General and Comparative Endocrinology 117, 299–312 (2000) doi:10.1006/gcen.1999.7419, available online at http://www.idealibrary.com on
Seasonal Changes in Sex and Adrenal Steroid Hormones of Gopher Tortoises (Gopherus polyphemus) Jeannine A. Ott,*,†,1 Mary T. Mendonc¸a,* Craig Guyer,* and William K. Michener† *Department of Zoology and Wildlife Science, Auburn University, Alabama 36849; and †J. W. Jones Ecological Research Center, Ichauway, Newton, Georgia 31770 Accepted November 9, 1999
We sampled a population of gopher tortoises (Gopherus polyphemus) from May to October 1997 to determine seasonal cycles of steroid hormones (testosterone, T; 17b-estradiol, E; and progesterone, P) and related them to observations of mating behavior. In males, plasma T levels peaked in July and August and remained elevated through October. This coincides with the reported time of peak mating and spermatogenesis, indicating that males display an associated pattern of reproduction. In females, E levels were high in September and October. Plasma T levels in females were elevated in May, decreased to basal levels in June and July, and rose again in August and September. Elevated E and T levels correspond to the reported time of peak vitellogenic activity, indicating that females also display an associated cycle. Plasma P in females remained basal throughout the active season, suggesting that ovulation occurs in late winter. We also determined levels of corticosterone (B) to assess the influence of capture stress on tortoises and correlated B levels with tortoise activity patterns and sex steroid levels. We found no seasonal variation in levels of B in males or females. Plasma B levels were not correlated with levels of T or E, but were positively correlated with female P levels. Further, we found no relationship between plasma B levels in males and mean distance moved, mean number of burrows used, or mean home
range size. However, there was a significant negative correlation between plasma B levels and male body size. In females, there was no relationship between B levels and mean distance moved, but B levels were significantly negatively correlated with the number of burrows females occupied. Lastly, there was no relationship between levels of B and the number of minutes required to obtain blood from an animal. However, B levels increased with the length of time that a tortoise spent in a trap, suggesting that trapped tortoises do exhibit capture stress. r 2000 Academic Press Key Words: turtle; testudinidae; gopher tortoise; 17bestradiol; testosterone; progesterone; corticosterone; seasonal cycle. The gopher tortoise (Gopherus polyphemus) is a large terrestrial turtle with a geographic range that extends from extreme southern South Carolina south along the Atlantic Coastal Plain through Florida and west along the Gulf Coastal Plain to extreme southeastern Louisiana (Ernst and Barbour, 1989; Ernst et al., 1994). The longleaf pine ecosystem that covered this range has been reduced as a result of habitat loss, fragmentation, and habitat degradation (Noss, 1989), resulting in an 80% reduction in gopher tortoise population densities range-wide (Auffenberg and Franz, 1982). For this reason, this species has been listed as protected in all states within its range (Ernst et al., 1994). In 1987, gopher tortoises located west of the Mobile and Tom-
1
To whom correspondence should be addressed at Jones Ecological Research Center, RR 2, Box 2324, Newton, GA 31770. Fax: (912) 734-4707. E-mail:
[email protected]. 0016-6480/00 $35.00 Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
299
300
bigbee Rivers in Alabama, Mississippi, and Louisiana were designated as Threatened under the Endangered Species Act of 1973 (TESII, 1995). Furthermore, the gopher tortoise has been considered a keystone species because over 360 species seek refuge in burrows that tortoises dig (Eisenberg, 1983; Jackson and Milstrey, 1987). Thus, the loss of gopher tortoises would have far-reaching effects. Gopher tortoises have an inherently low reproductive rate, primarily due to low clutch size and frequency (Landers et al., 1980). Further, it is estimated that a tortoise has a 94% chance of being predated before it reaches 1 year of age (Alford, 1980). Landers et al. (1980) estimated that a female gopher tortoise produces a clutch that is not depredated only once every 10 years. Few studies have focused attention on ways to improve gopher tortoise reproductive success because little is known about the basic reproductive biology of this species, including the mating system and timing of the mating season (see Iverson, 1980; Landers et al., 1980; Taylor, 1982; Palmer and Guillette, 1990; Diemer and Moore, 1994; Smith, 1995; Butler and Hull, 1996). An improved understanding of the reproductive biology of the gopher tortoise is necessary to allow determination of how or when one should concentrate efforts at recovery of this species. Detailed information on seasonal hormone levels is necessary for understanding the reproductive system of a species as well as the mechanisms regulating its reproductive cycle. Taylor (1982) examined seasonal steroid levels in gopher tortoises and reported cycles of reproductive hormones in both males and females. However, the seasonal concentrations of reproductive steroid hormones reported in his study may have been influenced by his use of a captive population. Stressors associated with captivity cause increased plasma glucocorticoid levels in many species, including reptiles, and this increase has been shown to have rapid and marked effects on plasma steroid levels (Mahmoud et al., 1989; Cree et al., 1990; Moore et al., 1991). Greatly increased levels of stress observed in captive populations may have dramatic effects on reproductive function because captivity might limit the range of activities an individual might use to eliminate the stressor (Norris, 1997). The effects of stress in gopher tortoises
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
Ott et al.
are unknown because Taylor (1982) did not measure corticosterone levels in his study. In this study, we report the levels of sex steroid hormones (testosterone, 17b-estradiol, progesterone) from a natural population of gopher tortoises during their active period (typically late May to early October) and correlate these levels with seasonal observations of mating behavior. We discuss these cycles in light of existing knowledge of the reproductive biology of tortoises reported in other studies. Furthermore, because adrenal hormones can influence levels of reproductive hormones (Mahmoud et al., 1989; Moore et al., 1991; Gregory et al., 1996; Cash et al., 1997; Romero et al., 1998), we report seasonal levels of corticosterone to assess the influence of trapping and handling on sex steroid levels. We also correlate plasma corticosterone levels with body size and seasonal movement patterns.
METHODS Study Site This study was conducted at Ichauway, an 11,700 ha private ecological reserve located in Baker County in southwestern Georgia. Longleaf pine (Pinus palustris) and wiregrass (Aristida stricta) associations comprise over 60% of the property, making it the one of the largest unbroken tracts of gopher tortoise habitat remaining in the United States. The focus of this study was Green Grove, a 70 ha mesic site located in the northeastern portion of Ichauway. Green Grove is comprised of longleaf pine-wiregrass habitat, interspersed with small abandoned agricultural food plots. Similar longleaf pine-wiregrass habitat borders Green Grove in all directions.
Study Animal Sexual maturity is reached at carapace lengths (CL) of 230–240 mm in males and 250–265 mm CL in females, sizes that require growth periods of 10–20 years (Auffenberg and Iverson, 1979; Iverson, 1980; Landers et al., 1982; Diemer and Moore, 1994; Mushinsky and Esman, 1994; Aresco and Guyer, 1999). The mating season has not been clearly defined for this species. Males visit female burrows throughout the
301
Seasonal Cycle in Gopher Tortoises
active season (late May through early November; Boglioli, 1999). However, visits that result in mating occur most often from July to November and mating events peak in September (Boglioli, 1999), suggesting that the mating season occurs from July to November. During cooler months, tortoises spend most of their time inside burrows. Ultrasound images (and subsequent surgical verification) of two females from Green Grove in October 1998 revealed numerous ovarian follicles; the largest follicles ranged from 14–32 mm in diameter. This suggests that vitellogenesis begins in late summer to early fall, as seen in many other turtle species (Wibbels et al., 1990; Rostal et al., 1994; Sarkar et al., 1996; Rostal et al., 1998). Oviposition has been reported in all months from April through July (Iverson, 1980). We used X-ray photography (Gibbons and Greene, 1979; Hinton et al., 1997) to determine the presence of shelled eggs in females from Green Grove, and observed eggs or egg laying from late April to mid-June of 1997. This suggests that oviposition occurred over a shorter time period than has been reported in other studies. Females lay one nest annually with mean clutch size ranging from 3.8 eggs in South Carolina to 8.9 eggs in south Florida (range: 1–12 eggs; Diemer and Moore, 1994; see also Butler and Hull, 1996; Smith et al., 1997). Shelled eggs average 42 mm in diameter and weigh between 30 and 49 g (40.9 6 5.19 g; Iverson, 1980). The Green Grove study site contains a large, freeranging population of gopher tortoises (N 5 75 males and 56 females). Mean straight carapace length of males in this population was 274 mm (222–306 mm CL). Female body size averaged 296 mm CL (210–325 mm CL). As part of a concurrent radio-telemetry study, each tortoise within Green Grove had a radio transmitter attached to its carapace, which allowed us to locate tortoises for collection of blood. All tortoises were monitored from April 1997 to June 1998. Home range size, number of burrows used, and mean distance moved between burrows were measured for all animals within Green Grove (Ott, 1999).
Blood Collection We collected blood biweekly from the marked population at Green Grove from May through November 1997. During each sampling period, we randomly selected 13 male and 13 female tortoises and placed
Tomahawk live traps at their burrow entrances. Traps were set the morning of the first sampling day, and remained open until the evening of the last sampling day. Each trap was checked initially at 0800 h to ensure that it was empty. We then checked traps at 2-h intervals until 1800 h. (Tortoises are typically below ground and are often disturbed by aboveground activity, including the sound of footsteps approaching the burrow. For this reason, we limited the number of times we approached each burrow to check traps.) Each sampling period lasted 3 days. Most samples (65%) were collected after 1300 h. Within 5 min after a tortoise was discovered in a trap, we collected 0.5 to 1.0 ml of blood from a femoral vein using a heparinized 1-ml syringe and a 26-gauge needle. Whole blood was stored on ice (,2 h) until the sample could be centrifuged. We then weighed each tortoise using a 10-kg pesola scale, identified its sex, and recorded any signs of trap stress (runny nose or eyes, presence of scat), trap injury (loss of nails, scraped feet or nose), or illness (upper respiratory disease; Jacobsen et al., 1991). When captured, tortoises either sat at the rear of the trap as if they were basking or they paced back and forth and/or attempted to climb the trap walls. In most cases, tortoises were cool to the touch and still had fresh soil on their carapace, suggesting that they had not been in the trap for long periods of time. After processing, each tortoise was released at the entrance of the burrow where it was captured. Blood samples were centrifuged at 3000 rpm for 10 min and the plasma was pipetted into vials and stored in a freezer for later hormone analyses.
Radioimmunoassay Plasma was thawed and incubated with 50 µL of a 1:10 solution of a specific hormone [testosterone (T), corticosterone (B), 17b-estradiol (E), or progesterone (P)] for 1 h to assess extraction efficiency. Extraction and radioimmunoassay procedures follow those described in Mendonc¸a et al. (1996a). Plasma was extracted with diethyl ether and dried down with nitrogen gas, and the hormone was resuspended in a phosphate buffer solution. The samples were aliquoted in duplicate (100 µL each) and incubated for 4 h at room temperature with the respective tritiated hormone and antibody. A third aliquot (50 µL) was used to
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
302
determine extraction efficiency. We measured circulating levels of T and B in plasma from males and females. In addition, we measured E and P levels from females. Values were corrected for volume of plasma and extraction efficiency and are reported in nanograms per milliliter of plasma. Interassay variation averaged 16% for all hormones, except E (all samples were analyzed for E in one assay). Intraassay variation averaged 11%. Assay sensitivity was 60 pg/ml for T, 140 pg/ml for B, 10 pg/ml for E, and 30 pg/ml for P. Percentage extraction efficiency averaged 89% for T, 87% for B, 78% for E, and 63% for P. We validated our steroid assays following the procedure described in Mendonc¸a et al. (1996b). Gopher tortoise plasma was charcoal-stripped until steroid levels were below the detectable limits of the assay. We then added known amounts of each steroid hormone to be assayed to the stripped plasma. Measured levels of each steroid were within 5% of the amounts added. We further validated our steroid assays for use in gopher tortoises with serially diluted control plasma (stripped and steroid-added), native plasma, and hormone standards and obtained parallel binding curves.
Assessment of Stress To assess the response to trap stress, we opportunistically sampled blood from unmarked, free-ranging tortoises (3 male and 6 female) found outside the Green Grove study area. We conducted two experiments to determine the influence of captivity on plasma B levels. In Experiment 1, an initial blood sample (0.5 ml) was taken from each tortoise within 5 min (n 5 4) of capture. Corticosterone levels from this sample were assumed to represent basal levels. These tortoises then were placed in live traps and a second 0.5-ml blood sample was taken after 12 h. Corticosterone levels from this sample were assumed to represent maximum stress levels associated with trapping and handling. In Experiment 2, each tortoise was placed in a live trap for 2 h (equal to the interval of time between trap inspections), and then an initial 0.5-ml blood sample was taken within 5 min after removal from the trap (N 5 4). Corticosterone levels from this sample were assumed to represent stress associated with trapping alone. Blood samples from tortoises in the Green Grove population that were captured by hand were used to represent basal levels
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
Ott et al.
of stress. We compared these basal levels from Green Grove animals to levels observed in Experiment 2 animals to determine if captivity affected B levels. In both experiments, tortoises were released in a burrow near their point of capture. Blood samples were processed as above.
Statistical Analysis Monthly T levels in male gopher tortoises and B levels in males and females were log10-transformed to improve normality (Male T: W 5 0.88, Pr , W: 0.0001; Male B: W 5 0.98, Pr , W: 0.35; Female B: 0.97; Pr , W: 0.39). In females, levels of T, E, and P were square-root transformed to improve normality (T: W 5 0.83, Pr , W: 0.0001; E: W 5 0.93, Pr , W: 0.02; P: W 5 0.96, Pr , W: 0.09). In this study, ‘‘season’’ refers to the time period between May and October (the active period). Results of each biweekly sample were pooled by month and are presented as a single mean. We used a two-way general linear model (GLM) procedure in SAS (SAS, 1996) to determine if there were any daily (morning versus afternoon sample) or monthly differences in plasma levels of T, B, E, or P; these tests were followed by a Tukey’s studentized (HSD) range test when significant results were observed. For animals in the Green Grove population, we looked at the effect of time (number of hours in a trap and number of minutes to obtain blood) on plasma B levels using a Pearson correlation. Additionally, we compared B levels between animals that exhibited behavioral signs of stress (including trap stress, trap injury, or illness; see above) and those that did not exhibit such stress. We correlated B levels with body size, mean monthly movement distances, mean number of burrows used, and mean home range size (Ott, 1999) to determine whether animals that moved more often or longer distances exhibited higher B levels. Finally, we examined relationships between plasma B levels and levels of T, E, and P. Levels of each hormone were log10transformed for consistency before correlations were performed. In Experiment 1, we assessed differences between initial B levels and B levels taken 12 h after placement in a live trap using a paired t test. We used a two-tailed t test to compare B levels from Experiment 2 animals after 2 h in captivity with B levels observed in hand-
Seasonal Cycle in Gopher Tortoises
captured animals from the Green Grove population. We also correlated B levels with the number of minutes required to bleed a tortoise (Pearson correlation) to determine if time to bleed influenced stress levels in experimental animals. All values are presented as the mean 61 standard error. Alpha level was set at 0.05.
RESULTS General (Green Grove) Plasma T levels in males varied significantly by month (F5,82 5 4.66, P 5 0.0009) but did not differ between morning and afternoon samples (F1,82 5 0.63, P 5 0.43). Levels of T began rising in June and were significantly higher in July and August than in May (Fig. 1). These levels remained elevated through October. There was no monthly or daily variation in plasma B levels (monthly: F5,87 5 1.72, P 5 0.14, Fig. 3; daily: F1,87 5 0.09, P 5 0.76). Further, we found no relationship between plasma B levels in males and mean number of movements (r 5 0.03, N 5 85, P 5 0.79), mean number of burrows used (r 5 20.05, N 5 46, P 5 0.74), or mean home range size (r 5 20.07, N 5 48, P 5 0.64) of males. However, there was a significant negative correlation between plasma B levels and male body size (r 5 20.22, N 5 85, P 5 0.047). In females, E levels varied significantly by month
303
(F5,61 5 19.77, P 5 0.0001) but there was no variation between morning and afternoon samples (F1,61 5 1.90, P 5 0.17). Plasma E levels remained at basal levels in May, June, and July, rose significantly in August, and remained elevated in September and October (Fig. 2B). Levels of T exhibited monthly variation (F5,58 5 2.45, P 5 0.047) but did not vary between morning and afternoon samples (F1,58 5 0.50, P 5 0.49). Plasma T levels were elevated in May 1997 when sampling began, but decreased to basal levels in June and July. In August, T levels began rising again and were significantly higher in September (Fig. 2A). Plasma T levels then decreased in October. Plasma P levels in females did not exhibit monthly or daily variation (monthly: F5,57 5 0.28, P 5 0.92, Fig. 2B; daily: F1,57 5 1.51, P 5 0.22). Plasma P levels of females that had shelled eggs (3.04 6 0.79 ng/ml; range 5 1.27–5.91 ng/ml) were slightly higher than females without eggs sampled at the same time (2.02 6 0.35 ng/ml; range 5 0–4.19 ng/ml), although these means are not statistically different (t test: t 5 21.38, df 5 17, P 5 0.18). There was no monthly or daily fluctuation in plasma B levels of females (monthly: F5,46 5 0.74, P 5 0.60, Fig. 3; daily: F1,87 5 0.19, P 5 0.67). Further, there was no relationship between plasma B levels and mean number of moves by females (r 5 20.13, N 5 45, P 5 0.41). However, plasma B levels were significantly negatively correlated with the number of burrows females used (r 5 20.49, N 5 28, P 5 0.008; Fig. 4). Home
FIG. 1. Mean plasma testosterone levels of male gopher tortoises (Gopherus polyphemus) during the active season from Green Grove, Baker County, Georgia. Numbers below data points indicate sample size. Vertical lines represent 61 standard error.
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
304
Ott et al.
FIG. 2. Mean plasma testosterone levels (A) and mean plasma 17b-estradiol and progesterone levels (B) of female gopher tortoises (Gopherus polyphemus) during the active season from Green Grove, Baker County, Georgia. Numbers indicate sample size for each data point. Vertical lines represent 61 standard error.
FIG. 3. Mean plasma corticosterone levels of male and female gopher tortoises (Gopherus polyphemus) during the active season from Green Grove, Baker County, Georgia. Numbers above lines indicate sample size for females; numbers below lines indicate sample size for males. Vertical lines represent 61 standard error.
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
305
Seasonal Cycle in Gopher Tortoises
FIG. 4. Relationship between mean number of burrows used and plasma corticosterone levels in female gopher tortoises (r 5 20.49, N 5 28, P 5 0.008).
range size increased with rising plasma B levels, but this relationship was not significant (r 5 0.34, N 5 31, P 5 0.12). There was no relationship between female body size and plasma B levels (r 5 20.005, N 5 31, P 5 0.98).
Assessment of Stress In Experiment 1, plasma B levels increased after animals were held in a trap for 12 h, but this rise was only marginally significant (mean difference between initial and posttrap samples: 7.22 6 2.9 ng/ml; t 5 2.49, P 5 0.07, N 5 5). In Experiment 2, plasma B levels did not differ significantly between animals placed in live traps for 2 h (1.21 6 0.48 ng/ml, N 5 4) and handcaptured animals from Green Grove (1.14 6 0.25 ng/ ml, N 5 10; t 5 20.12, N 5 12, P 5 0.90). For experimental animals only, there was no relationship between plasma levels of B and the number of minutes (maximum 5 25 min) it required to obtain blood (r 5 0.32, N 5 11, P 5 0.33). Within the Green Grove population, plasma B levels were significantly positively correlated with the length of time a tortoise spent in a trap (r 5 0.29, N 5 88, P 5 0.006; Fig. 5). There also was no relationship between the number of minutes required to obtain blood (maximum 5 24 min) and plasma levels of B (r 5 0.07, N 5 135, P 5 0.43) for animals within Green Grove. Tortoises that appeared stressed while blood
was being drawn (including trap stress, trap injury, or illness) had significantly higher levels of B than tortoises that did not appear stressed (stressed: 0.76 6 0.06 ng/ml, N 5 29; unstressed: 0.61 6 0.03 ng/ml, N 5 106; t 5 22.31, P 5 0.022, N 5 134). Plasma B levels were not correlated with levels of E (r 5 0.07, N 5 45, P 5 0.65) or T (male: r 5 20.05, N 5 77, P 5 0.66; female: r 5 0.08, N 5 45, P 5 0.61). However, plasma B levels in females were significantly positively correlated with plasma P levels (r 5 0.70, N 5 44, P 5 0.0001).
DISCUSSION Male Reproductive Cycle Males displayed a significant increase in plasma T in July, and T levels remained significantly higher during successive months of the active season than in May and June. Taylor (1982) reported a relationship between increasing T levels and spermatogenic stage, with spermiation and maximum testis weight coinciding with peak plasma T levels in September and October. This suggests that spermatogenesis may be androgen dependent (McPherson et al., 1982; Mendonc¸a and Licht, 1986). Additionally, the peak in plasma T observed in the present study immediately
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
306
Ott et al.
FIG. 5. Relationship between the number of hours a tortoise spent in a trap prior to being sampled and plasma corticosterone levels in gopher tortoises (r 5 0.29, N 5 88, P 5 0.006).
precedes the peak period of mating in south Georgia (late July to October; Boglioli, personal communication), indicating that male gopher tortoises display an associated reproductive cycle (Fig. 6). This cycle of testicular activity has been reported for most birds, lizards, and alligators, some snakes, and many turtles (Crews, 1984). The peak period of mating also corresponds to peak movement activity, with males moving significantly greater distances, occupying significantly more burrows, and sharing burrows with females significantly more often in August and September than in other months (Ott, 1999). A similar association between testicular and androgen cycles was observed in male desert tortoises (Gopherus agassizii) in the eastern Mojave Desert (Rostal et al., 1994). Testosterone levels rose significantly beginning in June and remained high until October, when tortoises became inactive. Further, spermatogenesis appears to begin in May, and seminiferous tubules reached their maximum diameter in October (Rostal et al., 1994). In the desert tortoise, the mating season is divided between fall and spring months, with fall mating immediately proceeding spermatogenesis and coinciding with peak T levels (Rostal et al., 1994). However, this associated cycle is not seen in all tortoises. In the tortoise Testudo h. hermanni from Yugoslavia and Austria, the main mating season occurs when T levels are low and testes are regressed (Kuchling et al., 1981).
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
The seasonal T cycle from field-captured gopher tortoises in this study is similar to the cycle observed in captive males from north Florida (Taylor, 1982), though the levels reported in the latter study were nearly an order of magnitude lower than those in the present study (31.6 ng/ml versus 325.5 ng/ml, respectively). Despite these lowered T levels, Taylor (1982) observed a seasonal gonadal and hormonal cycle in captive males, but his animals did not exhibit any reproductive behaviors. In the present study, we observed both courtship and mating by males. This may suggest that the high T levels observed in the Green Grove animals may be necessary to stimulate mating behavior. High levels of T also were reported for desert tortoises (G. agassizii) in the eastern Mojave Desert (Rostal et al., 1994). An analysis of the bound and free portions of plasma T may further elucidate the significance of the high T levels measured in the Green Grove population. Although Taylor (1982) did not measure plasma B during his study, the lower T levels observed might result from the influence of stress caused by animals being held in captivity. In the present study, we found no evidence that rising B levels cause a reduction in plasma T in animals that spent only a few hours in captivity. However, levels of B increased after tortoises were held in traps for 12 h, documenting a response to long periods of capture. Mendonc¸a and Licht (1986) reported a 35–60% decrease in plasma T in male musk turtles within 24 h of capture, and this decline in T was
Seasonal Cycle in Gopher Tortoises
307
FIG. 6. Schematic representation of fluctuations in plasma testosterone and testis weight during the reproductive cycle of the male gopher tortoise, Gopherus polyphemus. Large boxes represent behavioral and physiological events associated with the reproductive cycle, and shading within boxes indicates periods of maximum occurrence of that event. Small boxes indicate the months of the year. Modified from Taylor (1982), Palmer and Guillette (1990), Boglioli (1999), and this study.
correlated with increased levels of B over the captive period. Further, musk turtles held in captivity for several months had significantly lower levels of T, though a seasonal cycle was still exhibited. Similar results were found in painted turtles and snapping turtles (Licht et al., 1985; Mahmoud et al., 1989). Because elevated levels of B may negatively influence mating behavior in male gopher tortoises, further studies should examine natural and artificial sources of stress (e.g., poor quality habitat, reduced food availability, disease) to determine the potential indirect effects of stress on reproduction.
Female Reproductive Cycle Female gopher tortoises also exhibited a seasonal cycle of plasma steroid hormones. Plasma E levels remained basal from May through July, rising significantly in August and remaining elevated through October. Taylor (1982) reported analogous results for
captive females from north Florida, though peak levels were lower (2.5 ng/ml) than levels observed in this study (6.8 ng/ml). In this study, the elevated levels of E in late summer correlated with published reports of initial ovarian enlargement and the beginning of vitellogenesis in this species (Iverson, 1980; Taylor, 1982; Palmer and Guillette, 1990). In October 1998, we found ovarian follicles ranging from 14 to 32 mm in diameter (mean diameter of shelled eggs 5 42 mm) in two females from the Green Grove population. We did not sample during the winter months, but Palmer and Guillette (1990) reported an increase in ovarian and oviductal mass from October through April, although it is uncertain whether vitellogenesis proceeds during colder months. Vitellogenic activity then occurred rapidly over a period of 6 weeks in the spring, indicated by an increase in ovarian mass prior to ovulation, and E declined to basal levels postovulation (Palmer and Guillette, 1990). Assuming that the same patterns occur in the Green Grove population, this
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
308
cycle would explain the initial low E levels observed at the beginning of the active season (May) in this study (Fig. 2B). Plasma T levels were high at the beginning of the active season in May, but declined to basal levels in June and July. Levels of T then began rising again in August, and were significantly higher in September before declining again in October (Fig. 2A). Rostal et al. (1994) reported a similar rise of T in August in female desert tortoises (G. agassizii), but T levels did not peak until the following April. This may suggest that T levels in female gopher tortoises may only begin to rise in late summer, remain high during the winter months, and then decline in the spring, as indicated by the smaller peak of T observed in this study. Plasma P levels did not vary seasonally in this study, remaining at about 2 ng/ml (Fig. 2B) during the active season. Taylor (1982) reported two peaks of P in captive females from north Florida. One captive female had high P levels in late October, but the major peak occurred in April (Taylor, 1982). However, levels of plasma P (approximately 0.1–1.45 ng/ml) reported by Taylor (1982) were lower than the basal levels observed in wild-caught animals in the present study (1.42–2.57 ng/ml). It is possible that the some of the circulating P observed in females at Green Grove is of adrenal origin. Mahmoud et al. (1989) observed significant changes in P levels in Chelydra serpentina after being held captive, with P levels increasing after 6 h in captivity and returning to initial levels after 7 days. Similar results were found in male tuatara (Sphenodon punctatus), where P levels increased significantly after 3 h in captivity (Cree et al., 1990). In this study, plasma P levels also were positively correlated with plasma B levels. Further, levels of B increased significantly with the length of time an animal was in a trap (Fig. 5). The influence of captivity on plasma P levels was not examined directly during this study. However, our results suggest that, in gopher tortoises, P levels are influenced by capture stress, although it is unknown whether these effects would still be observed months after animals are placed in captivity. Plasma levels of P from females that had shelled eggs were slightly higher than those of females without eggs. Gravidity and nesting have been reported to occur from late-March to mid-June (Palmer and Guil-
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
Ott et al.
lette, 1990; Smith et al., 1997). X-ray photographs of females from Green Grove and anecdotal observations of nesting confirm the presence of eggs in the oviduct from late April to early June in this population. This indicates that plasma P levels decreased to basal by late April even in gravid animals. In most cases, we captured females as they first emerged from burrows in the spring. However, blood samples were not taken from females captured in late March or April. The low P levels observed in early May from gravid females in this study suggest that ovulation and egg shelling occur before animals emerged from their burrows in spring. Therefore, it is not surprising that females did not exhibit a peak in plasma P during this study. Following nesting, a period of reproductive quiescence occurs during summer (Palmer and Guillette, 1990). The low levels of P, E, and T during the summer months support the assumption that female gopher tortoises produce only one clutch annually (Landers et al., 1980; Taylor, 1982; Palmer and Guillette, 1990; Diemer and Moore, 1994; Smith, 1995; Butler and Hull, 1996; Smith et al., 1997). Similar patterns of ovarian growth and steroid hormone production are known for female desert tortoises, though some desert tortoises have been reported to produce two clutches a year (Rostal et al., 1994). It is unknown whether gopher tortoises could produce an additional clutch each year. Peak mating activity occurs in late summer (Boglioli, personal communication; Fig. 7) when ovarian follicles have just begun to reach preovulatory size. Because ovulation of follicles does not occur until spring, females most likely store sperm through the winter months and fertilize the ova after they ovulate in spring (Saint-Girons, 1973; Gist and Jones, 1987, 1989). Anecdotal information suggests that females may store sperm for several years (Porter, 1972; Iverson, 1980), indicating the potential for sperm competition to occur.
Assessment of Stress Gopher tortoises within Green Grove did not exhibit seasonal or daily cycles of plasma B. Further, we did not see a relationship between B levels and plasma levels of T or E or between B levels and the number of minutes it took to obtain blood from animals. However, plasma B levels were significantly positively correlated with levels of P in females. A similar rise in P
Seasonal Cycle in Gopher Tortoises
309
FIG. 7. Schematic representation of fluctuations in plasma 17b-estradiol and progesterone levels and combined ovarian and oviductal weight during the reproductive cycle of the female gopher tortoise, Gopherus polyphemus. Large boxes represent behavioral and physiological events associated with the reproductive cycle, and shading within boxes indicates periods of maximum occurrence of that event. Small boxes indicate months of the year. Modified from Taylor (1982), Palmer and Guillette (1990), and this study.
with increasing levels of B was seen in male tuatara (S. punctatus; Cree et al., 1990) and female snapping turtles (C. serpentina; Mahmoud et al., 1989). During our study, plasma P levels did not fluctuate within the active season. However, as noted earlier, levels of P observed in captive animals by Taylor (1982) were substantially lower than those observed in animals from Green Grove. It is possible that stress from capture and handling suppresses the production of P to produce lower basal levels than seen in the Green Grove tortoises. In the Green Grove population, levels of B were positively correlated with the length of time a tortoise spent in a trap. However, this relationship may be misleading because we did not know the exact amount of time a tortoise remained in a trap, only the amount of time that had elapsed since the last time we checked a trap. Therefore, we cannot be certain that higher B levels observed are actually in response to trap time. In Experiment 1, where the amount of time an animal
spent in a trap was controlled, plasma B levels increased an average of 7.2 6 2.9 ng/ml in tortoises held for 12 h. These levels were substantially higher than baseline levels observed in hand-captured animals from Green Grove (1.14 6 0.25 ng/ml). Therefore, it is possible that small sample sizes reduced our power to detect a significant difference between initial and final B levels in Experiment 1. However, in Experiment 2, we did not observe a rise in B after animals spent only 2 h in a trap (P 5 0.9). These results suggest that gopher tortoises do not respond to short-term trapping and handling stress with a rise in plasma B levels. Similar results have been reported for other reptiles (Mahmoud et al., 1989; Cree et al., 1990). However, because we found higher B levels after longer periods in captivity, a complete plasma profile of B secretion over time should be conducted to determine the influence of stress on reproductive hormones. We did not find relationships between rising B levels and any movement parameter in males. There was a
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
310
significant negative correlation between B levels and male body size. This relationship appears to be obscured by one male that exhibited substantially higher B levels. Removal of this animal from the data set does not affect this conclusion (r 5 20.28, P 5 0.054) between body size and B levels. This suggests that smaller males have a quicker stress response than larger males. However, body size explained only a small portion of the variance in B levels. In females, we found a significant negative relationship between burrow use and B levels, where females occupying fewer burrows exhibited higher levels of B. This suggests that females who changed burrows less frequently were either more stressed or exhibited a quicker stress response than animals that changed burrows more often. Boglioli (1999) reported that females may maintain a dominance hierarchy in which the largest females prohibit smaller females from inhabiting the best quality habitat within a landscape. Thus, larger females occupy burrows with the least canopy cover and have access to the best nesting sites (Boglioli, 1999). The higher levels of B observed in larger females may be associated with increased stress from trying to maintain these ‘‘high-quality’’ burrows. However, one female had substantially higher B levels than all other individuals, and this female may have obscured the true relationship between B levels and numbers of burrows used. Removal of this female did result in a weakened relationship (r 5 20.29, P 5 0.11). Thus, these results may only be artifacts of our data, since tortoises in this population appear to be relatively unstressed as a whole. The impacts of stress on reproduction should be examined more closely in this species. We found significantly higher levels of T, E, and P in wildcaptured gopher tortoises than Taylor (1982) found in captive animals, and these low levels in the captive animals are probably due to the stress of captivity. It is uncertain whether these lowered hormone levels negatively impair reproductive function. However, Taylor’s (1982) study does suggest that stress may negatively impact reproductive behavior. If gopher tortoises negatively respond to capture stress, it may be possible that ‘‘natural’’ stressors, like low-quality habitat, insufficient food availability, or increased canopy cover, can impact levels of sex steroid hormones. These stressors also may reduce the amount of energy individuals
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
Ott et al.
spend searching for mates or nesting sites. Because a decline in reproductive output would negatively influence recovery of gopher tortoise populations (see Congdon et al., 1994), future studies should examine the influence of habitat stress on reproduction. Further, recovery techniques should concentrate on improving conditions for tortoise populations during peak periods of reproduction to enhance reproductive output.
ACKNOWLEDGMENTS This paper is part of a thesis submitted by J.A.O. in partial fulfillment of the requirements for a Master’s degree from Auburn University. We thank Melissa Dills Boglioli, Tracey Tuberville, Derek Fussell, and the many undergraduate students who assisted in fieldwork. We especially thank Roger Birkhead for his assiduous help in the field and his knowledge of gopher tortoises. We are grateful to Dr. Ira Roth, D.V.M., for his interest in tortoise biology and for volunteering his time and supplies to take X-ray photographs of gopher tortoises, and Dr. Robert E. Cartee, D.V.M., for his help locating ovarian follicles with ultrasound technology. We thank Haley Burke, Rene´e Duckworth, and Jennifer Shelby for help with radioimmunoassays. This manuscript was improved by comments from Rene´e Duckworth, Melissa Boglioli, Jennifer Shelby, and two anonymous reviewers. Funding was provided by the Robert W. Woodruff Foundation through the J. W. Jones Ecological Research Center and by The Gopher Tortoise Council through the J. Larry Landers Student Research Award. Support to M.T.M. was through Auburn University’s Agricultural Experiment Station ALA-16-019.
REFERENCES Alford, R. A. (1980). Population structure of Gopherus polyphemus in northern Florida. J. Herpetol. 14, 177–182. Aresco, M. J., and Guyer, C. (1999). Burrow abandonment by gopher tortoises in slash pine plantations of the Conecuh National Forest. J. Wildl. Manage. 63, 26–35. Auffenberg, W., and Franz, R. (1982). Status and distribution of the gopher tortoise (Gopherus polyphemus). In ‘‘North American Tortoises: Conservation and Ecology’’ (R. B. Bury, Ed.). USDI FWS, Washington, DC. Auffenberg, W., and Iverson, J. B. (1979). Demography of terrestrial turtles. In ‘‘Turtles: Perspectives and Research’’ (M. Harless and H. Morlock, Eds.). Krieger, Malabar, FL. Boglioli, M. D. (1999). Burrow dispersion and occupancy patterns as they relate to habitat parameters and social behavior in the gopher tortoise, Gopherus polyphemus. M.S. thesis, Auburn Univ., Auburn. Butler, J. A., and Hull, T. W. (1996). Reproduction of the tortoise, Gopherus polyphemus, in northeastern Florida. J. Herpetol. 30, 14–18.
Seasonal Cycle in Gopher Tortoises Cash, W. B., Holberton, R. L., and Knight, S. S. (1997). Corticosterone secretion in response to capture and handling in free-living red-eared slider turtles. Gen. Comp. Endocrinol. 108, 427–433. Congdon, J. D., Dunham, A. E., and van Loben Sels, R. C. (1994). Demographics of common snapping turtles (Chelydra serpentina): Implications for conservation and management of long-lived organisms. Am. Zool. 34, 397–408. Cree, A., Guillette, L. J., Jr., Cockrem, J. F., and Joss, J. M. P. (1990). Effects of capture and temperature stresses on plasma concentrations in male Tuatara (Sphenodon punctatus). J. Exp. Zool. 253, 38–46. Crews, D. (1984). Gamete production, sex hormone secretion, and mating behavior uncoupled. Horm. Behav. 18, 22–28. Diemer, J. E., and Moore, C. T. (1994). Reproduction of gopher tortoises in north-central Florida. Fish and Wildlife Research 13 USDI National Biological Survey, Washington, DC. Eisenberg, J. F. (1983). The gopher tortoise as a keystone species. Proc. 4th Annual Meeting of the Gopher Tortoise Council. Florida State Museum, Gainesville. Ernst, C. H., and Barbour, R. W. (1989). ‘‘Turtles of the World.’’ Smithsonian Inst. Press, Washington, DC. Ernst, C. H., Barbour, R. W., and Lovich, J. E. (1994). ‘‘Turtles of the United States and Canada.’’ Smithsonian Inst. Press, Washington, DC. Gibbons, J. W., and Greene, J. L. (1979). X-ray photography: A technique to determine reproductive patterns of freshwater turtles. Herpetologica 35, 86–89. Gist, D. H., and Jones, J. M. (1987). Storage of sperm in the reptilian oviduct. Scanning Microscopy 1, 1839–1849. Gist, D. H., and Jones, J. M. (1989). Sperm storage within the oviduct of turtles. J. Morphol. 199, 379–384. Gregory, L. F., Gross, T. S., Bolten, A. B., Bjorndal, K. A., and Guillette, L. J., Jr. (1996). Plasma corticosterone concentrations associated with acute captivity stress in wild loggerhead sea turtles (Caretta caretta). Gen. Comp. Endocrinol. 104, 312–320. Hinton, T. G., Fledderman, P. D., Lovich, J. E., Congdon, J. D., and Gibbons, J. W. (1997). Radiographic determination of fecundity: Is the technique safe for developing turtle embryos? Chelonian Conserv. Biol. 2, 409–414. Iverson, J. B. (1980). Reproductive biology of Gopherus polyphemus (Chelonia: Testudinidae). Am. Midl. Nat. 103, 353–359. Jackson, D. R., and Milstrey, E. G. (1987). Fauna of gopher tortoise burrows. Nongame Wildl. Prog. Tech. Rep. 5 Florida Game and Freshwater Fish Commission, Gainesville, FL. Jacobsen, E. R., Gaskin, J. M., Brown, M. B., Harris, R. K., Gardiner, C. H., LaPointe, J. L., Adams, H. P., and Reggiardo, C. (1991). Chronic upper respiratory tract disease of free-ranging desert tortoises (Xerobates agassizii). J. Wildl. Dis. 27, 296–316. Kuchling, G., Skolek-Winnisch, R., and Bamberg, E. (1981). Histochemical and biochemical investigation on the annual cycle of testis, epididymis, and plasma testosterone of the tortoise, Testudo hermanni hermanni Gmelin. Gen. Comp. Endocrinol. 44, 194–201. Landers, J. L., Garner, J. A., and McRae, W. A. (1980). Reproduction of gopher tortoises (Gopherus polyphemus) in southwestern Georgia. Herpetologica 36, 353–361.
311 Landers, J. L., McRae, W. A., and Garner, J. A. (1982). Growth and maturity of the gopher tortoise in southwestern Georgia. Bull. Florida State Mus. 27, 38–94. Licht, P., Breitenbach, G. L., and Congdon, J. D. (1985). Seasonal cycles in testicular activity, gonadotrophin, and thyroxine in the painted turtle, Chrysemys picta, under natural conditions. Gen. Comp. Endocrinol. 59, 130–139. Mahmoud, I. Y., Guillette, L. J., Jr., McAsey, M. E., and Cady, C. (1989). Stress-induced change in serum testosterone, estradiol-17b and progesterone in the turtle, Chelydra serpentina. Comp. Biochem. Physiol. A 93, 423–427. McPherson, R. J., Boots, L. R., MacGregor III, R., and Marion, K. R. (1982). Plasma steroids associated with seasonal reproductive changes in a multiclutched freshwater turtle, Sternotherus odoratus. Gen. Comp. Endocrinol. 48, 440–451. Mendonc¸a, M. T., Chernetsky, S. D., Nester, K. E., and Gardner, G. L. (1996a). Effects of gonadal sex steroids on sexual behavior in the big brown bat, Eptesicus fuscus, upon arousal from hibernation. Horm. Behav. 30, 153–161. Mendonc¸a, M. T., and Licht, P. (1986). Seasonal cycles in gonadal activity and plasma gonadotropin in the musk turtle, Sternotherus odoratus. Gen. Comp. Endocrinol. 62, 459–469. Mendonc¸a, M. T., Tousignant, A. J., and Crews, D. (1996b). Pinealectomy, melatonin, and courtship behavior in male red-sided garter snakes (Thamnophis sirtalis parietalis). J. Exp. Zool. 274, 63–74. Moore, M. C., Thompson, C. W., and Marler, C. A. (1991). Reciprocal changes in corticosterone and testosterone levels following acute and chronic handling stress in the tree lizard, Urosaurus ornatus. Gen. Comp. Endocrinol. 81, 217–226. Mushinsky, H. R., and Esman, L. A. (1994). Perceptions of gopher tortoise burrows over time. Florida Field Nat. 22, 1–7. Norris, D. O. (1997). ‘‘Vertebrate Endocrinology.’’ Academic Press, San Diego. Noss, R. F. (1989). Longleaf pine and wiregrass: Keystone components of an endangered ecosystem. Nat. Areas J. 9, 211–213. Ott, J. A. (1999). Patterns of movement, burrow use, and reproduction in a population of gopher tortoises (Gopherus polyphemus): Applications to the conservation and management of a declining species. M.S. thesis, Auburn Univ., Auburn. Palmer, B. D., and Guillette, L. J. (1990). Morphological changes in the oviductal endometrium during the reproductive cycle of the tortoise, Gopherus polyphemus. J. Morphol. 204, 323–333. Porter, K. R. (1972). ‘‘Herpetology.’’ Saunders, Philadelphia, PA. Romero, L. M., Soma, K. K., O’Reilly, K. M., Suydam, R., and Wingfield, J. C. (1998). Hormones and territorial behavior during breeding in snow buntings (Plectrophenax nivalis): An arcticbreeding songbird. Horm. Behavior 33, 40–47. Rostal, D. C., Lance, V. A., Grumbles, J. S., and Alberts, A. C. (1994). Seasonal reproductive cycle of the desert tortoise (Gopherus agassizii) in the eastern Mojave desert. Herp. Monogr. 8, 72–82. Rostal, D. C., Owens, D. W., Grumbles, J. S., MacKenzie, D. S., and Amoss, M. S., Jr. (1998). Seasonal reproductive cycle of the Kemp’s Ridley sea turtle (Lepidochelys kempi). Gen. Comp. Endocrinol. 109, 232–243.
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
312 Saint-Girons, H. (1973). Sperm survival and transport in the female genital tract of reptiles. Biology of the Spermatozoa, INSERM International Symposium, Nouzilly. Sarkar, S., Sarkar, N. K., and Maiti, B. R. (1996). Seasonal pattern of ovarian growth and interrelated changes in plasma steroid levels, vitellogenesis, and oviductal function in the adult female softshelled turtle Lissemys punctata punctata. Can. J. Zool. 74, 303–311. SAS (1996). Statistical Analysis System. SAS Institute, Cary, NC. Smith, K. R., Hurley, J. A., and Seigel, R. A. (1997). Reproductive biology and demography of gopher tortoises (Gopherus polyphemus) from the western portion of their range. Chelonian Conserv. Biol. 2, 596–600. Smith, L. L. (1995). Nesting ecology, female home range and activity
Copyright r 2000 by Academic Press All rights of reproduction in any form reserved.
Ott et al. patterns, and population size-class structure of the gopher tortoise, Gopherus polyphemus, on the Katharine Ordway Preserve, Putnam County, FL. Bull. Florida Mus. Nat. Hist. 37, 97–126. Taylor, R. W., Jr. (1982). Seasonal aspects of the reproductive biology of the gopher tortoise, Gopherus polyphemus. Unpublished Ph.D. dissertation, Univ. of Florida, Gainesville. TESII (1995). ‘‘U.S. Threatened and Endangered Species. Volume 1: Animals.’’ Threatened and Endangered Species Information Institute, Golden, CO. Wibbels, T., Owens, D. W., Limpus, C. J., Reed, P. C., and Amoss, M. S., Jr. (1990). Seasonal changes in serum gonadal steroids associated with migration, mating, and nesting in the loggerhead sea turtle (Caretta caretta). Gen. Comp. Endocrinol. 79, 154–164.
