INCUBATION TEMPERATURE EFFECTS ON HATCHLING PERFORMANCE IN THE LOGGERHEAD SEA TURTLE (CARETTA CARETTA) A thesis submitted in partial fulfillment of the requirements for the degree

MASTER OF SCIENCE

in

MARINE BIOLOGY

by

LEAH FISHER OCTOBER 2012

at

THE GRADUATE SCHOOL OF THE COLLEGE OF CHARLESTON

Approved by:

Dr. David Owens, Thesis Advisor

Dr. Allan Strand

Dr. Eric McElroy

Dr. Melissa Hughes

Dr. Amy T. McCandless, Dean of the Graduate School

ABSTRACT

INCUBATION TEMPERATURE EFFECTS ON HATCHLING PERFORMANCE IN THE LOGGERHEAD SEA TURTLE (CARETTA CARETTA) A thesis submitted in partial fulfillment of the requirements for the degree

MASTER OF SCIENCE

in

MARINE BIOLOGY

by

LEAH FISHER OCTOBER 2012

at

THE GRADUATE SCHOOL OF THE COLLEGE OF CHARLESTON

Incubation temperature has significant developmental effects on oviparous animals, including determining sex for several species. For the Northwest Atlantic loggerhead sea turtle (Caretta caretta), apparent population-wide female-biased hatchling sex ratios contrast with observations of juvenile populations, where sex ratios have remained constant at about 2 to 1 female-biased over the past 30 years. It has been suggested that some unknown factor is affecting loggerhead survival resulting in an unexplained differential loss of ~60% of female hatchlings per year. One theory to explain this hatchling mortality is tested in this project, that incubation temperature affects traits that influence survival. Furthermore, there may be differential survival between male and female hatchlings. I conducted laboratory experiments to test for an effect of incubation temperature on performance of loggerhead hatchlings. I tested 68 hatchlings produced from eggs incubated at 8 different constant temperatures ranging from ~27ºC to ~32.5ºC. Following their emergence from the eggs, I tested righting response, crawling speed, and conducted a 24- hour long hatchling swim test. Data indicate an effect of incubation temperature on survivorship, righting response time, crawling speed, change in crawl speed, and overall swim activity, with hatchlings incubated at 27ºC showing decreased locomotor abilities. No hatchlings survived when incubated at 32ºC and above. Differences in survivorship of hatchlings incubated at high temperatures are important in light of projected higher sand temperatures due to climate change, and could indicate increased mortality from incubation temperature effects.

TABLE OF CONTENTS

Introduction 1 - 10 Materials and Methods 10 - 17 Egg Collection and Incubation Protocol Hatching Protocol Experimental Protocol Righting Response and Crawl Test Swim Test Video Analysis Statistical Analysis Results 17 - 27 Hatch Success and Survivorship Incubation Temperature Effects on Performance Variables Composite Rank Clutch Effect on Performance Variables Discussion 27 - 38 Hatch Success and Survivorship Performance + Incubation Temperature Implications

Beyond Controlled Incubation Facing Climate Change Conclusions 39 Future Directions 40 - 42 Acknowledgements 43 Literature Cited 44 - 52

INTRODUCTION

Temperature plays a critical role in animal development. In particular, incubation temperature has significant developmental effects on oviparous animals, including determining sex for several species (Deeming and Ferguson 1991). Temperature- dependent sex determination (TSD) is found in seven orders of fish (Conover 2004), and many reptiles: all crocodilians, tuataras, many turtles, and some lizards (reviewed in

Valenzuela and Lance 2004). TSD is a type of environmental sex determination (ESD), and contrasts with the more common genetic sex determination (GSD). For reptiles with

TSD, climate change impacts, particularly rising temperatures around the world, dictate a need for an improved understanding of how even subtle temperature changes in the nest environment will influence the survivorship of both individuals and species in the next few decades.

All sea turtles exhibit TSD, with temperatures greater than ~29°C producing females for six of the seven species, and temperatures below producing males (review in

Raynaud and Pieau 1985, Mrosovsky 1988). It has been predicted that females born in warm temperatures would have an increased fitness compared to males born at that same temperature, and the reverse for cool temperatures. Within this prediction is another hypothesis, stating that at the extreme incubation temperatures where only one sex is produced, hatchlings could reach optimal fitness for that sex. On the other hand, close to pivotal temperature may be optimal for both sexes. The fitness of sea turtle hatchlings in terms of lifetime fecundity is nearly impossible to test directly due to their slow growth and late maturity (Hamann et al. 2007). Thus, phenotypic measures have been used to estimate hatchling fitness based on other effects of incubation temperature.

1 Incubation Temperature Effects

Incubation temperature influences several factors in sea turtle egg incubation besides hatchling sex, including development time, incubation duration, size, mass, and amount of yolk content converted to hatchling tissue (Foley 1998, Reece et al. 2002,

Booth et al. 2004). Embryonic development is faster at warmer temperatures, resulting in shorter incubation periods (Davenport 1997). Incubation temperature can also affect physical traits of hatchlings, such as size and locomotor performance for turtles. It is generally assumed that an increase in locomotor activity, for example crawling faster or swimming longer, is advantageous because it may allow hatchlings to better avoid predators (Janzen 1993). In turn, this increased performance could affect lifetime fitness.

For example, a study on the snapping turtle (Chelydra serpentina) found that even after six months that included a period of hibernation, incubation temperature maintained an effect on growth rates and water temperature preferences for male and female turtles

(O’Steen 1998). Another study on the Ouachita map turtle (Graptemys ouachitensis) showed an effect of incubation temperature on performance both immediately after hatching and after one year (Freedberg et al. 2004).

For freshwater turtles, it has been reported that hatchlings show decreased performance ability at the upper and lower extremes of incubation temperature (Micheli-

Campbell et al. 2011, Booth et al. 2004, Freedberg et al. 2004). For loggerhead (Caretta caretta) and green sea turtles (Chelonia mydas), several studies (Booth and Astill 2001,

Booth et al. 2004, Booth 2006, Hamann et al. 2007) have suggested that hatchlings produced at warmer temperatures may be smaller, swim faster, and possibly even grow faster (Glen et al. 2003); however, these studies did not test the upper limit of incubation

2 temperature, and the fitness consequences remain untested or unconfirmed (Hawkes et al.

2009).

More specifically for green sea turtles, hatchlings incubated at 26°C were seen to have lower flipper stroke rate frequency and force output than hatchlings incubated at

28°C and 30°C (Burgess et al. 2006, Booth et al. 2004). This suggests that male hatchlings, coming from cooler temperatures, have a decreased locomotor ability compared to female hatchlings. In contrast for the loggerhead, one study reported that at cooler incubation temperatures, larger loggerhead eggs produced larger hatchlings, which swam faster but accumulated more lactate and exhibited slower growth rates (Foley

1998). No examination of sex was done in this study. A different study on loggerheads suggests that female hatchlings from warmer temperatures may be larger, possibly helping them evade predators (Reece et al. 2002). However, this study used different methods than similar studies, such as Burgess et al. (2006) and Booth et al. (2004), by physically moving eggs during incubation to manipulate incubation temperatures. Along with Foley (1998), these are the only two published studies on loggerhead hatchlings focusing on possible fitness consequences from physical differences based on incubation temperature. In an unpublished thesis experiment on loggerheads, a crawling test was used as a correlate to fitness, and it was reported that longer incubation duration, produced by cooler incubation temperatures, resulted in faster hatchlings (Nicar and

Weber 2002). This would suggest that male hatchlings perform better, contrasting with

Reece et al. (2002) and green turtle hatchling performance. These sparse and contradictory results indicate that a study designed to test physical and behavioral differences of loggerhead hatchlings incubated at a wide range of temperatures is needed.

3 Incubation temperature also determines if sea turtle hatchlings survive incubation because successful incubation only occurs within a small range of temperatures (Miller

1985). At temperatures above the upper range of optimal temperatures, approximately

33°C and higher for sea turtles, an increase in morphological abnormalities and decrease in hatching success are seen (Miller 1985, Miller et al. 2003). Even within the optimal range of incubation temperatures, Reid et al. (2009) found that hatchlings are more likely to have supernumerary, or irregularly patterned carapace scutes, with increased incubation temperature.

In Mon Repos, Australia, from 2005-2007, average incubation temperatures towards the hotter extreme were correlated with a decrease in hatching success for loggerhead hatchlings (Chu et al. 2008). Chu et al. (2008) state that a prolonged exposure to hotter incubation temperatures results in an overall weakening of hatchlings.

Furthermore, in Mon Repos, nest temperatures have reached 36°C for long periods of time, resulting in some death of eggs and hatchlings (Limpus 2006). In Japan, even though sand temperatures have not exceeded 33°C, hatchling mortality has been linked to high nest temperatures (Matsuzawa et al. 2002). Matsuzawa et al. (2002) note that this hatchling mortality occurred a few days prior to emergence, when metabolism of the hatchlings generates enough heat in the nest to raise nest temperature to lethal levels.

Metabolic heating has been reported and measured in more recent studies attempting to quantify the effects of climate change, such as for loggerheads nesting in North Carolina

(DeGregorio and Williard 2011) and Olive Ridleys (Lepidochelys olivacea) in Mexico

(Sandoval et al. 2011). The effects of high incubation temperatures combined with

4 metabolic heating raise concerns for all sea turtle species in light of global climate change.

Climate Change Connection

The fate of the Atlantic loggerhead sea turtle is uncertain. Anthropogenic threats combined with the impacts of rapid global climate change could have a major impact on the population of this species (Fuentes et al. 2009). Sea turtles are long-lived, slow growing, and late maturing animals, and in the past they have been able to adapt to changing climate, but with a significantly slower rate of climate change than is projected for the next 100 years (Hamann et al. 2007). Furthermore, sea turtle populations today remain at much lower levels compared to documented history (McClenachan et al. 2006,

Bell et al. 2007).

Due to these concerns and an unexpected decrease in nesting seen for the

Northwest Atlantic loggerhead population since ~2000, there was a recent push to upgrade the loggerhead to endangered species status rather than threatened (Conant et al.

2009). The decrease in nesting is believed to be linked to a decrease in overall population size; counting the number of nesting females has generally been used as the most reliable estimate of overall population size by estimating the number of adult females (Witherington et al. 2009). Counting sea turtles in the water is not considered as accurate as nesting beach data because sea turtles are distributed widely across an ocean basin (Meylan 1982, Schroeder and Murphy 1999). However, there is controversy surrounding the issue of the Atlantic loggerhead decline, with some believing that under current management practices, the loggerhead simply needs more time to demonstrate population growth since it takes 20-30 years for loggerhead sea turtles to reach sexual

5 maturity (TEWG 1998, Mike Arendt pers. comm.). In light of this controversy, research focusing on the potential effects of climate change is important and necessary to better understand, manage, and predict the population dynamics of this long-lived reptile into the future.

Rising sand temperatures on beaches are one way climate change may already be impacting sea turtle populations around the world. Temperatures at many nesting beaches worldwide have already been seen to be warming, for example in the Caribbean

(Glen and Mrosovsky 2004), South Atlantic (Hays et al. 2003), and Western Pacific (Chu et al. 2008). Due to TSD, there is a concern for all sea turtle species that the female to male sex ratio in the population is increasing and may approach 100% female, and that sea turtles will be unable to adapt at the rate needed to combat rapid climate change

(Hawkes et al. 2009). It is important to mention that sea turtle sex cannot be determined externally until adulthood, and that as hatchlings, the only accurate sexing method is to sacrifice hatchlings and perform histology on the gonads. Because of this, beach temperatures are used to estimate hatchling sex ratios. This concern is exacerbated by the knowledge that sea turtles are imprinted to their natal beach region and may not have the ability to adapt to climate shifts as rapidly as can other species that do not have natal homing (Hamann et al. 2010).

For the Atlantic loggerhead, eastern Florida beaches encompass 90% of the nesting productivity, and there is a significant possibility based on beach temperatures of a 100% female primary hatchling sex ratio emerging out of Florida (Shoop and Kenney

1992). In fact, female-dominated ratios have been reported on eastern Florida beaches since the 1980s. In 1986 Mrosovsky and Provancha (1989) determined the primary sex

6 ratio in Cape Canaveral to be greater than 93% female. This number was confirmed for three more years at Cape Canaveral, with sex ratios ranging from 87%-99% from 1986-

1988 (Mrosovsky and Provancha 1992). Recently, from 2004-2009, temperatures on

Florida beaches with the largest numbers of nesting loggerheads were seen to be consistently above the pivotal temperature, and generally hot enough to produce highly female-skewed hatchling sex ratios (Layton 2011). Although female-dominated sex ratios are often considered beneficial for population growth (Witzell et al. 2005), it is unknown how strongly female-biased ratios affect sea turtle populations. Furthermore, there are relatively few long-term sex ratio databases for any sea turtle species, essentially none that span multiple generations of sea turtle due to their late time to maturity and long lifespan. This makes it difficult to establish a baseline, “normal” sex ratio for a sea turtle population, complicating studying the effects of skewed sex ratios.

Thus, as Hawkes et al. (2009) state, future research should investigate the consequences of heavily female-biased primary sex ratios.

Loggerhead Sex Ratio Discrepancy

Due to the known skewed primary sex ratio, long-term studies on the juvenile

Atlantic loggerhead population monitor the sex ratio to look for an increase in female turtles. Hopkins-Murphy et al. (2003) determined the total expected female to male sex ratio of hatchlings entering the ocean to be 6F:1M using data compiled in the 1998

TEWG report. However, past studies on juveniles inhabiting neritic foraging grounds off the coast of North Carolina, South Carolina, Georgia, and Florida determined the sex ratio to be ~2F:1M (Owens 1997, Stabenau et al. 1996, Wibbels et al. 1987, 1991). More recent continuations of these monitoring programs are still reporting a fairly constant 2-

7 2.5F:1M juvenile sex ratio (Layton 2011, Arendt et al. 2010, Braun-McNeill et al. 2007).

Thus there is an apparent unexplained bias in hatchling sex ratios that is not maintained in the juvenile sex ratio. Juvenile loggerheads do exhibit slight natal homing, returning to waters in the vicinity of their natal rookery to forage (Bowen et al. 2004). This results in a slightly elevated number of female juveniles foraging off Florida compared to the South

Carolina or North Carolina coast, but still does not account for the difference seen in hatchling and juvenile sex ratios.

Furthermore, the genetic data for the juvenile population are not what is expected based on the proportion of hatchlings born in different nesting assemblages (Hopkins-

Murphy et al. 2003). Atlantic loggerheads are divided into subpopulations based on their nesting grounds: most Florida beaches represent the Southern subpopulation, and beaches from Northern Florida to North Carolina (29°-37.5°N) represent the Northern subpopulation (TEWG 1998). Genetic analyses on juvenile loggerheads have found that the Northern subpopulation is more highly represented than expected based on the relatively low nest density from Georgia to North Carolina (Hopkins-Murphy et al.

2003). Furthermore, recent studies conducted near Madeira Island in the Atlantic on younger pelagic-stage loggerheads also found the sex ratio to be 2.5F:1M (Delgado et al.

2008, Delgado et al. 2009), not significantly different from the neritic juvenile sex ratio; however, this is significantly different from the 6F:1M predicted primary sex ratio.

These studies on sex ratios suggest that there is a factor affecting loggerhead survival at an early stage, resulting in an unexplained differential loss of an estimated 2.4 million female hatchlings or about 60% of females per year (expanded from Hopkins-Murphy et al. 2003).

8 The predominant hypothesis to explain this potential differential mortality of loggerhead hatchlings is that the incubation environment is affecting trait(s) that influence hatchling survival. Other hypotheses include an undiscovered juvenile female foraging ground (Hopkins-Murphy et al. 2003, Arendt et al. 2009), or a genetic variable affecting hatchling dispersal (Norman et al. 1994, Wyneken et al. 1998). However, previous research discussed above has shown that the incubation environment, particularly incubation temperature, can affect numerous hatchling traits that could correspond to decreased survival.

Controlled Incubation Experiment Needed

Our main goal in this study was to test for performance differences between loggerhead hatchlings incubated at different controlled temperatures. A standardized laboratory experiment that incubates loggerhead eggs across the viable incubation temperature range can investigate potential incubation temperature effects on hatchling fitness. Fitness differences would be seen in performance differences between hatchlings, potentially indicating differing abilities to survive. Examples of performance that indicate survival ability of hatchlings include the Righting Response (Steyermark and Spotila 2001, Delmas et al. 2007), crawling speed (Foley 1998; Nicar and Weber

2002), swimming speed, and overall swimming activity (Salmon and Wyneken 1987;

Wyneken and Salmon 1992). The Righting Response refers to the ability of a hatchling to turn itself over after being placed upside down on its carapace (Delmas et al. 2007).

The other locomotor functions of crawling and swimming are considered variables indicating hatchlings’ ability to reach their offshore nursery grounds by escaping the predator-laden beach and near-shore predator-rich waters (Gyuris 1994).

9 An experiment testing these variables could also address the question behind the loggerhead sex ratio discrepancies if outside of the transitional range and near the upper limit of incubation temperature, female hatchlings were seen to have differing performance, such as crawling slower, or more variable behavior compared to hatchlings at other temperatures. This would imply that high beach temperatures are negating the potential benefit of TSD, or that being born a female in these conditions no longer conveys a fitness benefit because the temperature is simply too hot. Based on this idea, we would expect males born at the colder limit of incubation temperatures to show decreased performance ability as well. For the upper limit of incubation temperatures, this result could explain why, despite heavily female-biased primary sex ratios, the juvenile sex ratio of Atlantic loggerheads is not becoming more female-biased and population levels are not recovering as expected.

MATERIALS AND METHODS

Egg Collection and Incubation Protocol

Laboratory experiments were conducted to test for an effect of incubation temperature and sex on performance of 68 loggerhead hatchlings in 2011. The hatchlings came from two loggerhead clutches collected the morning after they were laid on North

Island, South Carolina (Yawkey Wildlife Center) on June 20th, 2011 (Clutch A, 89 eggs),

June 22nd, 2011 (Clutch B, 109 eggs). Volunteers monitoring the beach for the South

Carolina Department of Natural Resources (SC-DNR) collected the eggs. Eggs were placed carefully into buckets and insulated with sand from the nest site for transport on foot to a boat, and then transferred by car. Each clutch was transported in a single bucket. The clutches were driven to the Center for Marine Sciences and Technology

10 (CMAST) in Morehead City, North Carolina, where Dr. Matthew Godfrey supervised the incubation set-up. Total time between laying and placement in incubators was about 18 hours, assuming that nests were laid around midnight (Miller et al. 2003).

The eggs from each clutch were distributed randomly between five incubators set to temperatures ranging from 27°C – 32.5°C. Eggs were placed on three shelves, and the temperature varied slightly between these shelves. The most frequent temperatures maintained throughout the course of incubation for each shelf were 27.1°C, 28.3°C,

28.5°C, 28.7°C, 29.1°C, 29.2°C, 29.3°C, 29.6°C, 30.0°C, 31.2°C, 32.3°C, and 32.5°C.

Based on these temperatures, experimental temperatures were grouped categorically within half-degree intervals. Other than one faulty incubator, temperature variations per shelf and for each incubator overall never exceeded a half-degree. The experimental temperature groups tested were 27, 28, 28.5, 29, 29.5, 30, 31, 32, and 32.5°C.

Because of the slight variation in temperature across shelves within the incubators, we maintained an approximately equal number of eggs from each clutch on each shelf of the incubators. For each clutch, there were five to eight eggs per shelf, with a total of ten to sixteen eggs per shelf. Each egg was incubated individually in a labeled white plastic cup, sitting on a synthetic sponge, surrounded by moist vermiculite (see Mrosovsky 1988 for details). An Hg thermometer immersed in glycerin was located on each shelf of the incubators, and the temperature of each shelf was recorded daily. Humidity was also recorded daily. At days 15-17 and 38-40 of incubation, ~65 mL of DI water were added to the vermiculite of each egg.

Hatching Protocol

Once pipping started, eggs were checked twice daily, and the date and time of both

11 pipping and hatching were recorded. Pipping was defined as a break or cut in the shell.

A turtle was considered “hatched” when it had its head and at least one flipper out of the egg (Godfrey and Mrosovsky 1997). Hatchlings were left untouched for 24 hours post- hatching. After 24 hours, the hatchling was removed from its egg, measured and weighed, labeled, and then placed in an individual holding container in a separate incubator.

Curved carapace length and curved carapace width were recorded using a measuring tape to the nearest tenth of a centimeter, and weight was recorded to the nearest hundredth of a gram. Labeling was done on the carapace scutes with non-toxic whiteout paint. The holding containers were plastic cups surrounded in tinfoil to keep the hatchlings in the dark, and the holding incubator was set to 29.5°C. This simulated the natural warm and dark environment in a nest that a hatchling experiences before emerging (Wyneken pers. comm.). Hatchlings remained in the containers for a minimum of 72 hours and a maximum of 96 hours to simulate the actual course of hatching and emergence from a nest (Godfrey and Mrosovsky 1997; Miller et al. 2003). This three to four day period corresponds to the time a hatchling spends internalizing residual yolk

(Godfrey and Mrosovsky 1997). After this time period, hatchlings were transported in their individual holding containers by car 33.8 km to the Marine Aquaculture Research

Center (MARC) facility in Marshallburg, North Carolina. Experiments took place from

August 7th, 2011 – September 6th, 2011.

Experimental Protocol

Experiments started two hours after hatchlings were removed from the holding incubator. Hatchlings were tested one by one, and up to ten hatchlings could be tested

12 per trial. The order of hatchlings tested in a trial was random. All tests were conducted in one room in the MARC facility that was climate controlled, and air temperature in the room ranged from 24°C - 26°C to approximate nighttime air temperatures. There were

10 experimental trials grouped based on hatching order. The trials roughly corresponded to incubation temperature groupings, due to the fact that incubation temperature determines incubation duration (Foley 1998). For each hatchling, three tests were conducted in this order: Righting Response, Crawl, and Swim test. Before starting the

Righting Response test, non-toxic whiteout dots were placed on the upper surface ends of the hatchlings two front flippers to facilitate the video analysis of the Swim test.

Righting Response and Crawl Test

The first test performed was the Righting Response. A hatchling was placed upside down on its carapace at one end of a 0.3m x 2.44m runway. As soon as the hatchling was set down, a stopwatch was activated and the amount of time it took the hatchling to flip itself over was recorded in seconds. This was repeated four times for each hatchling. If after two minutes the hatchling could not turn itself over, this was noted and the hatchling was manually turned over.

Immediately after the hatchling righted itself for the fourth time, the Crawl test commenced. The hatchling would crawl from the top of the runway to the other end, where an attached night-light served as the only light source for hatchling orientation.

On a natural beach, hatchlings crawl towards the brightest light source, which is the ocean when there are no artificial lights on a beach (Witherington 1992). The runway was at an angle of 25° to simulate the inclination of a natural beach, and lined with sand- colored carpet. Once the hatchling reached a line drawn at the end of the runway, the

13 time was recorded in seconds. This process was repeated three times for a total distance crawled of 7.32 meters, and average speed calculated overall and for each 2.44m section

(modified from Foley 1998). There were no breaks between the three crawls to simulate a hatchling crawling continuously down a beach. The hatchling was then immediately moved to start the Swim test, simulating what happens post-emergence as hatchlings crawl down the beach to the ocean.

Swim Test

The methodology for the Swim test was derived from Salmon and Wyneken

(1987) and takes advantage of the hatchling swimming frenzy that occurs continuously for the first approximately 24 hours in the ocean. Each hatchling was fitted with a black lycra harness closed with Velcro that did not impede motion. This harness was attached to the center of a monofilament line tied across the center of a bucket. This served as a tether, allowing the hatchling to swim in any direction in the bucket without touching the bottom or sides, thereby simulating an open-water oceanic environment (Salmon and

Wyneken 1987). Each hatchling swam in its own circular 64.35L bucket of diameter

54cm containing filtered seawater maintained at ~26°C with a flow through water system.

As soon as the first hatchling was hooked into its harness, two Axis P1346 video surveillance cameras mounted approximately 2.1 meters above the buckets started recording at 30 frames per second. Each camera recorded five buckets at a time, allowing up to ten hatchlings to be tested simultaneously per trial. Recordings lasted for

24 hours after the last hatchling started swimming. Night-lights mounted on the wall opposite the buckets, at 15° and 30° above the hatchlings, served as an orientation light

14 cue for the hatchlings. The only other light sources were three high-powered infrared lamps spread throughout the room to enhance the video recordings. Infrared lamps were used because loggerheads do not respond to infrared wavelengths (Lohmann et al. 1997).

The experiment started with 12 hours of darkness to simulate natural events, that hatchlings typically enter the ocean at night after emerging from the nest. After 12 hours, the overhead fluorescent lights in the room turned on for 12 hours of “daytime”. After this 24-hour period, the hatchlings were removed from their harnesses, placed back into their holding containers, and returned to CMAST.

Video Analysis

I conducted video analysis of the Swim test by watching the recordings at 4x speed. A preliminary viewing of the videos allowed for a categorical index, or ethogram, to be developed describing the swimming behaviors of the hatchlings. These categories and the times of their occurrence were then recorded as the videos were watched. Any changes in swimming behavior that lasted longer than 30 seconds were recorded. The four types of swimming behaviors observed, totaling 98% of the 24 hours for all hatchlings, were called PowerStroke (PS), Rear-Flipper Kick (RFK), Mixed Rest (MR), and Rest (R). Breaths that lasted longer than 15 seconds were also recorded, along with any interference to the hatchlings (for example having to re-attach a harness).

Descriptions of PS and RFK behavior can be found in Salmon and Wyneken (1987). MR was described as hatchlings doing a mix of PS, RFK, breathing, and/or R behaviors, with all of these behaviors lasting less than 30 seconds. Rest (R) was simply when hatchlings were not moving any flippers. Overall swimming activity was quantified as [%PS] +

[%RFK] over 24 hours, because these were the two swim behaviors when hatchlings

15 exhibited oriented swimming. We also noted when hatchlings made their first major switch from PS to another form of swimming, which I defined as equal to or greater than

30 minutes of one hour, a sign of possible tiring during the frenzy.

The analysis of swimming speed was conducted over the first 4 hours of swimming, due to the fact that this was the time period during which all hatchlings were

PowerStroking. Flipper stroke rate frequency was calculated by counting strokes for one minute every half hour. This resulted in a nine-point time series that was used in the analysis. A proxy for distance was calculated using total percent time PS for each hatchling (%PS x stroke rate frequency), allowing generation of a “Distance” Index.

Statistical Analysis

Analyses were performed using “R” (R: A Language and Environment for

Statistical Computing, R Core Team, 2012). Percent hatched and percent survivorship binary data were analyzed using a logistic regression, with incubation temperature as a factor. Two-way analyses of variance (ANOVA) were performed on the Righting

Response, crawl, and swim data, using both incubation temperature and clutch as factors.

In cases where residuals and/or variances were not normally distributed, log- transformations were performed before running the ANOVA. An analysis of covariance

(ANCOVA) was used to analyze interactions between swimming speed, incubation temperature, and body mass. Change in swim speed over time was analyzed with multiple regression. Finally, we assembled a composite rank value for each hatchling by summing the ranks of all statistically significant performance variables. Low ranks were assigned to decreased locomotor abilities; for example, the hatchling that crawled the slowest would have a rank of 1, with ranks then increasing with crawl speed. Several

16 curves were fit to the data, and the best fit was assigned based on an Akaike Information

Criterion (AIC) value comparison and Akaike weight calculation.

Due to the high number of tests performed on each hatchling, I looked for correlations using pair-wise comparisons of performance variables, specifically between the three groups of tests performed (Righting Response, crawl, and swim test). There were no statistically significant correlations present, and the highest R2 value seen was

0.22. Thus I treated the performance tests as independent of each other, and statistical significance for Righting Response and crawl was set to p =/< 0.05. However, for the swim test, I performed a Sidak-Bonferroni Correction to account for the six variables tested on the swim data. The result of this correction was that statistical significance was set to p =/< 0.0085.

RESULTS

Percent Hatched and Survivorship:

Overall, there was a 62% hatch success rate (n=88/143), and 41% survivorship

(n=59/143). The difference in these values came from hatchlings that hatched but then died before starting a trial after the three-day holding period. There was one incubator that increased its temperature ~2°C in the middle of the incubation period; these hatchlings were excluded from all analyses. Percent hatched at different temperatures ranged from 33.3% at 32°C, to 69.2% at 29°C (Fig. 2). Incubation temperature had no significant effect on the number of hatched eggs (Fig. 2).

17 Incubation temperature did affect percent survivorship, specifically at 32°C and

32.5°C where no hatchlings survived to be tested (G = 36.61, p<0.0001, Fig. 2). Of the temperatures where hatchlings survived to be tested, percent survivorship ranged from

30.7% at 27°C to 61.5% at 29°C (Fig. 2). There was no significant difference in percent hatched between Clutch A and Clutch B; however, Clutch A had higher survivorship overall than Clutch B (F2,1=4.8476, p<0.05, Fig. 3,). Clutch A’s survivorship was 50%

(n=33/66), and Clutch B was 34% (n=26/77).

Figure 2: Overall percent of hatched eggs and percent of hatchlings that survived; darker bars are the percent hatched and lighter bars percent survivorship. Percent hatched and percent survivorship were maximized at 29°C, with minimums seen at the extremes of incubation temperature. A significant effect of incubation temperature is seen at 32 and 32.5°C, with stars denoting statistical significance.

18

Figure 3: Percent survivorship for Clutch A and Clutch B, showing Clutch A to have a higher survivorship at all incubation temperatures except 32°C and 32.5°C. The darker bars are Clutch A, lighter bars Clutch B.

Incubation Temperature Effect on Performance Variables:

Incubation temperature had a significant effect on hatchling curved carapace length (F2,6= 2.3963, p<0.05), mean Righting Response time (F2,6= 2.9692, p=0.01, Fig.

4), crawl speed for the third crawl (F2,6= 3.9792, p<0.01, Fig. 5), change in crawl speed

(F2,6= 7.0097, p=0.01, Fig. 6), overall swimming activity (F2,6= 8.0222, p<0.0001, Fig.

7), and %MR (F2,6= 4.5898, p=0.001). Hatchling body mass ranged from 16.00 g to

22.16 g, and there was no detectable effect of incubation temperature on body mass.

19 Curved carapace length (CCL) ranged from 4.2 cm – 4.9 cm. Hatchlings were smallest at

27°C and 31°C, and the largest at 29°C, 29.5°C, and 30°C.

Mean Righting Response time showed a similar pattern to CCL, with hatchlings from 27°C and 31°C taking significantly longer to right themselves than hatchlings from the middle incubation temperatures (Fig. 4). Righting Response time ranged from 0.65 s to 120 s (maximum time before hatchling was manually turned over). Mean Righting

Response time per individual ranged from 1.50 s to 90.63 s.

All crawl speeds ranged from 0.031 m/s to 0.080 m/s. Incubation temperature had no statistically significant effect on the first or second crawl speeds. During the third crawl (C3) incubation temperature did have a detectable effect, and speeds ranged from

0.040 m/s to 0.080 m/s. Hatchlings were significantly slower at 27°C (Fig. 5). Similarly, hatchlings from 27°C seemed to fatigue quicker by slowing down on average between the first and third crawl (C3-C1; Fig. 6). At all other temperature groupings, hatchlings sped up, demonstrating a linear increase in change in speed (Fig. 6). Change in speed ranged from slowing down by 0.015 m/s to speeding up by 0.048 m/s.

Overall swimming activity ranged from 31.45% active to 98.31% active. A clear pattern emerged, with most hatchlings spending about 80% to 90% of their time actively swimming; however, hatchlings at 27°C and 28°C were significantly less active (Fig. 7).

It should be noted that at 28°C, n=2, compared to a minimum of n=4 for all other incubation temperatures. With 28°C excluded from the analysis, incubation temperature maintains a significant effect (F2,5= 6.6241, p<0.001).

20 Incubation temperature had no effect on the other swim test performance variables, including %PS, %RFK, %R, or the first switch from PS to another form of swimming. There was also no effect on swimming speed over the first four hours of PS, nor did incubation temperature affect the “Distance” Index based on speed and %PS.

Figure 4: Mean Righting Response time ± SE, demonstrating that hatchlings incubated at 27°C and 31°C take significantly longer to right themselves than hatchlings from the middle range of incubation temperatures (F2,6= 2.9692, p=0.01). Hatchlings from 31°C also took significantly longer to right themselves than hatchlings from 27°C. [Sample size: 27°C n=5; 28°C n=5; 28.5°C n=12; 29°C n=15; 29.5 n=11; 30°C n=6; 31 n=4]

21

Figure 5: Mean crawl speed for the third crawl (C3) ± S.E.M., with hatchlings from

27°C being significantly slower than all other hatchlings (F2,6= 3.9792, p<0.01). This trend is seen for C1 and C2 as well, but the effect of incubation temperature is not detectable statistically. [Sample size: 27°C n=5; 28°C n=5; 28.5°C n=12; 29°C n=15;

29.5 n=11; 30°C n=6; 31 n=3]

22

Figure 6: Incubation temperature’s effect on mean change in crawl speed (C3-C1) ±

S.E.M. On average, hatchlings from 27°C slowed down while all other hatchlings sped up between crawls (F2,6= 7.0097, p=0.01). A linear relationship between these data is presented (R2 = 0.87). [Sample size: 27°C n=5; 28°C n=5; 28.5°C n=12; 29°C n=15;

29.5 n=11; 30°C n=6; 31 n=3]

23 Figure 7: The effect of incubation temperature on overall swimming activity ± S.E.M.

Hatchlings from 27°C and 28°C were significantly less active than those from all other incubation temperatures (F2,6= 8.0222, p<0.0001). Incubation temperature maintained a significant effect on hatchlings from 27°C when the two hatchlings incubated at 28°C were removed from the analysis. [Sample size: 27°C n=4; 28°C n=2; 28.5°C n=7; 29°C n=14; 29.5 n=10; 30°C n=6; 31 n=4]

Composite Rank:

There was no significant difference between AIC fit values or the nonlinear regression results for composite rank between a linear, quadratic, or 3rd order polynomial function (see Table 1). However, based on the prevalence of a quadratic pattern seen in the performance data results, a quadratic function was chosen for its potential predictive

24 power to estimate composite rank on new data (F1,44 = 4.671, p=0.01, Fig. 9). The

Akaike weight, or relative weight of evidence for each model, is also highest with the quadratic fit (Table 1). The composite rank value was obtained by summing the ranks of each RR, C3 speed, change in crawl speed (C1-C3), overall swimming activity, and

%MR value. These variables were considered independent measures of performance because there were no significant pair-wise relationships between each of these variables, with only a slight correlation between C3 speed and overall swim activity (R2=0.2).

Hatchlings from 27°C retained the lowest rank, with the curve leveling off around 30°C

(Fig. 9).

Table 1: Summary of the AIC values and fit of models applied to composite rank data.

Although there is no significant difference between AIC values, the quadratic function was seen to have the best fit based on relative Akaike weights. Furthermore, a quadratic curve better characterizes the data based on each individual performance result.

MODELS OF FIT AIC VALUE AKAIKE WEIGHT Linear 491.314 0.387 Quadratic 491.144 0.421 3rd Order Polynomial 492.708 0.193

25 Figure 9: The composite rank for each incubation temperature ± S.E.M., fit with a quadratic curve through the means (y = -7.256x2 + 439.13x - 6555.2 ; R² = 0.70).

Incubation temperature maintained a significant effect on the composite rank of hatchlings from 27°C. [Sample size: 27°C n=4; 28°C n=2; 28.5°C n=7; 29°C n=14;

29.5 n=10; 30°C n=6; 31 n=4]

Clutch Effect on Performance Variables

Clutch of origin had a significant effect on Righting Response time (F2,1= 8.7614, p<0.01), C1 speed (F2,1= 4.3602, p<0.05) and change in crawl speed (F2,1= 3.4830, p<

0.05). There were slight effects of clutch on overall swim activity (F2,1= 4.6860, p=0.038), first switch from PS (F2,1= 6.0500, p=0.019), and %MR (F2,1= 3.8751, p=0.057). These results are summarized in Table 2.

26 Table 2: Summary of all results. Bold variables in the left-hand column indicate where there are figures presenting the result. Bold test statistics and p-values indicate a significant effect of either incubation temperature or clutch on the performance variable.

Items in italics indicate a slight effect of clutch on the performance variable that did not meet the significance requirement after performing a Sidak-Bonferroni correction.

TEMPERATURE CLUTCH Percent Hatched G=11.999, p=0.21 F = 0.5582, p=0.81 2,1 Percent Survivorship G=36.61, p<0.0001 F2,1= 4.8476, p<0.05 Mass F = 0.5097, p=0.80 F = 2.0071, p=0.16 2,6 2,1 Curved Carapace Length F2,6= 2.3963, p<0.05 F2,1= 0.8227, p=0.37

Fulton’s K value F2,6= 2.8245, p=0.01 F2,1= 0.9353, p=0.34

Righting Response F2,6= 2.9692, p=0.01 F2,1= 8.7614, p<0.01 Crawl 1 F = 0.4631, p=0.83 F = 4.3602, p<0.05 2,6 2,1 Crawl 2 F2,6= 1.7682, p=0.13 F2,1= 0.6609, p=0.52 Crawl 3 F = 3.9792, p<0.01 F = 0.1463, p=0.86 2,6 2,1 Change in Crawl Speed F2,6= 7.0097, p=0.01 F2,1= 3.4830, p< 0.05 Overall Swim Activity F = 8.0222, p<0.0001 F = 4.6860, p=0.038 2,6 2,1 %Time PowerStroke F2,6= 1.8719, p=0.11 F2,1= 0.0161, p=0.90 %Time Rear Flipper Kick F = 0.9854, p=0.45 F = 3.2531, p=0.080 2,6 2,1 %Time Mixed Rest F2,6= 4.5898, p=0.001 F2,1= 3.8751, p=0.057

%Time Rest F2,6= 3.1064, p=0.015 F2,1= 0.8238, p=0.37

Switch from PowerStroke F2,6= 0.2098, p=0.96 F2,1= 6.0500, p=0.019

Average Swim Speed F3,6= 0.8867, p=0.52 F3,1= 0.1899, p=0.67

Approx. Distance Index F3,6= 1.2621, p=0.31 F3,1= 0.0726, p=0.79 Change in Swim Speed mult.R2=0.089, p=0.22 mult.R2=0.089, p=0.22

Swim Speed Hour 1 F2,6= 1.1115, p=0.37 F2,1= 1.8871, p=0.178

27 DISCUSSION

Using a controlled laboratory experiment manipulating incubation temperature, we observed an effect of incubation temperature on hatchling loggerhead performance.

In general, turtles incubated at 27°C showed a decreased locomotor ability; specifically they took longer to right themselves, crawled slower, slowed down while crawling, and swam for a shorter amount of time. Turtles incubated at 31°C took longer to right themselves but showed high locomotor ability in other tests. In the literature, hatchling turtles with decreased locomotor abilities have typically been interpreted to have a reduced chance of survival, and thus a lower overall fitness (Janzen 1993; Gyuris 1994).

Hatch Success and Survivorship

Both percent hatched and survivorship of hatchlings peaked at 29°C, in the middle of the incubation temperature range for loggerheads (Fig. 2). With 29°C producing an approximately equal number of male and female hatchlings, this result supports Fisher’s (1930) theory that a 1:1 male to female sex ratio is desirable.

Survivorship decreased both towards the hot and cold ends of incubation temperature further agreeing with Fisher (1930); however, we did not anticipate that no hatchlings would survive at 32°C and 32.5°C (Fig. 2). Other studies have reported decreased hatch success at the hot and cold temperature extremes (Michelli-Campbell et al. 2011; Janzen

1993).

The overall percent hatched at 62% was lower than expected based on a previous study incubating loggerhead eggs from Florida, Georgia, and North Carolina (Mrosovsky

1988), where percent hatched ranged from 72 – 96%. It is possible that the egg transport from North Island, South Carolina to Morehead City, North Carolina affected

28 development of the eggs. However great care was taken during transport; once the eggs were removed from the beach, they were transported by boat back to land, where wave action could have disturbed the embryos, and they were also transferred between three vehicles. Previous studies have maintained a high hatch success following a lengthy transport (e.g. from Brazil to Canada, hatch success 93.4%; Marcovaldi et al. 1997), so another possibility is that a bacterial or fungal infection affected hatching success.

Performance + Incubation Temperature

As predicted, decreased locomotor ability was seen for hatchlings incubated at the extremes of incubation temperature (Fig. 9). This was a uniform decrease at all performance measures for animals from 27°C, indicating that suboptimal hatchlings are produced below a certain incubation temperature threshold. These results match controlled incubation studies conducted on green sea turtle hatchlings (Burgess et al.

2006; Booth et al. 2004). Poor performance from animals incubated at controlled low temperatures has also been seen for freshwater turtle hatchlings, specifically the Ouachita map turtle (Graptemys ouachitensis), the smooth softshell turtle (Apalone mutica), and the Brisbane river turtle (Emydura signata), which also showed a decreased growth rate

(Freedberg et al. 2004, Doody 1999; Booth et al. 2004). For the Ouachita map turtle incubated at 25°C, the slower righting response was still observed after one year

(Freedberg et al. 2004).

At the warmer end of incubation temperatures tested, hatchlings from 31°C were seen to have a decreased ability to right themselves (Fig. 4). However, they maintained locomotor ability while crawling and swimming. This suggests that the upper threshold for suboptimal loggerhead hatchlings begins around 31°C; however, the lack of

29 survivorship at 32°C and above prevented us from testing performance of these animals.

The Mary River turtle (Elusor macrurus) showed decreased performance ability when incubated at a controlled elevated temperature (32°C); specifically these animals took 30 times longer to right themselves, swam for less amount of time, and exhibited less stroke force output (Micheli-Campbell et al. 2011). No controlled studies on loggerheads have tested hatchlings born at 32°C, but based on the hatchlings’ overall weakened state seen in their inability to survive past hatching, we expect to see similar results.

The Righting Response test has not commonly been used on sea turtle hatchlings

(one study, Hosier et al. 1981), although it has been frequently used on freshwater turtle hatchlings to estimate fitness (e.g. Hutchinson and Kosh 1965, Steyermark and Spotila

2001, Freedberg et al. 2004, Delmas et al. 2007, Micheli-Campbell et al. 2011). It should be noted that the Righting Response has not always correlated with incubation temperature for all freshwater turtle species, for example in the snapping turtle (Chelydra serpentina) (Steyermark and Spotila 2001). The Righting Response is most likely not an exact fitness estimate; however, in our study this simple and fast test showed a significant effect of incubation temperature on loggerhead hatchlings, and it is easily replicable both in laboratories and in the field. Furthermore, as Hosier et al. (1981) remark, features on a beach can cause hatchlings to be turned upside down onto their carapaces, so the ability to flip back over could be critical to surviving the crawl to the ocean. Using the Righting

Response test more frequently on sea turtle hatchlings could facilitate looking at effects of incubation temperature in more natural settings. We propose that this test should be incorporated into additional hatchling studies.

30 The Composite Rank measure is a novel way to estimate overall hatchling fitness based on the performance measures tested in this study. This type of analysis can be categorized as a “Performance Index”, and provides a uniform way to compare hatchlings using multiple correlates of survival (Harbour 2009). Although different curves fit equally well based on AIC values, the quadratic function has the highest Akaike weight and best accounts for the other performance results and our ecological understanding of incubation temperature effects on hatchling performance. In other words, the shape of the curve for a linear or 3rd order polynomial function does not reflect the effects of incubation temperature on performance observed for this species, whereas a quadratic curve agrees with trends observed in each separate performance variable. If more data can be collected on hatchlings in upcoming years, we may be able to fill in certain temperature gaps, bolster sample sizes, and even use the different functions to predict new composite values. Comparing these to the calculated composite values will further elucidate which curve appropriately fits the data. Our prediction is that any data that could be collected on hatchlings incubated at 32°C would have a similar composite value to animals incubated at 27°C. Furthermore, increasing the sample size at all temperatures, especially 28°C, could create a better fit for the quadratic function, or on the other hand illustrate the need to fit a different model. It is important to note that the low composite values seen for hatchlings incubated at 28°C resulted from their poor swim performance. Hatchlings incubated at 28°C showed increased performance ability in other areas on par with animals incubated at 28.5 - 31°C, but only two hatchlings were used in the swim test due to a loss of power in the lab from Hurricane Irene. With a

31 sample size of more than two animals, we may see an increased ability to swim and resultant shift in the Composite Rank measure.

Implications

Based on the hatch success, survivorship, and performance results, there seems to be an optimal range of hatchling loggerhead production at incubation temperatures from approximately 28.5 - 31°C. The decreased performance ability seen in hatchlings incubated at 27°C could indicate why loggerheads do not frequently nest north of North

Carolina. It has been suggested for snapping turtles (Chelydra serpentina) that incubation conditions limit the species range, and the poor condition of loggerhead hatchlings born at these low temperatures agrees with this theory (Bobyn and Brooks

1994).

With loggerhead nesting occurring from Florida to North Carolina, the incubation temperature range 28.5 - 31°C is seen for a significant amount of time only north of

Florida (except earlier in the summer, at the beginning of the nesting period) (Layton

2011). Even on South Carolina beaches, towards the northern end of loggerhead nesting, nest temperatures reach 32°C and above during the summer (Koepfler et al. 2008). On

Florida beaches on the Atlantic coast, sand temperatures are frequently above 31°C, sometimes being maintained at or above 32°C for about a week at a time (Layton 2011, data span years 2004-2009). With Atlantic coast Florida beaches accounting for about

90% of loggerhead hatchling production in the Northern Atlantic, it is feasible that many of these beaches are producing millions of hatchlings that perform sub-optimally.

Furthermore, with the onset of climate change and warming sand temperatures along the

32 Atlantic coast, this could result in an even greater production of suboptimal hatchlings, and an increase in nest mortalities from hot incubation temperatures. These effects will not facilitate the recovery of the Northern Atlantic loggerhead and could already be affecting the population.

However, we can only suggest this idea, and similar performance testing should be done on hatchlings born from Florida nesting females to test this hypothesis and see if there is geographic variation in performance ability. It will also be important to quantify hatching success and survivorship in the field, since these numbers could differ from our laboratory incubation results. The two clutches tested in this study were from South

Carolina nesting females, and it is possible that there is heritable variation resulting in a lowering of the upper incubation temperature limit for these eggs. In the literature, incubation temperatures higher than ~33°C are listed as lethal for loggerhead eggs, but the results of this study suggest that maintaining a nest temperature of 32°C drastically decreases hatch success and survivorship for South Carolina loggerheads. It would be interesting to see if there is a difference in this upper limit between Florida-originated hatchlings and the Northern subpopulation.

Beyond Controlled Incubation

From a conservation standpoint, a limitation of this study was the use of controlled incubation temperatures, making it difficult to extend our findings to hatchlings born in a natural incubation environment. However this experiment was an important first step for the loggerhead to better understand the basic effect incubation temperature has on hatchling performance. In this study we could only clearly address this question by controlling all variables except for incubation temperature. Shine et al.

33 (2006) demonstrated that the same effects of incubation temperature on phenotype and performance observed in the laboratory for hatchling lizards (Bassiana duperreyi) were also observed in the field. This was also the case for the Mary River turtle, when incubation temperatures were both controlled and fluctuated in a lab (Micheli-Campbell et al. 2012). Thus it can be argued that the results of our controlled experiment should be applicable to hatchlings born naturally, and we still suggest that it is important to build off of these results and investigate hatchlings’ performance abilities under fluctuating incubation temperature regimes.

For green sea turtle hatchlings, studies either based in the field or that included fluctuating incubation temperatures in the lab have observed an increase in variation of performance ability. Booth and Evans (2011) detected only small differences in mean swim thrust between hatchlings incubated in natural warm and cool nests. Similarly,

Ischer et al. (2009) observed significant amounts of variation within one nest both in crawl speed and swim thrust. Other studies showed an increase in locomotor performance with an increase in daily temperature fluctuation, for the skink (Bassiana duperreyi), the Keelback snake (Tropidonophis mairii) and the smooth softshell turtle

(Shine and Harlow 1996, Webb et al. 2001, Ashmore and Janzen 2003).

We would predict that fluctuating temperatures could increase loggerhead performance ability, but that like the green sea turtle hatchlings, this scenario would also introduce more variation. In our study, hatchlings from the incubator that fluctuated

~2°C showed the most variation for many of the performance measures (data not shown).

This could be explained ecologically, if fluctuating incubation temperatures in the wild produce hatchlings with a wide range of performance abilities. It may severely decrease

34 the survival chances for some, but increase the ability of others to survive, thereby potentially increasing the mother’s lifetime fitness. Furthermore, this could help offset the potentially damaging effects of temperatures at or above 32°C.

Future studies testing loggerhead hatchling performance could start by continuing laboratory incubations to control certain variables, but build in fluctuating incubation temperatures; then based on these results, hatchlings should be tested from natural nests.

The results of these studies, building off of the results from our controlled experiments, would provide a more complete picture of the natural effects of incubation temperature on the early life history of the loggerhead sea turtle. In the face of climate change and diminishing sea turtle populations, this information could be critical to inform management of nesting beaches in coming years.

Facing Climate Change

Assuming the results of our study are applicable to natural nesting scenarios, it is worth discussing how the Northwest Atlantic loggerhead will adapt, or not be able to adapt, to oncoming climate change. Climate change was recognized as a potential problem for sea turtles in the 1980’s (Mrosovksky 1984, Davenport 1989), and recently much research has focused on this field (reviews in Hamann et al. 2007 and Hawkes et al.

2009). Over the last 100 years, global average surface temperatures have increased

0.8°C, and future surface increases of 2-3°C are expected by 2100 (Hansen et al. 2006).

As stated in the Introduction section of this paper, these rising temperatures could already be affecting sea turtle populations globally. We have illustrated in this study the potential on-going production of millions of suboptimal loggerhead hatchlings as well as potential hatchling mortality coming off of Florida beaches. If this scenario holds true, it creates

35 dire outcomes for the loggerhead in the face of increasing nest temperatures. This problem is exacerbated by loggerhead life history, specifically the late time to maturity, long-lived nature of the species, and natal homing of nesting females (Hamann et al.

2007, Hamann et al. 2010). In theory, a female loggerhead that starts nesting in 2012 may continue to be reproductively active for 20 – 50 years. This means all of the female turtles nesting on Florida beaches will continue to nest there while sand temperatures rise, leading to further decreases in the loggerhead population, and increasing the female to male sex ratio of the population.

It is possible that the loggerhead will be able to shift its nesting range north, allowing the production of more male hatchlings and potentially increasing hatchling survival. However there is no clear evidence that this is already occurring for the loggerhead, and due to the life history characteristics described above, the shift may not occur rapidly enough to recover population numbers. Another adaptational response from loggerheads would be to nest earlier, and hence the nests would experience lower sand temperatures. Evidence supporting this has been controversial, with some studies reporting an earlier onset of nesting (e.g. Pike et al. 2006, Weishampel et al. 2004) while other have not (e.g. Hawkes et al. 2007, Pike 2009). A more recent thesis study by

Bowers and Crowder (2010) reported a temporal shift in loggerhead nesting when the effect of the North Atlantic Oscillation is taken into account, suggesting a behavioral response to climate variability. Overall, it is unknown if these responses to climate changes by loggerheads will be sufficient to maintain population levels in the future.

If we continue to see a decrease in nesting female numbers and a seeming inability of the loggerhead to fully recover its population levels, there are management

36 actions that could be taken. These actions will have to be site-specific, due to varying degrees of monitoring, funding, and management in different regions (Fuentes et al.

2011). Here, I will mention two options that could be used where nesting beach monitoring already occurs, as these are most likely the most feasible sites for any management measures (Fuentes et al. 2011). It is important to remember that funding must be taken into consideration for any management action taken, along with other management options at different stages in the loggerhead life history. Managers must prioritize different management options, as this will be critical to identify and implement the options that will best help the loggerhead in the long-term (Fuentes et al. 2011, Witt et al. 2009).

One of the more obvious and easier management options is to relocate nests after they are laid on high-priority beaches. Nests could be moved into hatcheries, an option that has already been shown to maintain a comparable level of hatch success to in-situ nests (Wyneken et al. 1988, Bimbi 2009). Shade could be easily and cheaply created in the hatcheries to increase hatch success and the number of males produced (e.g. Naro-

Maciel et al. 1999). This option would help loggerheads in the short term, by attempting to increase survival of hatchlings in the first phase of their lives. However, it would be labor-intensive for volunteers because to have an effect on the population, a large number of clutches would have to be moved every year.

An obvious flaw in this management option is the lack of long-term benefit to the loggerhead, as female hatchlings born on these beaches would return to that area to nest, leaving their offspring to experience even higher nest temperatures. A more drastic, expensive, and labor-intensive option could be taken, which would be to transport

37 clutches of eggs north to cooler beaches. This would take advantage of natal homing and artificially accelerate the process of establishing new nesting colonies at higher latitudes

(Hoegh-Guldberg et al. 2008). Preliminary work would have to be conducted to study the habitat and ensure enough similarities between beaches to facilitate survival of the eggs. For example, one might move eggs from the Carolinas to Virginia, where sand composition and moisture content may be similar. Although this would be quite an extensive operation, it has been seen to work for the Kemp’s ridley sea turtle

(Lepidochelys kempii). This species was considered the most endangered sea turtle, and

Kemp’s ridleys only nested on one primary beach, Rancho Nuevo, Mexico. Eggs were relocated to Padre Island, Texas to establish a new nesting colony, and females have since returned to Padre Island (Francis 1978, Shaver 1996).

When deciding if costly management measures are worth the time and financial investment, managers must consider all threats facing the Northwest Atlantic loggerhead.

Climate change is only one such threat to the population, and others, such as fisheries by catch and habitat loss, may hinder population recovery and negate any benefit of efforts made on nesting beaches. The survival of the loggerhead, and indeed of all sea turtle species, will depend on our ability to target all of these threats and develop effective, sustainable management plans for the future. In future years targeting the climate change threat may be our most significant challenge, to ensure the survival of sea turtles as well as many other species around the world.

38 Conclusions

1. Selection seems to favor incubation temperatures near the pivotal sex-determining

temperature of 29°C, based on the survivorship and performance results of the

hatchlings.

2. An incubation temperature of 27°C for hatchlings resulted in a decreased

locomotor ability by all performance measures, perhaps indicating reduced ability

to survive and a lower threshold for incubation temperature. At 31°C, hatchlings

showed a decreased ability to right themselves, indicating the start of the upper

threshold.

3. Incubation at 32°C and above was too hot for hatchlings to survive to be tested.

We assume this corresponds to a reduced viability at these incubation

temperatures, and marks a lower upper limit to incubation temperature than

expected.

4. Global climate change could already be impacting loggerhead hatchling survival

based on measured nest incubation temperatures at the dominant nesting beaches.

Further research should be done to investigate this hypothesis in more natural

settings.

39 Future Directions

Another dimension can be added to this study by looking at the sex of the hatchlings tested, which will be possible after Dr. Godfrey finishes his work on sex determination with these hatchlings. The potential to compare male and female hatchling performance at temperatures where males and females are both produced could help elucidate the adaptive significance of TSD, and addresses an unanswered question focusing on the evolution of TSD that is still seen in animals today. The controversy over TSD is explored below.

TSD Evolution and Persistence

TSD seems to have evolved multiple times in evolutionary history, and there is an ongoing controversy over the existence and persistence of TSD in reptiles. Fisher (1930) predicted that for most species, male and female offspring would be produced in equal number regardless of the mechanism of sex determination. Fisher’s hypothesis has been widely supported in animals with GSD, but it has been suggested that under certain conditions Fisher’s principle does not always hold, particularly in species with environmental sex determination (Bull and Charnov 1989). However, it has been a challenge to determine which conditions or traits of a species undermine Fisher’s principle, resulting in ESD instead of GSD (Crews et al. 1988). Due to environmental influence resulting in skewed population sex ratios, ESD may be an evolutionary relic, but it has persisted as TSD in numerous reptiles including all seven sea turtle species

(Janzen and Paukstis 1991).

Other hypotheses to explain the persistence of TSD in reptiles argue for an adaptive significance of TSD. These include the group-adaptation, inbreeding avoidance,

40 and differential fitness hypotheses (Shine 1999). The differential fitness hypothesis is considered the most robust, and includes within it several sub-hypotheses to explain how incubation temperature could differentially affect offspring fitness (Crews et al. 1988,

Shine 1999).

The differential fitness hypothesis (also called “Charnov-Bull model”) broadly states that ESD would be favored over GSD if one sex were born with an increased fitness in a certain environment (Charnov and Bull 1977, Bull 1981). More specifically for TSD, maternal fitness would be enhanced if incubation temperature differentially affects the fitness of male and female offspring, because TSD would allow the embryo to develop as the sex better suited to the incubation conditions (Shine 1999). Support of the

Charnov-Bull model has been seen in fish, most notably for the Atlantic silverside

(Menidia menidia). The silverside produces females early in the spawning season and males later, resulting in females of a larger size because they experience a longer growing season (Conover and Kynard 1981). Larger female fish have enhanced reproductive success, thereby supporting the predictions of the Charnov-Bull model (Conover 1984).

Using a short-lived lizard, Warner and Shine (2008) provided the first empirical evidence in reptiles that there is an adaptive advantage of TSD, supporting the differential fitness hypothesis. They found that incubation temperatures that naturally produced a certain sex in the jacky dragon (Amphibolurus muricatus) resulted in higher lifetime reproductive output for that particular sex (Warner and Shine 2008).

For sea turtles under the differential fitness hypothesis, it would be predicted that females born in warm temperatures would have an increased fitness compared to males born at that same temperature, and the reverse for cool temperatures (Fig. 10). Another

41 hypothesis depicted in Fig. 10 is that at the extreme incubation temperatures where only one sex is produced, hatchlings could reach optimal fitness for that sex. However, the results of this study indicate that extreme incubation temperatures are not ideal in terms of performance or survivorship for either sex. The other possibility that seems more likely is that close to pivotal temperature may be optimal for both sexes (Fig. 10). This possibility can be explored when the final sex data on these hatchlings becomes available.

It will be interesting to investigate any potential sex-linked performance effects at incubation temperatures near 29°C where both sexes are produced.

Figure 10: Illustration of differential fitness hypothesis as it could apply to sea turtles

(from Valenzuela 2004). Hypothetical relationships between fitness, incubation temperature, and sex ratios that may be observed within species. Sex ratios at different temperatures are shown with the gray curve; male fitness possibilities with the black solid line, and female fitness possibilities with the dashed line. Solid circles and triangles denote the fitness value of the sexes that are produced at the extreme temperatures, while the opposite ends indicate the fitness of the sex that is no longer produced at the extreme temperatures.

42 Acknowledgements

This research was supported by the PADI Foundation (#5040) and the Charleston

Scientific and Cultural Society. Personal support came from the McLeod Frampton

Foundation and Joanna Foundation. All research was performed in accordance with the

Institutional Animal Care and Use Committee (College of Charleston protocol #2011-

005; North Carolina State University protocol #11-078-O). Eggs were collected and transported with permission of the SC-DNR. The help and cooperation of the SC-DNR state staff and volunteers is greatly appreciated, including the SC-DNR Sea Turtle program staff, Jamie Dozier and SC-DNR Yawkey Wildlife Center staff, and the North

Island Sea Turtle Project Team.

I’d like to particularly thank Dr. Matthew Godfrey, Dubose Griffin, Dr. Craig

Harms, Heather Broadhurst, Dr. Marc Turano, Nick Reynolds, and Ryan Kelly for their critical assistance and support. Diversified Computer Solutions in Columbia, SC helped me research and purchase surveillance cameras, and Jeanette Wyneken provided me with invaluable advice on hatchling methodology. I’d like to thank Olivia Ahearn and Ariel

Cristensen for assisting with the video analysis. Finally a big thank you to everybody at

Grice Marine Lab, especially my advisor and committee, Craig Plante, Shelly Brew, and

Pete Meier, as well as my friends and family for their support.

43 Literature Cited

Arendt, M., J. Schwenter, J. Byrd, J.D. Whitaker, L. Liguori, L. Parker, D.W. Owens. 2010. Relative abundance, distribution, and health of sea turtles in near-shore coastal waters of the Southeastern United States. Final project report to NMFS and NOAA.

Arendt, M., J. Byrd, A. Segars, P. Maier, J. Schwenter, D. Burgess, J. Boynton, J. Whitaker. 2009. Examination of local movement and migratory behavior of sea turtles during spring and summer along the Atlantic coast off the Southeastern United States. Final project report to NMFS and NOAA.

Ashmore, G., F. Janzen. 2003. Phenotypic variation in smooth softshell turtles (Apalone mutica) from eggs incubated in constant versus fluctuating temperatures. Oceologica. 134: 182-188.

Bell, C., J. Solomon, J. Blumenthal, T. Austin, G. Ebanks-Petrie, A. Broderick, B. Godley. 2007. Monitoring and conservation of critically reduced marine turtle nesting populations: lessons from the Cayman Islands. Anim Conserv. 10: 39-47.

Bimbi, M. 2009. Effects of relocation and environmental factors on loggerhead sea turtle nests on Cape Island. Masters Thesis, College of Charleston, Charleston, SC.

Bobyn, M., R. Brooks. 1994. Incubation conditions as potential factors limiting the northern distribution of snapping turtles, Chelydra serpentina. Can J Zool. 72: 28- 37. Booth, D. 2006. Influence of incubation temperature on hatchling phenotype in reptiles. Physiol Biochem Zool. 79(2): 274-281. Booth, D. and K. Astill. 2001. Incubation temperature, energy expenditure and hatchling size in the green turtle (Chelonia mydas), a species with temperature-sensitive sex determination. Aust J Zool. 49: 389-396. Booth, D., E. Burgess, J. McCosker, J. Lanyon. 2004. The influence of incubation temperature on post-hatching fitness characteristics of turtles. Int Congr Ser. 1275: 226-233. Bowen, W., A. Bass, S. Chow, S. Bostrom, K. Bjorndal, A. Bolten, et al. 2004. Natal homing in juvenile loggerhead turtles (Caretta caretta). Molecular Ecology. 13: 3797–3808. Bowers, M., L. Crowder. 2010. Phenological shifts in loggerhead sea turtle (Caretta caretta) first nesting dates. Masters project, Nicholas School of the Environment. Duke University, Beaufort, NC.

44 Braun-McNeill, J., S. Epperly, D. Owens, L. Avens, E. Williams, C. Harms. 2007. Seasonal reliability of testosterone radioimmunoassay (RIA) for predicting sex ratios of juvenile loggerhead (Caretta caretta) turtles. Herpetologica. 63: 275- 284. Bull, J. 1981. Sex ratio evolution when fitness varies. Heredity. 46(1): 9-26.

Bull, J. and E. Charnov. 1989. Enigmatic reptilian sex ratios. Evolution. 43:1561-1566. Burgess, E., D. Booth, J. Lanyon. 2006. Swimming performance of hatchling green turtles is affected by incubation temperature. Coral Reefs. 25:341-349. Charnov, E. and J. Bull. 1977. When is sex environmentally determined? Nature. 266:828-830. Chu, C., Booth, D., and Limpus, C. 2008. Estimating the sex ratio of loggerhead turtle hatchlings at Mon Repos rookery (Australia) from next temperatures. Aust. J. Zool. 56, 57–64. Conant, T., P. Dutton, T. Eguchi, S. Epperly, C. Fahy, M. Godfrey, S. MacPherson, E. Possardt, B. Schroeder, J. Seminoff, M. Snover, C. Upite, and B. Witherington. 2009. Loggerhead sea turtle (Caretta caretta) 2009 status review under the U.S. Endangered Species Act. Report of the Loggerhead Biological Review Team to the National Marine Fisheries Service.

Conover, D. 1984. Adaptive significance of temperature-dependent sex determination in a fish. Am Nat. 123: 297-313. Conover, D. 2004. Temperature-dependent sex determination in fishes. Temperature-Dependent Sex Determination in Vertebrates. Eds. N. Valenzuela, V. Lance. Washington: Smithsonian Books. 11-20. Conover, D. and B. Kynard. 1981. Environmental sex determination: interaction of temperature and genotype in a fish. Science. 213: 577-579. Crews, D., J. Bull, A. Billy. 1988. Sex determination and sexual differentiation in reptiles. Handbook of Sexology Vol 6: The Pharmacology and Endocrinology of Sexual Function. Ed. J.M.A. Sitsen. Elsevier Science Publishers: Amsterdam. Davenport, J. 1989. Sea turtles and the greenhouse effect. Brit Herpetol Soc Bull. 29: 11-15. Davenport, J. 1997. Temperature and the life-history strategies of sea turtles. J Therm Biol. 22: 479-488. Deeming, D.C. and M.W.J. Ferguson. 1991. Physiological effects of incubation temperature on embryonic development in reptiles and birds. Egg incubation: its effects on embryonic development in birds and reptiles. Eds. D.C. Deeming, M.W.J. Ferguson. Cambridge: Cambridge University Press. 147-171.

45 DeGregorio, B.A. and Williard, A.S. 2011. Incubation temperatures and metabolic heating of relocated and in situ loggerhead sea turtle (Caretta caretta) nests at a Northern rookery. Chelonian Conserv Biol. 10(1): 54-61. Delgado, C., A. Canario, T. Dellinger. 2008. Sex-ratio of juvenile pelagic loggerheads Caretta caretta off Madeira Island (Portugal), NE Atlantic: a 3-method approach. NOAA Technical Memorandum NMFS SEFSC. 569:207. Delgado, C., A. Canario, T. Dellinger. 2009. Sex ratios of loggerhead sea turtles Caretta caretta during the juvenile pelagic stage. Mar Biol. DOI 10.1007. Delmas, V., E. Baudry, M. Girondot, A. Prevot-Julliard. 2007. The righting response as a fitness index in freshwater turtles. Biol J Linnean Soc. 91: 99-109. Doody, J. 1999. A test of the comparative influences of constant and fluctuating incubation temperatures on phenotypes of hatchling turtles. Chelonian Conserv Biol. 3: 529-531. Fisher, R. 1930. The genetical theory of natural selection. Clarendon Press, Oxford. Foley, A. 1998. The nesting ecology of the loggerhead turtle (Caretta caretta) in the Ten Thousand Islands, Florida. Ph.D. Dissertation, University of South Florida, Tampa, Florida. Francis, K. 1978. Kemp’s ridley sea turtle conservation programs at South Padre Island, Texas and Rancho Nuevo, Tamaulipas, Mexico. Florida Marine Research Publications. 33: 51-52. Freedberg, S., A. Stumpf, M. Ewert, C. Nelson. 2004. Developmental environment has long-lasting effects on behavioural performance in two turtles with environmental sex determination. Evol Ecol Research. 6: 739-747. Fuentes, M., J. Maynard, M. Guinea, I. Bell, P. Werdell, M. Hamann. 2009. Proxy indicators of sand temperature help project impacts of global warming on sea turtles in northern Australia. Endang Species Res. 9: 33-40. Glen, F., A. Broderick, B. Godley, G. Hays. 2003. Incubation environment affects phenotype of naturally incubated green turtle hatchlings. J Mar Biol Assoc U.K. 83: 1183-1186. Glen, F., and Mrosovsky, N. 2004. Antigua revisited: The impact of climate change on sand and nest temperatures at a hawksbill turtle (Eretmochelys imbricata) nesting beach. Glob. Change Biol. 10: 2036–2045. Godfrey, M. and N. Mrosovsky. 1997. Estimating the time between hatching of sea turtles and their emergence from the nest. Chelonian Conserv Biol. 2: 581-585.

Gyuris, E. 1994. The rate of predation by fishes on hatchlings of the green turtle. Coral Reefs. 13: 137-144.

46 Hamann, M., C. Limpus, M. Read. 2007. Vulnerability of marine reptiles in the Great Barrier Reef to climate change. Climate change and the Great Barrier Reef: a vulnerability assessment, Great Barrier Reef Marine Park Authority. Eds. J. Johnson, P. Marshall. 465-496. Hamann, M., Godfrey, M. H., Seminoff, J., Arthur, K., Barata, P., Bjorndal, K., Bolten, A. B., et al. 2010. Global research priorities for sea turtles: informing management and conservation in the 21st century. Endangered Species Research, 11(3): 245-269.

Hansen, J., M. Sato, R. Ruedy, K. Lo, D. Lea, M. Medina-Elizade. 2006. Global temperature change. Proc Natl Acad Sci USA. 103: 14288-14293. Harbour, J. 2009. The Basics of Performance Measurement. 2nd ed. Productivity Press: New York. Hays, G. C., Broderick, A. C., Glen, F., and Godley, B. J. 2003. Climate change and sea turtles: A 150-year reconstruction of incubation temperatures at a major marine turtle rookery. Glob. Change Biol. 9: 642–646. Hawkes, L., A. Broderick, M. Godfrey, B. Godley. 2009. Climate change and marine turtles. Endang Species Res. 7: 137-154.

Hosier, P., M. Kochar, V. Thayer. 1981. Off-road vehicle and pedestrian track effects on the sea approach of hatchling loggerhead turtles. Environmental Conservation. 8(2): 158-161. Hutchinson, V. and R. Kosh. 1965. The Effect of Photoperiod on the Critical Thermal Maxima of Painted Turtles (Chrysemys picta). Herpetologica. 20(4): 233-238.

Ischer, T., K. Ireland, D. Booth. 2009. Locomotion performance of green turtle hatchlings from the Heron Island Rookery, Great Barrier Reef. Marine Biology, 156(7): 1399-1409.

Janzen, F. and G. Paukstis. 1991. A preliminary test of the adaptive significance of environmental sex determination in reptiles. Evolution. 45(2): 435-440. Janzen, F. 1993. The Influence of Incubation Temperature and Family on Eggs, Embryos, and Hatchlings of the Smooth Softshell Turtle (Apalone mutica). Physiological Zoology. 66(3): 349-373.

Hoegh-Guldberg, O., L. Hughes, S. McIntyre, D. Lindenmayer, C. Parmesan, H. Possingham, C. Thomas. 2008. Assisted colonization and rapid climate change. Science. 321: 345-346. Hopkins-Murphy, S. D. Owens, T. Murphy. 2003. Ecology of immature loggerheads on foraging grounds and adults in internesting habitat in the Eastern United States. Loggerhead Sea Turtles. Eds. A. Bolten, B. Witherington. 79-92.

47 Koepfler E., P. Hoffman, S. Dawsey. 2008. Caretta caretta nest temperatures in hatcheries at Cape Romain – determinants of spatial and temporal variability. Poster presentation, 28th Annual Symposium on Sea Turtle Biology and Conservation. Loreto, Baja California Sur, Mexico. Layton, J. 2011. Biological, Ecological, and Conservational Implications of Temperature- Dependent Sex Determination in Sea Turtle Populations. PhD Dissertation. University of Alabama, Birmingham, Alabama.

Limpus, C. 2006. Impacts of climate change on sea turtles: a case study roundtable on migratory species and climate change. CMS COP, Nairobi, November 2005. Lohmann, K., B. Witherington, C. Lohmann, M. Salmon. 1997. Orientation, navigation, and natal beach homing in sea turtles. The Biology of Sea Turtles. eds. P. Lutz and J. Musick, 107-135. CRC Press: Boca Raton.

Marcovaldi, M., M. Godfrey, N. Mrosovsky. 1997. Estimating sex ratios of loggerhead turtles in Brazil from pivotal incubation durations. Canadian Journal of Zoology. 75: 755-770.

Matsuzawa, Y., K. Sato, W. Sakamoto, K. Bjorndal. 2002. Seasonal fluctuations in sand temperature: effects on the incubation period and mortality of loggerhead sea turtle (Caretta caretta) pre-emergent hatchlings in Minabe, Japan. Mar Biol. 140: 639–646.

McClenachan, L., J. Jackson, M. Newman. 2006. Conservation implications of historic sea turtle nesting beach loss. Front Ecol Environ. 4: 290-296. Meylan, A. 1982. Estimation of population size in sea turtles. Biology and conservation of sea turtles. Smithsonian Institution Press, Washington, D.C. 135–138. Micheli-Campbell, M., H. Campbell, R. Cramp, D. Booth, C. Franklin. 2011. Staying cool, keeping strong: incubation temperature affects performance in a freshwater turtle. Journal of Zoology, 285: 266-273.

Micheli-Campbell, M., M. Gordos, H. Campbell, D. Booth, C. Franklin. 2012. The influence of daily temperature fluctuations during incubation upon the phenotype of a freshwater turtle. Journal of Zoology, 1-8.

Miller, J. 1985. Embryology of marine turtles. Biology of the Reptilia. Eds. C. Gans, F. Billett, P. Maderson. 269-328. Miller, J. and C. Limpus. 1981. Incubation period and sexual differentiation in the green turtle Chelonia mydas. Proceedings of the Melbourne herpetological symposium zoological board of Victoria, 1981. Zoological Board of Victoria, Parkville. 66–73.

48 Miller, J., C. Limpus, M. Godfrey. 2003. Nest site selection, oviposition, eggs, development, hatching, and emergence of loggerhead turtles. Loggerhead Sea Turtles. Eds. A. Bolten, B. Witherington. 125-143. Mrosovsky, N., S. Hopkins-Murphy, J. Richardson. 1984. Sex Ratio of Sea Turtles: Seasonal Changes. Advancement Of Science. 225, 4663: 739-741.

Mrosovsky, N. 1984. Editorial. Mar Turtle News. 28: 1-2. Mrosovsky, N. 1988. Pivotal temperatures for loggerhead turtles from northern and southern nesting beaches. Can J Zool. 66: 661-669. Mrosovsky, N. and Provancha, J. 1989. Sex ratio of loggerhead sea turtles on a Florida beach. Can J Zool. 67:2533-2539. Naro-Maciel, E., N. Mrosovsky, M. Marcovaldi. 1999. Thermal profiles of sea turtle hatcheries and nesting areas at Praia do Forte, Brazil. Chelonian Conserv Biol. 3: 407-413. Nicar, E. and L. Weber. 2002. Hatchling fitness of loggerhead sea turtles from nests with different incubation durations. Undergraduate thesis, Warren Wilson College. Norman, J., C. Moritz, C. Limpus. 1994. Mitochondrial DNA control region polymorphisms: genetic markers for ecological studies of marine turtles. Mol Ecol. 3(4): 363-373. O’Steen, S. 1998. Embryonic temperature influences juvenile temperature choice and growth rate in snapping turtles Chelydra serpentina. Journal of Experimental Biology. 449:439-449.

Owens, D. 1997. Hormones in the life history of sea turtles. The Biology of Sea Turtles. Eds. P. Lutz, J. Musick. 315-341. Raynaud, A. and Pieau, C. 1985. Embryonic development of the genital system. Biology of the Reptilia. Eds. C. Gans, F. Billett, P. Maderson. 150-300. Reece, S., A. Broderick, B. Godley, S. West. 2002. The effects of incubation environment, sex and pedigree on the hatchling phenotype in a natural population of loggerhead turtles. Evol Ecol Res. 4: 737-748. Reid, K., D. Margaritoulis, J. Speakman. 2009. Incubation temperature and energy expenditure during development in loggerhead sea turtle embryos. J Exp Mar Biol Ecol. 378: 62-68. Pieau, C., M. Girondot, N. Richard-Mercier, G. Desvages, M. Dorizzi, P. Zaborski. 1994. Temperature sensitivity of sexual differentiation of gonads in the European pond turtle: hormonal involvement. J Exp Zoology. 270, 1:86-94.

49 Pike, D. 2009. Do green turtles modify their nesting seasons in response to environmental temperatures? Chelonian Conserv Biol. 8: 43-47. Pike, D., R. Antworth, J. Stiner. 2006. Earlier nesting contributes to shorter nesting seasons for the loggerhead sea turtle, Caretta caretta. J Herpetol. 40: 91-94. Salmon, M. and J. Wyneken. 1987. Orientation and swimming behavior of hatchling loggeread turtles Caretta caretta during their offshore migration. J Exp Mar Biol Ecol. 109: 137-153. Sandoval, S., V. Munoz-Gomez, J. Gutiérrez, M.A. Porta-Gándara. 2011. Metabolic heat estimation of the sea turtle Lepidochelys olivacea embryos. Journal of Thermal Biology. 36: 138-141.

Shaver, D. 1996. Head-started Kemp’s ridley turtles nest in Texas. Marine Turtle Newsletter. 74: 5-7.

Shine, Richard. 1999. Why is sex determined by nest temperature in many reptiles? TREE. 14(5): 186-189. Shine, R., M. Elphick, P. Harlow. 2006. The influence of natural incubation environments on the phenotypic traits of hatchling lizards. Ecology. 78(8): 2559- 2568.

Schroeder, B., and S. Murphy. 1999. Population surveys (ground and aerial) on nesting beaches. Research and management techniques for the conservation of sea turtles. IUCN/SSC Marine Turtle Specialist Group Publication Number 4. Eds. K. Eckert, K. Bjorndal, F. Abreu-Grobois, M. Donnelly.

Shoop, C. and Kenney, R. 1992. Seasonal distributions and abundances of loggerhead and leatherback sea turtles in waters of the Northeastern United States. Herpetol Monogr. 6: 43-67.

Stabenau, E. K. Stanley, A. Landry. 1996. Sex ratios from stranded sea turtles on the upper Texas coast. J Herpetol. 30: 427-430.

Steyermark, A., and J. Spotila. 2001. Body Temperature and Maternal Identity Affect Snapping Turtle (Chelydra serpentina) Righting Response. Copeia. 2001, 4: 1050-1057.

TEWG (Turtle Expert Working Group). 1998. As assessment of the Kemp's Ridley (Lepidochelys kempii) and Loggerhead (Caretta caretta) sea turtle populations in the western North Atlantic. NOAA Technical Memorandum.

Valenzuela, N. and V.A. Lance. 2004. Temperature-Dependent Sex Determination in Vertebrates. Washington: Smithsonian Books.

50 Warner, D.A., and R. Shine. 2008. The adaptive significance of temperature-dependent sex determination in a reptile. Nature. 451: 566-568.

Webb, J., G. Brown, R. Shine. 2001. Body size, locomotor speed and antipredator behaviour in a tropical snake (Tropidonophis mairii): the influence of incubation environments and genetic factors. Functional Ecology. 15: 561-568.

Weishampel, J., D. Bagley, L. Ehrhart. 2004. Earlier nesting by loggerhead sea turtles following sea surface warming. Glob Chang Biol. 10: 1424-1427.

Wibbels, T. R. Martin, D. Owens, M. Amoss. 1991. Female-biased sex ratio of immature loggerhead sea turtles inhabiting the Atlantic coastal waters of Florida. Can J Zool. 69: 2973-2977.

Wibbels, T., D. Owens, M. Amoss. 1987. Sexing techniques and sex ratios for immature loggerhead sea turtles captured along the Atlantic coast of the United States. Ecology of East Florida sea turtles, NOAA Technial Report. 65-74.

Witherington, B., M. Salmon. 1992. Predation on Loggerhead Turtle Hatchlings after Entering the Sea. Society for the Study of Amphibians and Reptiles. 26(2): 226-228.

Witherington, B., Kubilis, P., Brost, B., & Meylan, A. 2009. Decreasing annual nest counts in a globally important loggerhead sea turtle population. Ecological Society of America. 19(1): 30-54.

Witt, M., L. Hawkes, M. Godfrey, B. Godley, A. Broderick. Predicting the impacts of climate change on a globally distributed species: the case of the loggerhead turtle. Journal of Experimental Biology. 213: 901-911.

Witzell, W., Geis, A., Schmid, J., & Wibbels, T. 2005. Sex ratio of immature Kempʼs ridley turtles (Lepidochelys kempi) from Gullivan Bay, Ten Thousand Islands, south-west Florida. Journal of the Marine Biological Association of the UK. 85(1): 205-208.

Wyneken, J., T. Burke, M. Salmon, D. Pedersen. 1988. Egg Failure in Natural and Relocated Sea Turtle Nests Egg Failure in Natural and Relocated Sea Turtle Nests. Society for the Study of Amphibians and Reptiles. 22(1): 88-96.

Wyneken, J. S. Madrak, M. Salmon, J. Foote. 1998. Migratory activity by hatchling loggerhead sea turtles (Caretta caretta): evidence for divergence between nesting groups. Mar Biol. 156: 171-178. Wyneken, J. and M. Salmon. 1992. Frenzy and postfrenzy swimming activity in loggerhead, green, and leatherback hatchling sea turtles. Copeia. 1992(2):478 484.

51 Yntema, C. and N. Mrosovsky. 1980. Sexual differentiation in hatchling loggerheads (Caretta caretta) incubated at different controlled temperatures. Herpetologica. 36: 33-36.

52