Effects of the antiandrogens cyproterone acetate and flutamide on male reproductive behavior in a lizard (Anolis sagrei)

HORMONES

AND

BEHAVIOR

21, 1-16 (1987)

Effects of the Antiandrogens Cyproterone Acetate and Flutamide on Male Reproductive Behavior in a Lizard (Anolis sagrei) RICHARD

R. TOKARZ

Department of Biology, University of Miami, Coral Gables, Florida 33124 This study examined the effects of the antiandrogens cyproterone acetate (CA) and flutamide (F) on male reproductive behavior in the lizard Anolis sagrei. Reproductively active males were implanted with subcutaneous pellets of either CA, F, or placebo (P). Pellets delivered CA and F at a constant rate of 0.1 mg/day. Three weeks after implantation, males were tested with stimulus males and two days later with stimulus females. Cyproterone acetate inhibited aspects of male aggressive and sexual behavior, and reduced testis weight and size of the renal sex segment. Plasma testosterone (T) levels in CA-treated males were not significantly different than those of P-treated males. Flutamide did not inhibit aggressive or sexual behavior, but did decrease testis weight as well as the size of the renal sex segment. Plasma T levels were significantly higher in F-treated males than in P-treated males. These data suggest that CA, an antiandrogen with antigonadotropic activity, may be used to inhibit reproductive behavior in male lizards. o 1xi7 AC&YGC &SS, IX.

Antiandrogenic compounds can be divided into two functional classes based on whether they have only antiandrogenic activity or whether they also possess antigonadotropic properties (Neumann, G&f, Hasan, Schenck, and Steinbeck, 1977). Antiandrogens with only antiandrogenic activity interfere with the androgen negative feedback system controlling gonadotropin release and consequently long-term treatment with these compounds increases gonadotropin release and increases circulating androgen levels (Neuman et al., 1977). On the other hand, antiandrogens with antigonadotropic activity have properties which inhibit gonadotropin release, and consequently long-term treatment with these compounds does not increase circulating androgen levels (Neumann et al., 1977). Because of their ability to compete for androgen receptors, both functional classes of antiandrogens have been employed to investigate how androgens mediate male reproductive behavior (Steinbeck and Neumann, 1975; Davidson, Smith, and Damassa, 1977; Neumann et al., 1977). FMamide, an antiandrogen that possessesonly antiandrogenic activity, has proven to be ineffective in inhibiting male reproductive behavior in 0018-506X/87 $1.50 Copyright 0 1987 by Academic Press, Inc. AN rights of reproduction in any form reserved.

2

RICHARD

R. TOKARZ

adult guinea pigs and rats, although it markedly inhibits androgen-dependent accessory reproductive structures (Sodersten, Gray, Damassa, Smith, and Davidson, 1975; Davidson et al., 1977; Neumann et al., 1977). In contrast to these negative findings from studies of mammals, Searcy and Wingfield (1980) found that male red-winged blackbirds implanted with flutamide were less aggressive and had significantly lower dominance ranks than did control males. Based on their findings and the results from other studies of birds (see Adkins and Mason, 1974; Silver, 1977), Searcy and Wingfield (1980) suggested that the nervous system of birds may be more susceptible to the inhibitory effects of antiandrogens than is the nervous system of mammals. The ability of antiandrogens with antigonadotropic activity to inhibit male reproductive behavior also appears to vary between mammals and birds. For example, treatment with cyproterone acetate, an antiandrogen with progestational properties and thus antigonadotropic activity, has been found to be ineffective in inhibiting agressive and/or sexual behavior in adult male mammals (Zucker, 1966; Beach and Westbrook, 1968; Whalen and Edwards, 1969; Whalen and Luttge, 1969; Edwards, 1970; Bloch and Davidson, 1971; Sayler, 1970; but see Steinbeck, Elger, and Neumann, 1967). On the other hand, cyproterone acetate has been found to inhibit male reproductive behavior in birds. For example, copulatory behavior in male Japanese quail and bow-cooing as well as nest soliciting behavior in male ring doves are reduced by cyproterone acetate treatment (Adkins and Mason, 1974; Silver, 1977). This study had three objectives. The first was to test whether the antiandrogenic compounds, cyproterone acetate and flutamide, can inhibit male reproductive behavior in an anoline lizard (Anolis sagrei). Such data would be useful in determining whether a significant difference exists in the ability of antiandrogens to inhibit male reproductive behavior between mammals and nonmammalian vertebrates. Reptiles are a pivotal group in this regard because their progenitors gave rise to birds and mammals. The second objective was to compare the relative abilities of cyproterone acetate and flutamide to affect reproductive behavior. The last objective was to establish whether cyproterone acetate and/or flutamide can be employed to alter male social dominance in Anolis lizards, a genus where males can be under intense intrasexual selection to obtain and defend territories (see Stamps, 1983). MATERIALS AND METHODS Subjects and Housing

Adult male and female A. sagrei were collected from the Miami area during June 1985, a period when this species is reproductively active (Licht and Gorman, 1970). Lizards were brought to the laboratory and housed in an environmentally controlled animal room. This room had a

ANTIANDROGENS

AND LIZARD

BEHAVIOR

3

14L:lOD photoperiod, and the temperature varied from 30°C t 1°C during the photophase to 24°C + 1°C during the scotophase. Average relative humidity was 50-55% during midday and 60-65% during midevening. These conditions are similar to those of Miami during the early spring and summer months (Terhune, 1983). Males were housed singly in holding cages (20 cm wide x 32 cm high x 32 cm long) or in testing cages (30 x 42 x 32 cm). Each cage contained a food and water dish, and a twig (holding cages) or dowel (testing cages) for perching. Female lizards were housed in glass aquaria (32 x 42 x 61 cm) which contained food and water dishes, twigs, and potted plants (Aglaonema). Lizards were fed ad libitum a diet of meal worms supplemented with crickets. Other details of housing are described in Tokarz (1985). Procedures Males were assigned randomly to either stimulus or treatment categories. Treatment males received a single subcutaneous pellet containing either flutamide (F), cyproterone acetate (CA), or placebo (P). Pellets were purchased from Innovative Research of America (Rockville, Md.). The F and CA were put into pellet form as a custom service. These pellets have been demonstrated in studies with mammals to deliver hormones or drugs at a dose-dependent controlled rate. Each pellet contained 2.1 mg of either F or CA and delivered antiandrogen for 3 weeks at a controlled rate of 0.1 mg/day. A pellet was implanted by making a small incision behind the shoulder of a cold-anesthetized male and placing the pellet subcutaneously with surgical forceps. The incision was sutured closed. Each treatment male was placed in a testing cage following pellet implantation. Males were assigned to testing cages such that every three cages contained a male from each of the three treatment groups. This pattern of placement was repeated until all testing cages contained a male. Alternation among treatment groups was done to minimize positional effects of cage location and/or changes in behavioral responses due to the time of day when testing was conducted. Behavioral Tests Behavioral tests between treatment males and stimulus males were conducted on Day 22, 3 weeks after pellet implantation. Behavioral tests between treatment males and stimulus females were run on Day 24. No attempt was made to determine whether a female was sexually receptive prior to its use. Each behavioral test was 15 min long and began when a stimulus lizard was placed manually into a testing cage housing a treatment male. Stimulus lizards were removed at the end of the test. A different stimulus lizard was used for each test. Behavioral tests were run in sequence beginning with the first cage and continuing sequentially until all treatment males had been tested.

4

RICHARD

R. TOKARZ

Videotaping

The behavioral responses of the treatment males were videotaped using a Panasonic time-lapse videotape recorder, video timer, and a tripodmounted Panasonic TV camera with Sony zoom lens. During videotaping, a video monitor was used to watch the movements of the treatment male and to guide appropriate camera adjustments. Recording was at 60 fields/set in normal time. Videotapes were transcribed at normal time or more slowly. Behaviors

Transcribed from Videotapes

In tests with stimulus males, the number of males in each treatment group that displayed, the number that approached, the number that erected a nuchal and dorsal crest, and the number that showed biting behavior were recorded. An approach was defined as movement of more than one body length by the treatment male toward the stimulus animal. The latencies in minutes to the first display, to the first approach, and to the first biting behavior were also recorded. To quantify display behavior, the total number of three display types was recorded. These were (1) dewlap fanning without postural modiflers, (2) dewlap fanning with postural modifiers, and (3) postural movements only without dewlap fanning. The dewlap fanning display without postural modifiers was defined as the extension and retraction (i.e., fanning) of the dewlap in the absence of dynamic postural modifiers such as headbobs or body pushups. The dewlap fanning with postural modifiers display was defined as dewlap fanning accompanied by headbobs and/or pushups. Finally, the postural movements without dewlap fanning display was defined as the occurrence of headbobs and/or pushups in the absence of dewlap fanning. Although all three of these displays were sometimes performed in the absence of static modifiers such as nuchal and dorsal crests and lateral body compression, these static modifiers usually appeared in the more intense aggressive encounters between males. These three display types are based on the dewlap only, bob only, and bob plus dewlap displays described in Scott (1984). Displays were counted as separate displays if they were separated by at least a 3-set interval with no display behavior. In tests with stimulus females, the number of males in each treatment group that displayed, the number that approached, the number that courted, the number that obtained a neck grip, and the number that achieved intromission were recorded. A male was delined as courting if it performed a series of rapid, low amplitude movements of the head (i.e., jiggle-bob) as it approached the female. The occurrence of jiggle-bobs is a highly characteristic feature of anoline courtship display (Crews, 1975; Scott, 1984). The latencies in minutes until the appearance of these behaviors were also noted. The total number of the three display types (see above) was recorded during the 15-min tests or until the male obtained a neck

ANTIANDROGENS

AND LIZARD BEHAVIOR

5

grip of the female. The displays of males in tests with females did not include crests or lateral compression: static display modifiers which are characteristic of aggressive encounters between males. Physiological Measurements

Following completion of behavioral testing, each treatment male was weighed and plasma samplesfor testosterone (T) and dihydrotestosterone (DHT) radioimmunoassay were obtained by cutting the tail with a scalpel blade. Blood was collected into heparinized microhematocrit capillary tubes and the plasma separated from the cellular components of blood by centrifugation. Plasma samples were stored at -70°C before being shipped in dry ice to Hazleton Biotechnologies Corporation (Vienna, Va.) where the T and DHT radioimmunoassayswere performed. Prior to assay, T was separated from DHT by using phase partition chromatography with celite columns. Males were sacrificed by decapitation after being bled. The left testis from each male was removed and weighed. The renal sex segment was also visually examinedusing a Bausch and Lomb Stereo-Zoommicroscope. The renal sex segment was classified as being either stimulated or unstimulated. This classification was based on whether the extent of renal sex segment hypertrophy more closely matched that observed in reproductively active males or that observed in reproductively inactive males. Statistical Analyses

Frequency data were analyzedby using a computer program to calculate the exact cumulative probability of obtaining an outcome as extreme or more extreme than the one obtained (see p. 709, Sokal and Rohlf, 1981). This procedure was used instead of a G test becausesomeof the expected outcomes in 2 x 2 contingency tables were less than five (Sokal and Rohlf, 1981). Latencies of behavioral responses were compared using the median test with Fisher’s exact method (Downie and Heath, 1965). Display behavior data and plasma T levels were compared with MannWhitney U tests. Analysis of variance (ANOVA) and Student-NeumanKeul’s procedure were employed to analyze testis weight data as these were normally distributed and homoscedastic.To minimize the possibility of making a type I error, comparisons were made only between the placebo-treated group and each one of the two antiandrogen-treatment groups. Results were considered to be significant if P s 0.05 (one-tailed test). RESULTS Aggressive Behavior

The aggressivebehavior of F-treated males in tests with stimulus males did not differ significantly from that of P-treated males (Tables 1-4,

RICHARD R. TOKARZ TABLE 1 Effects of Cyproterone Acetate and Flutarnide on the Aggressive Behavior of Male Anolis sagrei during 15min Tests with Stimulus Males Number of males Treatment group”

Displaying

Approaching

Erecting a crest

Biting

12 12

10 10

9 6

8 10

12

8

Placebo Flutamide Cyproterone acetate

Ob

3b

* Each treatment group had 12 animals. b Significantly fewer than in the placebo group (exact test).

Fig. IA). Cyproterone acetate-treated males, on the other hand, showed significantly less behavior in some of the high intensity components of aggressive behavior. Significantly fewer CA-treated males erected a crest (P < 0.001, exact test) or attempted to bite (P < 0.05, exact test) than did P-treated males (Table 1). Significantly more CA-treated males than P-treated males had a latency to approach (Table 2) that was longer than the group median for the two treatment groups (P < 0.025, median test using Fisher’s exact method). The median latency to bite in the CA treatment group was also longer than in the P treatment group, although this difference could not be tested statistically because there were insufficient values to calculate the group median which is required for the median test. Although the mean number of total displays in the CA treatment group was not significantly lower than in the P treatment group TABLE 2 Effects of Cyproterone Acetate and Flutamide on the Latency to the Appearances of Aggressive and Sexual Behavior in Male Anolis sagrei during 15-min Tests with Stimulus Males and Females Median latency (min)* Treatment group” Placebo Flutamide Cyproterone acetate

Median latency (mitt)

To display

To approach

To bite

0.78 1.02

0.78 0.48

3.48 0.78

0.91

7.W

>15.00

To display

To approach

To court

0.71 0.71

2.24 1.04

6.92 >15.00

2.54d

9.78“

>15.00

o Each treatment group had 12 animals. b In tests with stimulus males. ’ In tests with stimulus females. d Significantly longer than in the placebo group (median test).

ANTIANDROGENS

AND LIZARD BEHAVIOR

TABLE 3 Effects of Cyproterone Acetate and Flutamide on the Display Behavior of Male Anolis sagrei during 15-min Tests with Stimulus Males Type of displayb

Dewlap fanning without postural modifiers

Dewlap fanning with postural modiflers

Postural movements without dewlap fanning

Treatment group”

Mean no. of displays

Mean no. of dewlap extentions

Mean no. of displays

Mean no. of dewlap extensions

Mean no. of displays

Placebo Flutamide Cyproterone acetate

8.5 f 1.8 6.3 + 1.9

44.7 f 11.4 24.7 f 7.8

3.2 ” 1.0 2.4 e 0.7

17.3 f 5.5 15.3 2 5.4

2.7 2 0.8 4.4 f 1.2

7.8 f 1.9

32.4 f

8.4

0.8’ f 0.3

4.3 k 1.8

1.3’ f 0.4

’ Each treatment group had 12 animals. b Each mean is followed by the standard error of the mean. ’ Significantly less than in the placebo group (Mann-Whitney U test).

(Fig. lA), analysis of the three display types (Table 3) indicated that CA-treated males showed signikantly fewer dewlap farming with postural modifier displays (U = 41, N1 = ZVz = 12, P < 0.05) and postural movements without dewlap fanning displays (U = 41, N1 = ZV2= 12, P c 0.05). Social dominance was also adversely affected by CA treatment (Table 4). The number of CA-treated males supplanted from their perch (7/12) at the end of the 15min behavioral tests was significantly greater (P < TABLE 4 Effects of Cyproterone Acetate and Flutamide on Perch Site Dominance in Male Anolis sagrei during 15-min Tests with Stimulus Males Location of males relative to perchb Treatment group” Placebo Flutamide Cyproterone acetate

Treated male on perch

Stimulus male on perch

Both males on perch

Neither male on perch

10 10

1 1

0 1

1 0

7d

2

1

2d

’ Each treatment group had 12 animals. b At the end of the 15-min tests. ’ Number of occurrences in 12 tests. d Signitkantly different from the placebo group (exact test).

RICHARD R. TOKARZ IA

T a

h

12

12

F

CA

FIG. 1. Mean total display frequencies for male Anolis sagrei in different treatment groups during 15-min tests with stimulus males (A) and stimulus females (B). Treatment groups were placebo (P), flutamide (F), and cyproterone acetate (CA). In tests with stimulus females, the display frequencies for those males that obtained a neck grip were calculated by dividing the total number of displays that occurred prior to the neck grip by the elapsed time from the start of the test until the neck grip. Vertical bars above and below each mean represent the standard error of the mean. The number of males is shown at the base of each bar. A letter at the top of a bar indicates that the group had a mean display frequency significantly less than that of the placebo group (P < 0.05, Mann-Whitney U tests).

0.01, exact test) than the number of P-treated males supplanted from their perch (l/12). Correspondingly, the number of CA-treated males that had sole occupancy of their perch (2/12) at the end of the tests was significantly less (P < 0.01, exact test) than the number of P-treated males that had exclusive occupancy of their perch (10/12). Sexual Behavior The sexual behavior of F-treated males in tests with stimulus females did not differ significantly from that of P-treated males (Tables 2, 5, Fig. 1B). There were, however, significant differences between the CA and P treatment groups. For example, significantly fewer CA-treated males (P < 0.05 exact test) approached stimulus females than did P-treated males (Table 5). In addition, significantly more CA-treated males than P-treated males had a latency to display or to approach (Table 2) that was longer than the group medians (P C 0.025, median test using Fisher’s exact method). Although the median latencies to court for both the F and CA treatment groups were longer than the median latency for the P group, it was impossible to evaluate these differences statistically

ANTIANDROGENS

9

AND LIZARD BEHAVIOR

TABLE 5 Effects of Cyproterone Acetate and Flutamide on the Sexual Behavior of Male Anolis sagrei during 15-min Tests with Stimulus Females Number of males Treatment group” Placebo Flutamide Cyproterone acetate

Displaying

Approaching

Courting

Neck g~ppinis

Intromitting

12 11

8 6

5 6

3 6

4

2

2

12 11 9

86

’ Each treatment group had 12 animals. b Significantly fewer than in the placebo group (exact test).

becausethere were too few valuesto calculategroup medians.Cyproterone acetate treatment also had a negative effect on sexual display behavior (Fig. 1B) as the mean number of displays in CA-treated males was significantly less than in P-treated males (U = 38.5, ZV, = IV2 = 12, P < 0.05). Physiological Measurements

Body weights at the end of the experiment did not differ significantly among treatment groups. There was, however, a signifkant treatment effect on left testis weight (F2,t2 = 5.93, P < 0.01). Left testis weight was significantly lower in both F-treated (P < 0.05) and CA-treated (P < 0.01) males than in P-treated males (Fig. 2). However, this difference in testis weight was greater in CA-treated males than in F-treated males. The degree of stimulation of the renal sex segmentsin F-treated males

a b

P

12 lh F

12 CA

FIG. 2. Mean left testis weight for male Anolis sagrei treated with either placebo (P), flutamide (F), or cyproterone acetate (CA). Vertical bars above and below each mean represent the standard error of the mean. The number of males is shown at the base of each bar. A letter at the top of a bar indicates that the group had a mean left testis weight signitlcantly less than that of the placebo group (a = P < 0.05, b = P < 0.01; StudentNeuman-Keul’s test).

10

RICHARD R. TOKARZ

and CA-treated males differed significantly from P-treated males. Significantly fewer males had stimulated renal sex segmentsin the F treatment group (7/12, P < 0.05, exact test) and in the CA treatment group (4/12, P < 0.001, exact test) than in the P treatment group (12/12). Plasma T levels in CA-treated males were not significantly different from those of P-treated males (Fig. 3). But plasma T levels in F-treated males were significantly higher than in P-treated males (U = 20, N, = N2 = 12, P < 0.001). The data on plasma levels of DHT in each of the treatment groups were not compared statistically because a majority of males had nondetectable amounts of this androgen. The mean plasma levels for the males of each treatment group with detectable amounts of DHT were: F-treated males (1.26 rig/ml t 0.26 SEM, 12 = 5) and P-treated males (0.73 ng/mi & 0.07 SEM, IZ = 3). All CA-treated males had nondetectable amounts of DHT. DISCUSSION Cyproterone acetate treatment inhibited components of male aggressive and sexual behavior in the lizard A. sagrei. Because T and DHT, but not estradiol, have been found to restore reproductive behavior in castrated male A. sagrei (Tokarz, 1986), it is likely that these inhibitory effects of CA are the result of its antiandrogenic activity. In tests with stimulus males, significantly fewer CA-treated males than P-treated males erected a crest or showed biting behavior. Males treated with CA did not, however, differ from P-treated males in terms of the number of males that displayed or approached. Because crest erection and biting behavior normally occur only in those agonistic encounters between males that have escalated beyond initial assertion or threat displays (Ortiz and Jenssen, 1982;

fL 12

CA

FIG. 3. Mean plasma testosterone levels for male Anofis sagrei treated with either placebo (P), flutamide (F), or cyproterone acetate (CA).‘Vertical bars above and below each mean represents the standard error of the mean. The number of males is shown at the base of each bar. A letter at the top of a bar indicates that the group had a mean plasma testosterone level significantly higher than that of the placebo group (P < 0.001, Mann-Whitney U test).

ANTIANDROGENS

AND LIZARD

BEHAVIOR

11

Greenberg and Crews, 1983; Scott, 19&I), these behaviors probably reflect an increase in motivational state. As such, crest erection and biting behavior may be more androgendependent than is dewlap display behavior of a resident male that occurs when it is initially exposed to an intruding male. This hypothesis is supported by the finding that castrated male A. sagrei, when paired with stimulus males, do not erect a crest and bite infrequently, although they do continue to dewlap display (Tokarz, 1986). Males treated with CA also had longer latencies to approach and to bite than P-treated males, again suggesting that CA-treated males are less likely to escalate agonistic encounters, or at least to escalate encounters more slowly than are P-treated males. Mason and Adkins (1976) have also suggested that androgen dependency of male reproductive behavior may vary with the intensity of the behavioral responses. For example, the intense levels of aggressive behavior observed in male A. cadinensis during the breeding season are believed to be more androgen-dependent than the less intense behaviors typically seen in males during the nonbreeding season (Mason and Adkins, 1976). Although total display frequency did not differ significantly between CA-treated and P-treated males, analysis of the three component display types indicated that CA-treated males performed fewer dewlap fanning with postural modifier displays and fewer postural movements without dewlap fanning displays. This is noteworthy because it suggests that CAtreated males performed fewer of the display types that have been found to be given by the more dominant male in an agonistic encounter between two males (see Scott, 1984). The fact that CA inhibited crest erection and biting, also suggests that a male’s ability to compete with other males during agonistic encounters might be androgen dependent. Additional evidence supporting the hypothesis that androgens play an important role in determining the outcome of agonistic contests between males comes from the finding that CA treatment significantly decreased the number of males that were able to retain their perch site. Male anoles use their perch for surveillance and for social interaction, and consequently males compete for territories with a conspicuous perch site (see Scott, Wilson, Jones, and Andrews, 1976). Resident males are usually able to defend their perches from similar-sized intruders (Ruibal, 1967). This was not true for CA-treated male A. sagrei where a highly significant number of resident males were supplanted by stimulus males. This is an important finding because it suggests that androgens are required for successful territorial defense. Cyproterone acetate also inhibited components of male sexual behavior. Significantly fewer CA-treated than P-treated males approached stimulus females; they also had significantly longer approach latencies than Ptreated males. Although fewer CA-treated males courted, obtained a neck grip, or intromitted than did P-treated males, these differences were not

12

RICHARD

R. TOKARZ

significant. It should be noted, however, that most stimulus females were nonreceptive (i.e., ran from the approaching male). Thus, even in the P treatment group very few males obtained a neck grip or intromitted. This fact makes it difficult to detect differences, if these exist, between Ptreated males and CA-treated or F-treated males. Although it would have been possible to produce receptive females by estradiol treatment, or to select for receptive females by conducting a pretest, these methods were not employed because a large proportion of female anoles in nature are normally nonreceptive at any given time (Ruby, 1984). Hence, it is likely that males in nature court nonreceptive as well as receptive females. The total display frequency during tests with stimulus females was significantly lower in CA-treated males than in P-treated males. Thus, CA significantly inhibited the display frequency of resident males in tests with stimulus females but not in tests with stimulus males. Courtship display behavior has been found to decline much more quickly than does aggressive display behavior following castration of male A. carolinensis (Crews, 1979a,b). This finding suggests that the maintenance of courtship display behavior requires higher circulating androgen levels than does the maintenance of aggressive display behavior. Cyproterone acetate treatment also affected peripheral reproductive tissues. Left testis weight in CA-treated males was markedly lower than in P-treated males. Similar effects of CA treatment on testis weight have been observed in other vertebrates (see Neumann, von Berswordt-Walhabe, Elger, Steinbeck, Hahn, and Dramer, 1970; Silver, 1977). Cyproterone acetate has both antiandrogenic as well as antigonadotropic activity. It is thus reasonable to assume that the inhibitory effect of CA treatment on the testis was mediated, at least in part, by its negative feedback action on the hypothalamic-pituitary axis and consequent suppression of gonadotropin secretion. Nevertheless, sufficient gonadotropin must have been released to maintain androgen synthesis by the testis because plasma T levels in CA-treated males at the end of the experiment were not significantly lower than in P-treated males. Significantly, the mean plasma T levels found in the P-treated males and in the CA-treated males are similar to those that have been reported for intact male A. carolinensis (see Crews, Traina, Wetzel, and Muller, 1978). Thus, the plasma levels of T observed in P-treated and CA-treated male A. sagrei appear to be physiologically reasonable. Cyproterone acetate treatment did have an antiandrogenic effect on the renal sex segment, a highly androgen sensitive organ (Prasad and Sanyal, 1969), as the size of this structure in CA-treated males was significantly less than in P-treated males. Cyproterone acetate has been found to inhibit a variety of accessory reproductive structures ranging from the seminal vesicles of rodents (Neumann et al., 1970)to the cloaca1 gland of male Japanese quail (Adkins and Mason, 1974).

ANTIANDROGENS

AND LIZARD

BEHAVIOR

13

Unlike CA-treated males, F-treated males showed no significant reduction in either aggressive behavior or sexual behavior. This finding is consistent with what has been found in studies of mammals. It differs, however, from results obtained from a study of male dominance in red-winged blackbirds (Seamy and Wield, 1980),where F had a significant inhibitory effect on male aggressive behavior and dominance. It is difficult to understand why F treatment does not inhibit male reproductive behavior in light of the fact that F can act centrally to block the negative feedback effects of androgens (Bloch and Davidson, 1971; Sodersten et al., 1975; Davidson et al., 1977) and peripherally to inhibit androgen-dependent reproductive tissues (Davidson et al., 1977). Davidson et al. (1977) have proposed that in the rat brain there may be two classes of androgen receptors, with one class mediating the negative feedback effects of androgens and the other mediating male reproductive behavior. If these two classes of receptors were to diier iu their ability to bind antiandrogen, then it would be possible to explain why an antiandrogen such as F could block the negative feedback action of androgens while not affecting male reproductive behavior. Others have argued that antiandrogens such as F do not inhibit male reproductive behavior because these compounds increase gonadotropin release and consequently elevate plasma androgen levels (Steinbeck and Neumann, 1975;Neumann et al., 1977).This increase in plasma androgen levels is believed to diminish or prevent any central inhibitory effects of these antiandrogens. Although the results of the present study cannot be used to address the issue of possible differences in brain androgen receptors, they do indicate that three weeks of F treatment signilicantly elevated plasma T levels, presumably by interfering with the negative feedback action of androgens on gonadotropin release. Assuming that F is capable of acting centrally to inhibit androgendependent behavior, an increase in the T to F ratio might be the reason why F failed to inhibit male reproductive behavior in A. sugrei. It is, perhaps, significant that in red-winged blackbirds, a species where F treatment inhibited aspects of male aggressive behavior, plasma T levels in Ftreated males did not differ from controls, and the levels of immunoreactive LH were even lower than in controls (Searcy and Wingfield, 1980). The absence of elevated T levels might explain why F treatment had a detectable inhibitory effect on male aggressive behavior of this species. Although F treatment did not affect male reproductive behavior in A. sugrei, it did have effects on peripheral reproductive structures. The finding that testis weight was significantly lower in F-treated males than in P-treated males was unexpected in light of the marked increase in plasma T levels. However, F may have exerted a direct inhibitory action on the testis before there was a significant rise in gonadotropin levels. Significantly more F-treated males had smaller renal sex segments than did P-treated males, indicating that F had an antiandrogenic effect on

14

RICHARD R. TOKARZ

this tissue. However, the inhibitory effect of F treatment on the renal sex segment was not as pronounced as the inhibitory effect observed with CA treatment. The higher plasma T levels in F-treated males may have diminished the ability of F to inhibit this highly androgen-sensitive tissue. In conclusion, CA, an antiandrogen with both antiandrogenic and antigonadotropic activity effectively inhibited important components of male reproductive behavior in an Anolis lizard. In addition, CA decreased testis weight and size of the highly androgen-dependent renal sex segment. However, CA treatment did not reduce plasma T levels. In contrast, F, which does not possess antigonadotropic activity, did not inhibit reproductive behavior, although it did reduce testis weight and size of the renal sex segment and elevated plasma T levels. Because CA significantly increased the number of males supplanted from their perch, this antiandrogen may be useful in studies addressing the role androgens play in the control of male social dominance in free-living lizards. ACKNOWLEDGMENTS I thank Maureen Donnelly, Craig Guyer, and Marc Hayes for their critique of this manuscript. I also thank John Beck, Jr. for developing the computer program used in the statistical analyses. Flutamide was obtained as a gift from the Schering Corporation (Bloomfield, N.J.). Cyproterone acetate was obtained as a gift from Schering Aktiengesellschaft (Berlin, FRG). This is contribution No. 217 from the Program in Ecology, Behavior, and Tropical Biology of the Department of Biology, University of Miami.

REFERENCES Adkins, E. K., and Mason, P. (1974). Effects of cyproterone acetate in the male Japanese quail. Horm. Behav. 5, l-6. Beach, F. A., and Westbrook, W. H. (1968). Morphological and behavioural effects of an “antiandrogen” in male rats. J. Endocrinol. 42, 379-382. Bloch, G. J., and Davidson, J. M. (1971).Behavioral and somatic responsesto the antiandrogen cyproterone. Horm. Behav. 2, 11-25. Crews, D. (1975). Inter- and intraindividual variation in display patterns in the lizard, Anolis carolinensis. Herpetologica 31, 37-47. Crews, D. (1979a). Neuroendocrinology of lizard reproduction. Biol. Reprod. 20, 51-73. Crews, D. (197913).Endocrine control of reptilian reproductive behavior. In C. Beyer (Ed.), Endocrine Control of Sexual Behavior, pp. 167-222. Raven Press, New York. Crews, D., Traina, V., Wetzel, F. T., and Muller, C. (1978). Hormonal control of mate reproductive behavior in the lizard, Anolis carolinensis: Role of testosterone, dihydrotestosterone, and estradiol. Endocrinology 103, 1814-1821. Davidson, J. M., Smith, E. R., and Damassa, D. A. (1977). Comparative analysis of the roles of androgen in feedback mechanisms and sexual behavior. In L. Martini and M. Motta (Eds.), Androgens and Antiandrogens, pp. 137-150. Raven Press, New York. Downie, N. M., and Heath, R. W. (1%5). Basic Statistical Methods. Harper & Row, New York. Edwards, D. A. (1970). Effects of cyproterone acetate on aggressive behaviour and the seminal vesicles of male mice. J. Endocrinol. 46, 477-481.

ANTIANDRGGENS

AND LIZARD BEHAVIOR

15

Greenberg, N., and Crews, D. (1983). Physiological ethology of aggression in amphibians and reptiles. In B. B. Svare (Ed.), Hormones and Aggressive Behavior, pp. 469-506. Plenum, New York. Licht, P., and Gorman, G. C. (1970). Reproduction and fat cycles in Caribbean Anolis lizards. Univ. Calif. Pub. 2001. 95, l-52. Mason, P., and Adkins, E. K. (1976). Hormones and social behavior in the lizard, Anolis carolinensis. Harm. Behav. 7, 75-86. Neumann, F., Graf, K.,-J., Hasan, S. H., Schenck, B., and Steinbeck, H. (1977). Central actions of antiandrogens. In L. Martini and M. Motta, (Eds.), Androgens and Antiandrogens, pp. 163-177. Raven Press, New York. Neumann, F., von Berswordt-Wallrabe, R., Elger, W., Steinbeck, H., Hahn, J. D., and Dramer, M. (1970). Aspects of androgen-dependent events as studied by antiandrogens. Rec. Progr. Harm. Res. 26, 337-410. Ortiz, P. R., and Jenssen, T. A. (1982). Interspecific aggresion between lizard competitors, Anolis cooki and Anolis cristatellus. Z. Tierpsychol. 60, 227-238. Prasad, M. R. N., and Sanyal, M. K. (1969). Effect of sex hormones on the sexual segment of the kidney and other accessory reproductive organs of the Indian house lizard Hemidactylus jlaviviridis Ruppell. Gen. Camp. Endocrinol. 12, 110-l 18. Ruby, D. E. (1984). Male breeding success and differential access to females in Anolis carolinensis. Herpetologica 49, 272-280. Ruibal, R. (1%7). Evolution and behavior in West Indian anoles. In W. W. Milstead (Ed.), Lizard Ecology: A symposium, pp. 116-140.. Univ. Missouri Press, Columbia. Sayler, A. (1970). The effect of anti-androgens on aggressive behavior in the gerbil. Physiol. Behav. 5, 667-671. Scott, M. (1984). Agonistic and courtship displays of male Anolis sagrei. Breviora Museum Camp. Zool. 479, l-22. Scott, N. J., Jr., Wilson, D. E., Jones, C., and Andrews, R. M. (1976). The choice of perch dimensions by lizards of the genus Anolis (Reptilia, Lacertilia, Iguanidae). J. Herpetol. 10, 75-84. Searcy, W. A., and Wingfield, J. C. (1980). The effects of androgen and antiandrogen on dominance and aggressiveness in male red-winged blackbirds. Harm. Behav. 4, 126135. Silver, R. (1977). Effects of the antiandrogen cyproterone acetate on reproduction in male and female ring doves. Harm. Behav. 9, 371-379. Sadersten, P., Gray, G., Damassa, D. A., Smith, E. R., and Davidson, J. M. (1975). Effects of a non-steroidal antiandrogen on sexual behavior and pituitary-gonadal function in the male rat. Endocrinology 97, 1468-1475. Sokal, R. R., and Rohlf, F. J. (1981). Biometry. Freeman, New York. Stamps, J. A. (1983). Sexual selection, sexual dimorphism, and territoriality. m R. B. Huey, E. R. Pianka, and T. W. Schoener (Eds.), Lizard Ecology: Studies of a Model Organism, pp. 169-204. Harvard Univ. Press, Cambridge. Steinbeck, H., Elger, W., and Neumann, F. (1967). Sexual activity of male rats under the intluence of oestradiol and antiandrogens and recurrence of libido after cessation of treatment. Acta Endocrinol. Suppl. 119, 63 (Abstract). Steinbeck, H., and Neumann, F. (1975). Androgen antagonists: Chemistry and influence on neural-gonadal function. In R. Vokaer and G. de Bock (Eds.), Reproductive Endocrinology: Hormone Relations and the Physiopathology of Pregnancy, Hormone Investigations, Hypothalamic-Hypophyseal Relations, pp. 135-144. Pergamon, New York. Terhune, F. W. (Ed.) (1983). Florida Statistical Abstract, 17th ed. The University of Florida Bureau of Economic and Business Research. The Univ. Presses of Florida, Gainesville. Tokarz, R. R. (1985). Body size as a factor determining dominance in staged agonistic encounters between male brown anoles (Anolis sagrei). Anim. Behav. 33, 746-753.

16

RICHARD R. TOKARZ

Tokarz, R. R. (1986). Hormonal regulation of male reproductive behavior in the lizard Anolis sagrei: A test of the aromatization hypothesis. Harm. Behav., in press. Whalen, R. E., and Edwards, D. A. (1969). Effects of the anti-androgen cyproterone acetate on mating behavior and seminal vesicle tissue in male rats. Endocrinology t34, 155156. Whalen, R. E., and Luttge, W. G. (1969). Contraceptive properties of the anti-androgen cyproterone acetate. Nature (London) 223, 633-634. Zucker, I. (1966). Effects of an antiandrogen on the mating behavior of male guinea pigs and rats. J. Endocrinol. 35, 209-210.