C. & E.H. Chan, 1995. Radio-tracking leatherback hatchlings during their swimming frenzy. In: Procs.l2th Annual Workshop on Sea Turtle Biology and Conservation, NOAA, Tech. Memo. NMFS-SEFSC-361: 274.

Lohmann, K.J., C.M.E. Lohmann & J.C. Callaway, 1995a. Evidence that beach crawl direction sets the magnetic compass in loggerhead hatchlings. Tech. Memo. NMFSSEFC-361: 71p.

Lohmann, K.J. & C.M.E. Lohmann, 1992. Orientation to oceanic waves by green turtle hatchlings.J. Exp. Biol. 171: 1-13.

Lohmann, K.J. & C.M.E. Lohmann, 1996. Orientation and open-sea navigation in sea turtles. J. Exp. Biol. 199: 73-81.

Lohmann, K.J., M. Salmon & J. Wyneken, 1990. Functional autonomy of land and sea orientation systems in sea turtle hatchlings. Biol. Bull 179: 214-218.

Lohmann, K.J., A.W. Swartz & C.M.E. Lohmann, 1995b. Perception of ocean wave direction by sea turtles. J. Exp. Biol. 198: 1079-1085.

Mrosovsky, N. & A. Carr, 1967. Preference for light of short wavelengths in hatchling green sea turtles, Chelonia mydas, tested on their natural nesting beaches. Behaviour 28:217-231.

Salmon, M. & K.J. Lohmann, 1989. Orientation cues used by hatchling loggerhead sea turtles (Caretta caretta) during their offshore migration. Ethology 83: 215.

Salmon, M. & J. Wyneken, 1987. Orientation and swimming behavior of hatchling loggerhead turtles (Caretta caretta L.) during their offshore migration. J Exp. Mar. Biol. Ecol 109: 137.

Salmon, M. & J. Wyneken, 1990. Do swimming loggerhead sea turtles (Caretta caretta L.) use light cues for offshore orientation? Mar. Behav. Physiol. 17: 233-246.

Stancyck, SE., 1982. Non-human predators of sea turtles and their control. In: Biology and Conservation of Sea Turtles, (K.A. Bjorndal, ed): Smithsonian Institution Press, Washington, D.C.: 139-152.

Witherington, B.E., 1991. Orientation of hatchling loggerhead turtles at sea off artificially lighted and dark beaches. J. Exp. Mar. Biol Ecol 149:1-11.

Witherington, B.E. & K.A. Bjorndal, 1990. Influences of artificial lighting on the seaward orientation of hatchling loggerhead turtles Caretta caretta. Biol Cons. 53: 139-149.

Witherington, B.E. & M. Salmon, 1992. Predation on Loggerhead turtle hatchlings after entering the sea. J. Herp. 26(2): 226-228.

Witzell, W.N., 1981. Predation on juvenile green turtles, Chelonia mydas, by a grouper Promicrops lanceolatus (Pisces; Serranidae) in the Kingdom of Tonga, South Pacific. Bull. Mar. Sci. 31(4): 935-936.

Wyneken, J., 1996. Sea Turtle locomotion: Mechanics, Behavior amd Energetics. In: The Biology of Sea Turtles, (P.L. Lutz & J.A. Musick, eds.): CRC Press, New York, 165-198.

Wyneken, J., M. Goff & L. Glenn, 1994. The trials and tribulations of swimming in the near-shore environment. In: Procs.14th Annu. Symp. Sea Turtle Biology and Conservation, Miami, FL, NOAA, Tech. Memo. NMFS-SEFC-351: 169-171.

Wyneken, J. & M. Salmon, 1990. Swimming the offshore marathon: analyses of early swimming activity in loggerhead, green turtle and leatherback hatchlings. In: Procs.10th Annu. Workshop on Sea Turtle Biology and Conservation, Hilton Head, NOAA, Tech. Memo. NMFS-SEFC-278: 109-110.

Zar, J.H., 1984. Biostatistical Analysis, 2nd, Prentice-Hall, Englewood Cliffs, New Jersey: 718.