Annual and diurnal cycles in plasma testosterone and thyroxine in the male green sea turtle Chelonia mydas

GENERAL

AND

COMPARATIVE

ENDOCRINOLOGY

57, 335-344 (1985)

Annual and Diurnal Cycles in Plasma Testosterone and Thyroxine the Male Green Sea Turtle Cheionia mydas PAUL LICHT,JAMES

in

E WOOD,* AND FERN E. WOOD*

Department of Zoology, University of California, and *Cayman Turtle Farm, Ltd, P.Q. Box 645 Grand

Berkeley, California Cayman Island, British

94720, West Indies

Accepted May 8, 1984 Male plasma testosterone (T) and thyroxine (T$ were monitored over several annual cycles in a captive breeding colony of green sea turtles, Chelonia mydas. Daily and annual water temperatures varied by only -1 and 3”, respectively. A pronounced season cycle in plasma T was evident in the population as a whole and in individual animals: plasma T was at a nadir (-3 nglml) in September-November and then increased progressively to a peak (27-39 rig/ml) in April; levels began declining immediately thereafter, coincident with the onset of copulatory behavior. By contrast, plasma T4 remained uniform (-9 ngiml) throughout the year and, thus, could not readily account for the decline in androgen levels. Plasma hormones were relatively stable over a 24-hr period at three times a year, and there was a correlation for individual plasma T levels sampled in April and May. Thus, limited sampling should allow identification of seasonal rhythms and individual variability in plasma T levels. Testis mass and spermatogenic activity were significantly greater in January than in September; i.e., spermatogenesis and androgen secretion were not “uncoupled.” Copulatory activity began in April but did not peak until May-June, after plasma T had significantly declined. However, there was a significant (but weak) correlation between individual peak levels of plasma T (i.e., in April) and the quantitative level of mating activity (time spent mounting and number of mates) measured for the entire subsequent season. Thus. green turtles do not exhibit the ‘“postnuptial” type of testis cycle typical of many temperate-zone turtles, and the levels of plasma androgen may be important for initiating and maintaining sex behavior, although t#hey are not tightly linked during the mating SeaSOIl.

0 1985 Academic

Press, Inc.

Seasonal changes in gonadal histology have been described in a variety of chelonian species (reviewed by Moll, 1979), but there are few direct measurements of the dynamics of circulating hormone levels (reviewed in Licht, 1982). The endocrinology of the chelonian testicular cycle is of special interest because of the apparent seasonal dissociation between spermatogenesis and androgen secretion. The spermatogenic cycle was considered to be “postnuptial” with maximal spermatogenesis occurring in late summer before entry into hibernation, and androgen secretion was postulated to be high after emergence in spring when testes were spermatogenitally inactive, coincident with mating activity. However, this scheme was derived

largely from histological studies and the few data on seasonal cycles in plasma androgen levels are inconsistent (cf. Licht, 1982). Several factors may underlie discrepancies among studies. In addition to valid interspecific differences in reproductive modes, lack of control over collection and blood sampling protocols may lead to artifacts due to “stress” effects associated with capture which may reduce plasma androgen levels (e.g., Licht et ul., 1983). Also, there are few data regarding the individual variability and especially the temporal dynamics (e.g., diurnal rhythms) of plasma steroid levels. In the one chelonian species where this was examined, the magnitude of diurnal fluctuations varied seasonally and were complicated by diurnal

335 00166480185 $1.50 Copyright Q 1985 by Academic Press, Inc. All rights of reproduction in any form reserved.

336

LICHT,

WOOD, AND WOOD

cycles in body temperature (Kuchling, 1982, 1984). Moreover, most seasonal studies in turtles deal with temperate-zone species in which seasonal temperature cycles confuse interpretation of hormonal variations and, even for these species, incomplete information on the actual mating period confounds interpretation of the significance of androgen cycles. Despite their importance from both an economic and conservation standpoint, there is remarkably little information on the annual testicular cycles of any of the tropical sea turtles (Cheloniidae). Although the green sea turtle, Chelonia mydas, is one of the few turtles for which extensive endocrine studies have been performed (Licht, 1980), data still do not permit conclusions regarding annual endocrine cycles, especially for males. Seasonality is well known in the nesting activity of several species of sea turtles (e.g., Rebel, 1974), and, in males of some species like C. mydas there is also evidence of seasonality in mating behavior. Owens (1980) postulated a seasonal testicular cycle for the male green turtle based on a few gonadal samples at one time of year and a very limited sampling of plasma androgen taken over only a few months (Licht et al., 1979), but there have been no systematic studies on male reproduction in this species, The steroid data were of special interest from the standpoint of behavioral endocrinology because circulating androgens were lower during the peak of the copulatory period than they were in the premating season. However, lack of data for other times of year limits interpretation of these data. The relation between thyroid hormone (e.g., thyroxine, T4) and androgen cycles is of special interest because of the suggestion derived from avian studies (Jallageas et al., 1978a,b; Jallageas and Assenmacher, 1979) that the two hormones may be functionally linked; in particular, elevated thyroid hormone may contribute to the suppression of androgen levels. An inverse correlation be-

tween annual peak plasma T4 and declining androgen in the cobra (Bona-Gall0 et al., 1980) and experimental demonstrations of an inhibitory effect of T, on testis activity in lizards (reviewed in Licht, 1984) support the hypothesis that thyroid and gonadal cycles in reptiles might show a similar relation to that described for birds. The present study extends information on the endocrinology of the green turtle, C. mydas, by examining seasonal and diurnal patterns of plasma androgen (testosterone, T) and T, in males. Data are based on adult turtles from an established captive breeding colony on Grand Cayman Island (Wood and Wood, 1980). Previous studies confirmed that the hormonal changes in this captive colony closely paralleled those in wild populations for events surrounding nesting and the levels of androgens in copulating males (Licht et aE., 1979, 1980). This study also examines the quantitative relationship between androgen levels and male sex behavior. MATERIALS

AND METHODS

Animals. The breeding colony of C. mydas at the Cayman Turtle Farm, Grand Cayman Island, and methods for handling and sampling animals, have been described previously (Licht ei al., 1979). Briefly, animals were either farm-reared or long-term wild captives; data are based on animals >9 years old unless stated otherwise. Turtles are housed in large open outdoor pools supplied with running natural seawater. Thus, all animals were exposed to the same ambient conditions. Seasonal fluctuations in water temperatures and photoperiod are shown in Fig. 1. Water temperatures were relatively uniform throughout the year: diurnal fluctuations were only 0.5-X and the maximum annual change was only 3”C, with a peak in July-September. Males were separated from females in January and then reintroduced in early March or April. Copulation typically begins in April and extends through August, but with a peak in activity in May and June. Extensive behavioral observations quantified the amount of time each male spent mounting females and the number of females (mates) selected throughout this season; males show attention only to females that are sexually receptive (Wood and Wood, 1980). Annual cycles. The present study examined both seasonal changes as well as diurnal cycles in plasma T and T, at different times of year. To obtain data on

SEA TURTLE

TESTIS

possible annual rhythms, blood samples were taken at 18 different times, including most months of the year, over 3 years (between March 1978 and September 1982). T4 was measured for the samples collected between April 1981 and May 1982). Seventy-three different males (5-55 males per period) were involved in this blood sampling program, and some animals were sampled several times over the 4.5year period. Blood sampling was routinely performed in the morning (0830-1100 hr). Diurnai cycles. Changes throughout the day were examined in both sexes at three times of year (January, May, and September) corresponding to different phases of the reproductive cycle. Six individuals of each sex were used at each time with some overlap in animals between the three times. In each season, five sequential blood samples were taken at 6-hr intervals from each animal over a single 24-hr period: at - 1000, 1600, 2200, 0400, and 1000 hr. Turtles were corralled in smaller pens within the pond for this study. Blaod handling and analysis. Blood samples were taken from a dorsal cervical sinus (Owens and Ruiz, 1980) and allowed to clot for several hours at room temperature before freezing of serum at - 20”; we previously confirmed that this treatment did not alter hormone levels (cf. Licht et al., 1980). Plasma T and T, were estimated by radioimmunoassay (RIA) as described previously (T, Licht et al., 1979; T,, MacKenzie et a/., 1981). Attempts were also made to measure plasma luteinizing hormone by RIA (Licht et al., 1979), but levels remained below detectable limits (-1 ng LHlml) at all times. Slaughter of ‘adult’ males to thin the herd in January and September provided an opportunity to examine testicular weight and histology. Combined weight of testes and epididymides was taken and tissues were preserved in Formalin, embedded in Paraplast, and stained with hematoxylin and eosin. Statistics. ‘One-way analysis of variance (ANOVA) was used, to test for differences among sampling times and individual times were compared by Duncan’s multiple-range test. Correlation coefficients (r) and ANOVA to test for significance of regressions were performed with the Microstat program (ECOSOFT, Inc.) on a microcomputer. A probability level of 0.05 was taken ta indicate significance in these tests unless stated otherwise.

337

AND THYROID

27 /

, T

n

9

n

hT;I

A

9

0

N

Y 3 I’

FIG. 1. Annual photoperiod (squares, dashed line) and mean maximum water temperatures (circles, solid line) in the holding pond at the Cayman Turtle Farm. Grand Cayman Island.

sistent among years; there were no significant differences between months for different years. Consequently, presentation and discussion of data is simplified by combining months for the different years on a single plot (Fig. 2). A similar annual cycle in plasma testosterone was evident for, the few individuals for which multiple samples were obtained throughout the study period (e.g., Fig. 3) and from the individuals used in the cohort for the diurnal study (Fig. 4). These data reveal a single annual peak in plasma testosterone. This peak typically occurs in the spring, but some individuals reach peaks as early as January (e.g., Fig. 3). Circulating testosterone was minimal (-3 ngiml) in later summer and early fall (August-November) and then began to rise progressively starting in December ‘to a sharp peak in April (-33, 27, and 39 ng! ml for the 3 years). Mean peak plasma, T in April did not differ significantly between captive-wild and farm-reared males. bevels fell rapidly thereafter; for both years for which data are complete, mean values for March and April are similar to one another RESULTS and significantly higher than in May :and ah Annual Cycles other months (P < 0.05; Duncan’s multipleTestosterone. Highly significant differrange test). ences in mean plasma testosterone was evAdditional evidence for the decline after ident among months (P < O.OOl), with a April comes from paired samples on 10 clear annual cycle in each year of study. males bled in April and May 1982. Each one Moreover, the annual cycle was very con- showed a drop between the two times

338

LICHT,

WOOD,

AND

WOOD

which paralleled the average change in the population. There was a highly significant correlation between these two values (u = 0.83, F = 18.2, P = 0.002). Thyroxine. Unlike T, plasma T, was essentially uniform throughout the year (Fig. Z), averaging between 8.9 and 10.5 rig/ml in all months. There were no significant differences in T4 among months (F = 0.99, P = 0.45). Multiple samples for individuals were equally uniform in different months (data not shown). T, in females showed similar uniformity at approximately the same mean levels. Diurnal

Cycles

The diurnal cycles in male plasma testosterone at three different times of year are shown in Fig. 4. No consistent diurnal rhythms were evident at any sampling time. In September, three turtles exhibited a parallel increase in T up to a peak at about midnight but only one (male 213) showed a subsequent decline that could be interpreted as a cycle; the magnitudes of these individual changes were small compared to annual cycles. There was generally little individual fluctuation over the first four sampling times in other animals in September or at the other times of year. There was a trend toward declining T levels in all of the last samples of each series, which corresponded in time of day to the first; this decline may reflect the effects of repeated sampling. Fortuitously, individuals representing the extremes of the variation in plasma testosterone in each month were included in this diurnal study. The results show that this individual variation is not due to temporal fluctuations in hormone levels, i.e., individual levels tended to remain at their starting values (either high or low) throughout the day and night. However, note that individuals were not consistently at the high or low end of the seasonal spectrum, e.g., plasma T in male 106 was typical

FIG. 2. Monthly levels of plasma testosterone (lower graph) and thyroxine (upper graph) in adult male Chelonia my&s. Values represent mean 2 95% confidence limits for the sample sizes shown next to each entry. Data are taken from three different years: 1978 (squares), 1981 (circles), 1982 (triangles).

in September, among the highest in January, but one of the lowest in May. Plasma T, in both sexes also showed little diurnal fluctuations in all three seasons (data not shown). Relation of Plasma Mating Activity

Testosterone

to

Although mating behavior increases as plasma T falls, the levels of peak levels of plasma T attained in the prebreeding period (e.g., April) may influence overall sex behavior. This relationship became apparent when the captive-wild and farm-reared males were treated separately. Although peak plasma T and the number of mates selected were similar for the two stocks, the duration of mating was much greater in

SEA TURTLE

339

TESTIS AND THYROID

captive-wild animals (e.g., 2828 vs 870 min in 1981). No correlation could be found between plasma T and mating indices (duration or number of mates) for the 31 farmreared males sampled in 1981. However, separate analyses for the 24 captive-wild males revealed a highly significant correlation between individual levels of plasma T measured in April and both the total number of females mounted (F = 6.89, r = 0.49, P = 0.015) and the duration of mounting (F = 8.739, Y = 0.533, P = 0.0073) for the entire subsequent breeding season (April-September) in 1981 (Fig. 5). The sample size and range of plasma T value5 were too small for meaningful analysis of this phenomenon in 1982, but a similar trend was evident for the number of mates. Table 1 summarizes gravimetric and histological data for tissues taken from animals slaughtered in September and January 1981 e Animals represented two age cohorts: <8 years old and >lO years old. Compared to the older animals, those under 8 years of age had both absolutely and relatively smaller gonads, a smaller percentage exhibited active spermatogenesis, and there was no difference in either measure between the two sampling periods. The older animals, which presumably best reflect the breeding cycle of the species, show a clear seasonal change in relative testis size, with mass of combined testis and epididymis being signifiqantly (P < 0.025) larger in winter. In winter, testes also appeared to be spermatogenically active in all but 2 of the 19 tissues studied histologically; germinal epithelia were well developed and contained all stages of spermatogenesis; all epididymides contained sperm. One of the two inactive testes ‘@peared to be in the process of spermatogenic regression and the second wa’s comlpletely quiescent. Unfortunately, only a single adult tissue was preserved for histological examination in September; it was fully ,regressed in appearance.

AMJJ

ASONDJ

FMAMJ I

1981

J

1982

MONTH

FIG. 3. Annual cycle in plasma testosterone for three individual male green turtles sampled repetitively over a 2-year period.

DISCUSSION Annual Endocrine Cycle

The male green turtle exhibits pronounced seasonal changes in plasma androgen that are comparable to those observed in the few temperate-zone turtles that have been examined (Licht, 1982). This pattern of androgen and its relation to other aspects of testicular activity do not clearly conform to the postnuptial scheme initially suggested for turtles, Temperate turtles typically show a brief period of ‘testitular growth and spermatogenesis in late summer (MolI, 1979; Licht, 1982), while plasma T may peak at the same time it independently in the spring (Licht, 1982); Although data on the gonadal weight ,an spermatogenic cycle are still incomplete for C. mydas, the limited information obtained here combined w.ith the few observations from wild turtles suggests that androgerrsecretion is not seasonally uncoupled from spermatogenesis. In the captive population, testes were clearly spermatogenically active in winter, immediately before the onset of the mating ,season and at a time when androgen levels are rapidly increasing. A sample of 15 adult wild males slaughtered in January at Chicoia, Mexico had testes which were similar in size and condition to those obtained from the farm stock (M. Cliffton;unpublished data). Testes ifrom two adults killed on the Nicaraguan fee&rag

340

LICHT,

,,q,

SEPTEMBER \

/’ .m’

\

\

\ ‘\

1000

1600

2200

WOOD, AND WOOD

0600

\\

1000

TIME OF DAY (HRS) 4. Plasma testosterone levels over a single 24hr period for groups of six male green turtles sampled at three times of year, corresponding to different phases of the annual androgen cycle (cf. Fig. 2). FIG.

grounds in April were also at early or peak spermiogenesis with large testis showing the beginning of spermiation (Owens, 1980). By contrast, testis mass is clearly reduced in late summer when androgens are at their nadir (Table 1). Thus, this turtle appears to show a “prenuptial” testicular weight aqd androgenic cycle that is similar to that of lizards, with a coincidence between spermatogenic and androgenic activities. Diurnal Rhythms

Hormone levels may be highly labile and diurnal rhythms may confound interpretation of samnles taken at onlv2 one time. , I

especially if peak values occur at different times in different seasons. For example, in mammals, photoperiod may change the phase relationships from diurnal to nocturnal peaks in androgen, and maxima tend to occur when animals are inactive (Rowsemitt and Berger, 1983). There was a hint of a nocturnal peak in the turtles sampled in September (Fig. 4). Daily T, rhythms have been reported in a variety of mammals and nonmammalian species, but there is no consistent time when the peak occurs (e.g., Kuhn et al., 1983). The sampling protocol used here was not designed to address the issue of a short-term pulsatile release of hormones, but there is little evidence for a distinct daily rhythm in either plasma T or T, in the green turtle. Studies by Kuchling (1984) suggested that diurnal cycles in the tortoise may occur only at certain times of this year; we observed a slightly greater tendency for a diurnal pattern in September than at the other two times, but even then, it was inconsistent. For most individuals the variation between times (except for the last sample) was within the repeatability of the RIA. The drop in plasma T at the end of the sampling period could be attributed to the effects of repeated sampling as suggested in the tortoise study by Kuchling (1984). The greater stability of androgen levels in the green turtle compared to that in the tortoise may be related to the greater thermal stability in this study. It should be noted, however, that the green turtle does show distinct and typical diurnal cycles in plasma melatonin (Owens et al., 1980). While androgen levels may not be absolutely constant throughout a day, these data co&m that single daily samples are adequate for discerning seasonal androgen cycles in the green turtle, and for identifying individual variation within the population. Mechanisms Underlying the Annual Gonadal Cycle

From the standpoint of environmental stimuli, minimal androgen secretion coin-

SEA TURTLE

f

TESTIS

341

AND THYROID

= 0.53

P = 0.007 0

0

6

e 0

0

0 0

0 0 8

0 0

0 0 0 0 6 0

‘-k-++r

---x7---5-74[j--TESTOSTERONE

PLRSMR

50

(NG/MLI

FIG. 5. Relationship between plasma testosterone measured at the peak of the androgen cycle in April 1981 and two measures of the intensity of mating behavior exhibited over the following breeding season (April-September) in a group of 24 captive-wild stock males. Values indicated by “ x ” in the lower figure did not show any mounting behavior. Correlation coefficients (v) and corresponding probability values (p) from regression analysis are indicated for each.

tides with the highest ambient temperature and a decline in day length, although the drop in androgen levels occurs much earlier. The only experimental evidence regarding the environmental regulation of chelonian gonadal cycles deals with a temperate species and implicates temperature as the main proximate cue controlling testicular recrudescence; photoperiod had no apparent effect in Chrysemys picta (Ganzhorn and Licht, 1983). Although thermal changes in the present study are small by comparison to those in temperate habitats,

their impact on the endocrine cycie cannot be discounted, since the reproductive system of tropical reptiles may be hi$hly temperature sensitive (Licht, 1973). Bowever, if temperature changes mediate, the testis cycle of C. mydas, their effect is distinct from that in C. picta in which warm temperatures accelerated gonadal growth (Ganzhorn and Licht, 1983) The regularity of the androgen cycle from year to year argues for some synchronization by a refatively constant external cue, but in ‘the turtle farm environment, temperature “and

342

LIGHT,

WOOD,

AND

TABLE RELATIVE

Season January September

TESTIS

WOOD

1

MASS (GSI) AND SPERMATOGENICACTWITY IN YOUNG (5.5-7.5 YEARS) AND ADULT (11-14 YEARS) CheIonia nzydus Age

Young Adult Young Adult

Body wt (kgY

GSI Wkd

16 22 17

96.5 -c 3.8 141.4 _’ 6.8 109.6 t 5.4”:

0.51 i 0.11 3.08 + 0.25 0.69 _e 0.19

4115 17119 114

145

151.4 i 7.3

1.33 t 0.2*

Oil

N

Proportion activeb

a Means t SEM for the entire sample (N). b The number of animals in which seminiferous tubules appeared to be in maximal phases of spermatogenic activity/number of animals examined histologically. * Significantly (P < 0.025) different from animals of same age group in winter sample.

photoperiod are probably equally regular in this regard. Since most, if not all, of the male population showed active spermatogenesis in January and a marked elevation of plasma androgen in spring, we may conclude that the males have an annual cycle. Information on breeding effort in females suggest that caution must be exercised in extrapolating these results to natural populations in which nutritional conditions may be inferior (Wood and Wood, 1980). The potential involvement of thyroid hormone in the regulation of the testicular, or at least the androgen, cycle demonstrated for the cobra (Bona-Gallo et al., 1980) is not evident in the green turtle. In the cobra, rising T4 levels offered an explanation for why the drop in plasma androgen preceded the fall in gonadotropin. Unlike this snake, T, levels remain very uniform throughout the year and there is no indication of a rise that might explain the abrupt drop in androgen levels after May. Comparative data for reptiles are too sparse to allow any generalizations about T, cycles at this time. For example, in contrast to the cobra, Sellers et al. (1982) reported a subtle change in T, in one lizard, with highest IeveIs during hibernation, whereas JohnAlder (1983) found a bimodal cycle, with early and late summer peaks in another lizard. Again, all of these previous data concern temperate-zone species in which major temperature changes are involved and any relationship between androgen and

T4 cycles may simply reflect their independent response to such environmental changes rather than a causal relationship between the two. Relation between Androgen Behavior

and Sex

A lack of precise information on the exact time, and certainly the magnitude, of sexual activity in the few species of turtles studied previously seriously limits interpretation of the potential role of androgen in chelonian sex behavior (Licht, 1982). It has been assumed that androgens would be maximal during the mating season, although direct evidence for a dependence of sex behavior on androgen in turtles is limited. Indeed, testosterone treatment of sexually immature green turtles induced elongation of the penis and tail and precocious mating attempts (Owens, 1980). However, mating activity cannot be taken as an indicator of circulating androgen since the present results confirm previous data (Licht et al., 1979) which show that copulatory behavior in the green turtle occurs well after the peak in circulating androgen levels. Indeed, it is now evident that the most intense mating behavior (e.g., June and July) occurs as androgen levels approach their nadir and mating may even continue sporadically into September when plasma T is minimal. Thus, androgens clearly need not be maximally elevated to support mating behavior.

SEA TURTLE

TESTIS

Since the attractiveness of females appears to vary seasonally, depending on their receptivity (Wood and Wood, 1980), the level of male sexual “‘arousal” cannot be accurately assessed at all times of year. However, since all males had access to the same large pool of females, individual variation in male sexual activity during the breeding season presumably reflects the condition of the male. Since females are generally unreceptive before March, it is not clear that the rise in androgen level up through April is essential to initiate male sexual behavior. However, the correlation between individual plasma T values determined at the time of the peak in April and the subsequent intensity of copulatory activity of captive-wild males during the next 4-5 months (Fig. 5) suggests that the early peak in testosterone may have important behavioral consequences. This correlation, albeit weak, is impressive in that it is based on a single measurement of plasma T. The relative stability of plasma T over time may contribute to the value of such single measurements. The lack of a marked diurnal rhythm (Fig. 4) confirms the validity of estimating individual variation on the basis of single daily samples. While we cannot rule out the possibility of major fluctuations over days or weeks, the demonstration of a hiih correlation between individual levels measured a8month apart (in April and May) also suggests that reIatively infrequent sampling may be adequate to discern individual variability and seasonal trends. Thus, elevated testosterone may be generally important for the initiation of the mating season and subsequent expression of seg behavior, although high circulating androgens are clearly not required for continued sexual activity. Also, the ,high levels of testosterone is not by itself sufficient to ensure mating activity. ACKNOWLEDGMENTS We thank Mr. Raymond Pang for technical assistance in hormone assays and Dr. D. Owens for reviewing this manuscript. This work was supported in

AND THYROID

343

part by a grant (PCM 8109846) from the National Sci” ence Foundation.

REFERENCES Bona-Gallo, A, Licht, P., MacKenzie, D. S., and Lofts, B. (1980). Annual cycles in levels of pituitary and plasma gonadotropin, gonadal steroids, and thyroid activity in the Chinese cobra (Naju naja). Gen. Cornp. Endocrinol. 42, 477-493. Ganzhorn, D., and Licht, P. (1983). Regulation of seasonal gonadal cycles by temperature in the painted turtle Chvysenzys pictn. Copein 1983, 347-358. Jallageas, M.. and Assenmacher, I. (1979). Further evidence for reciprocal interactions between the annual sexual and thyroid cycles in male Peking ducks. Gen. Comp. Endocrinal. 37, 44-51. Jallageas, M.. Astier, H., and Assenmacher, I. (1978a). Thyroid-gonadal interactions during the postnuptial phase of the sexual cycle in maie ducks. Gen. Cornp. Endocrinol. 34, 68-69 (Abstract). Jallageas, M., Tamisier, A., and Assenmacher, I. (197&b). A comparative study of the annual cycles in sexual and thyroidal function in male Peking ducks (Anas platyrhynchos) and teal (Arms crecca). Gen. Comp. Endocrinol. 36, 201-210. John-Alder, H. B . (1983). “The Physiological Basis of Activity in Lizards: Influence of Body Temperature and Thyroid Hormones.” Ph.D. Thesis. Univ. California, Irvine. Kuchling, G. (11982). Environmental temperature, SpermatQgenesis, and plasma testosterone concentration in the tortoise Testudo hermanni hermanni Gremlin. Acta Endocrinol. 99, (Suppl. 246), 29-30. Kuchling, G. (1984). Seasonal and diurnal fluctuation of the plasma testosterone concentration in the male tortoise, Testudo hermanni hermanni Gremlin. Gen. Cornp. Endocrinol., in press. Kuhn, E. R., Delmotte, N. M. J., and Darras, V. M. (1983). Persistence of a circadian rhythmicity for thyroid hormones in plasma and thyroid of hibernating male Rana ridibunda. Gen. Camp. Endocrinol. 50, 383-394. Licht, P. (1973). Thermal and photic influences on reptilian reproduction. & “Int. Endocrinol. Congr., Washington, D+C. Excerpta Med., Int, Congr. Ser.” Vol. 273, pp. 185-190. Licht, P. (1980). Evolutionary and functional aspects of pituitary gonadotropins in the green turtle, Chelonia mydas. Amer. ZooE. 20, 56.5-574. Licht, P. (1982). Endocrine patterns in the reproductive cycles of turtles. Herperologica 38, 51-61. Licht, P. (1984). Reptiles. In. “MarshalI’s Physiology of Reproduction” (G. E. Lamming, ed.). 4th ed.,

344

LICHT,

WOOD, AND WOOD

Vol. A, Chap. 3, pp. 206-282. Churchill-Livingstone, Edinburgh. Licht, P., Rainey, W., and Cliffton, K. (1980). Serum gonadotropin and steroids associated with breeding activities in the green sea turtle Chelonia mydas. II. Mating and nesting in natural populations. Gen. Comp. Endocrinol. 40, 116-122. Licht, P., Wood, J., Owens, D. W., and Wood, F. (1979). Serum gonadotropin and steroids associated with breeding activities in the green turtle Cheloaia mydas. I. Captive animals. Gen. Comp. Endocrinol. 39, 274-289. Licht, P., McCreery, B. R., Barnes, R., and Pang, R. (1983). Seasonal and stress related changes in plasma gonadotropins, sex steroids and corticosterone in the bullfrog, Rana catesbeiana. Gen. Comp. Endocrinol. 50, 124-145. MacKenzie, D. S., Licht, P., and Papkoff, H. (1981). Purification of thyrotropin from the pituitaries of two turtles: The green sea turtle and the snapping turtle. Gen. Comp. Endocrinol. 45, 39-48. Moll, E. 0. (1979). Reproductive cycles and adaptations. In “Turtles: Perspective and Research” (M. Harless and H. Morlock, eds.), pp. 30.5-331. Wiley, New York.

Owens, D. W. (1980). The comparative reproductive physiology of sea turtles. Amer. Zool. 20, 549563. Owens, D. W., and Ruiz, G. J. (1980). New methods of obtaining blood and cerebrospinal fluid from marine turtles. Herpetologica 36, 17-20. Owens, D. W., Gem, W. A., and Ralph, C. L. (1980). Melatonin in the blood and cerebrospinal fluid of the green sea turtle (Chelonia mydas). Gen. Comp. Endocrinol. 40, 180-187. Rebel, T. P. (1974). “SeaTurtles.” Univ. Miami Press, Coral Gables, Fla. Rowsemitt, C. N., and Berger, P. J. (1983). Die1 plasma testosterone rhythms in male Microtus montanus, the Montane vole, under long and short photoperiods. Gen. Camp. Endocrinol. 50, 354-358. Sellers, J. C., Wit, L. C., Ganjam, V. K., Etheridge, K. A., and Ragland, I. M. (1982). Seasonal plasma T4 titers in the hibernating lizard Cnemidophorus sexlineatus. Gen. Camp. Erzdocrinoi. 46, 24-28. Wood, J. R., and Wood, E E. (1980). Reproductive biology of captive green sea turtles Chelonia mydas. Amer. Zool. 20, 499-506.