- Email: [email protected]
Prostaglandin-F2α stimulates reproductive behavior of female paradise fish (Macropodus opercularis)
HORMONES
AND
BEHAVIOR
19, 21-35 (1985)
Prostaglandin-F,, Stimulates Reproductive Behavior of Female Paradise Fish (Macropodus opercuhris) TRUDY ANN VILLARS, Department
of Psychology,
NANCY HALE, Wheaton
College,
AND DEBORAH CHAPNICK Norton,
Massachusetts
02766
The function of prostaglandin-F,, (PGF,,) in reproductive behavior of adult female paradise fish (Macropodus opercularis) was evaluated. In experiment 1, females were allowed to spawn normally with a male and were left with that male until the following day when testing occurred. Subjects were then randomly assigned to one of three groups: PGF*, (N = 8), vehicle control (N = 3), or a handling control (N = 3); for testing, females were removed, injected, then returned to the same tank. Seven of the eight subjects injected with 500 ng PGFz, initiated complete reproductive behavior following injection. Vehicle and handling control subjects did not show any female reproductive behaviors. Prostaglandininduced increases in sexual behaviors were seen from 15 to 75 min after the injection and were maximal approximately 45 min postinjection. In experiment 2, mature but unspawned (and presumably sexually naive) females were placed for 1 hr with males which were actively nestbuilding and parenting, following which subjects were either injected with PGFZ, (N = 6) or given a vehicle injection (N = 6). Experimental females did not behave differently from control females. Neither groups was observed to approach the male or initiate spawning behaviors following treatment; all subjects in both groups spawned on the following day. In experiment 3, sexually mature females were given extended familiarity with the male and the test tank. Pairs which had not spawned after 6 days (unspawned) were randomly assigned to experimental (N = 6) or control (N = 5) groups. Pairs which had spawned 1 day previously (N = 6) formed a “prespawned” comparison group. Females from the experimental and prespawned groups were injected with PGF,, (500 ng), while the control females received a vehicle injection. Half of the “prespawned” females performed spawning acts while none of the unspawned or control females did so. Detailed behavioral analyses showed little or no effect of PG treatment, despite greater familiarity with the male. In experiment 4, females were placed with males and observed until the early signs of spawning, at which time they were injected with indomethacin (a PG synthesis inhibitor) or vehicle. The behavior of females treated with 35pg (N = 3) or 7Opg (N = 6) of indomethacin was unaffected by indomethacin and largely indistinguishable from controls (N = 6). In summary, exogenous prostaglandins reinstate sexual behavior in female paradise fish. However, responsiveness to prostaglandins is influenced by prior sexual experience. Moreover, inhibition of PG synthesis with indomethacin following onset of sexual behavior appears insufficient to block behavior or egg release. o 1985 Academic PKZSS, IIK.
21 0018/506X/85 $1.50 Copynghf 0 1985 by Academic Press. Inc. All rights of reproductmn in any form reserved.
22
VILLARS,
HALE,
AND
CHAPNICK
Prostaglandins (PGs) are involved at several points in vertebrate reproductive physiology (Turner and Bagnara, 1976), and appear also to modulate reproductive behavior. At the level of the ovary PGs have been implicated in several stages of ovarian function including ovum maturation, ovulation and breakdown of the corpus luteum. In the CNS exogenous PGs stimulate the adenylate cyclase system and can induce release of LRF as well as pituitary gonadotropins (Behrman, 1979). One of the earliest demonstrations for a role of PGs in reproductive behavior was seen with goldfish (Stacey, 1976; Stacey, 1981, for review). Early work had indicated that female goldfish sexual behavior is tied to the presence of eggs in the reproductive tract, since spawning behavior of female goldfish can be terminated by removal (“stripping”) of ovulated eggs, and reinstated by injecting ovulated eggs through the ovipore into the ovarian lumen (Stacey and Liley, 1974). Indomethacin, a PG inhibitor, blocks the onset of spawning behaviors following egg injection (Stacey, 1976), while injection of PGs induces spawning behaviors in both male and female goldfish and can overcome the indomethacin-induced blockade of reproductive behavior in females (Stacey, 1976; 1981). The behavioral effect of PGs is enhanced if the same dose of PG is introduced intraventricularly rather than via systemic injection (Stacey and Peter, 1979), suggesting a central site of action. Prostaglandins have also been shown to influence female proceptivity and receptivity in several tetrapod vertebrates. In Rana pipiens prostaglandins induce female receptivity to male clasps and inhibition of prostaglandin synthesis depresses AVT-induced receptivity (Diakow and Nemiroff, 1981), suggesting that AVT induces receptivity by stimulating PG synthesis. In contrast, receptivity of the female is blocked by prostaglandin treatment in the female lizard Anolis carolinesis (Tokarz and Crews, 1981) and in the guinea pig (Man-one, Rodriguez-Sierra, and Feder, 1979). Prostaglandins stimulate the occurrence of lordosis and solicitation in estrogen-primed, ovariectomized female rats and hamsters (Dudley and Moss, 1976; Hall, Luttge, and Berry, 1975; Rodriguez-Sierra and Komisaruk, 1978; Buntin and Lisk, 1979). Moreover, prostaglandins can induce lordosis responding within 20 min in ovariectomized female rats in the absence of estrogen priming. (RodriguezSierra and Komisaruk, 1977). Interestingly, the time course of the PG effect on lordosis reported by Rodriguez-Sierra and Komisaruk is similar to the time course seen in goldfish, suggesting that the PG effects on reproductive behavior may represent a general phenomenon in vertebrates. The anabantoid teleosts provide an appropriate model with which to test the generality of the effects of PGs on female teleost reproductive behavior, since the pattern of their reproductive behavior allows detailed analysis of hormone-behavior interactions. The elaborate reproductive sequence shown by M. opercularis (cf., Hall, 1966; Villars and DeNeff,
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
23
1979) includes a courtship of 1 to 4 days, a spawning sequence extending over 3 to 5 hr, involving tight synchronization between male and female behavior, and a parental phase of approximately 3 days in which the male defends and cares for the young. During the Courtship phase the male builds a nest of bubbles clustered at the water surface. He may frequently approach the female to “lead” her to the nest or to attack if she is unresponsive. This time period apparently is a function of developing female reproductive readiness. A study by Cheal and Davis (1974) suggested that female reproductive readiness of Trichogaster trichopterus, a species closely related to M. opercularis, is crucial to the initiation of the spawning sequence. If the female from a pair of T. trichopterus is replaced daily, spawning does not occur, while spawning occurs within the normal time if the male is replaced daily. The female M. operculuris also appears to play a major role in the initiation and termination of spawning (personal observation). The female initiates spawning by approaching a nestbuilding male, and she terminates the behavioral sequence when she has apparently depleted the store of ovulated eggs available for egg release. The Spawning phase begins when the female approaches the nestbuilding male. It continues for 3 to 5 hr and consists of repeated bouts in which the male curves his body around the female and the pair inverts under the nest (Villars and DeNeff, 1979). These curves and inversions are analogous to the mounts and intromissions of the rodent-mating sequence. Following each inversion a stream of approximately 10 to 50 eggs are released and fertilized in the stream of milt released concurrently by the male; the first inversions do not result in egg release, and earlier bouts of egg release produce fewer eggs. The male places the eggs in the bubble nest and cares for them by continually rearranging them and adding further bubbles to the nest. The male’s Parental behavior continues for several days after spawning is complete until the young are free swimming. The 300 to 500 young hatch in approximately 3 to 5 days. In a laboratory setting pairs may spawn repeatedly at intervals of 1 to 3 weeks (Hall, 1966; personal observation). The present studies were designed to determine whether prostaglandins play a role in the regulation of female reproductive behavior of M. opercularis comparable to the effect seen in goldfish. Experiment 1 tested whether exogenous PGF2, could induce sexual activity in females which had recently terminated spawning. Experiments 2 and 3 examined the effects of PGs on unspawned females which were placed with actively nestbuilding males. In Experiment 2, females were adapted to the test situation for 1 hr before testing. In experiment 3 females were adapted for 6 days prior to testing. Experiment 4 examined the ability of indomethacin, a PG synthetase inhibitor, to interfere with ongoing sexual behavior.
24
VILLARS,
HALE,
AND
CHAPNICK
METHODS Subjects
Subjects were sexually mature Mucropodus opercularis obtained from a local supplier. Experimental females weighed from 1.4 to 2.9 g (x: 1.9 + 0.4) and were 4.5 to 5.7 in. (x 5.0 + 0.3) in experiments 1 and 2. In experiment 4, mean body weight was 3.1 + 0.67. Subjects in experiment 3 were not weighed, but were from the same population as those in experiment 4. Fish of both sexes were housed together, in groups of 3050, in 80- or 120-liter aquaria, and fed once or twice daily with freezedried worms or Tetramin flakes. Under these holding conditions, males do not establish territories or nests, and mating does not occur. A photoperiod of 14L-10D was used in all experiments and was supplemented by natural light in experiments 1 and 2. Temperatures in the holding and test aquaria ranged from 21 to 24°C and pH ranged from 6.7 to 7.2. Spawning
Procedure
Pairs of fish were placed in 40-liter aquaria half-filled with distilled water into which 4 mg of sodium bicarbonate had been added (experiments 1 and 2) or with filtered well water (experiments 3 and 4) and were allowed to spawn normally according to the procedure of Villars and Davis (1977). Females were each placed in an aquarium first and the males were introduced 24 hr later to minimize male aggression against the female. Pairs generally spawned 2 to 10 days after the introduction of the male, somewhat more slowly than in earlier work (Villars and Davis, 1977). Occurrence of spawning is indicated by direct observation of spawning acts or the presence of eggs in the nest. On the day following spawning, “prespawned” females in experiment 1 were randomly assigned to one of three groups: PG treated (N = 8) vehicle control (N = 3) or handling control (N = 3). Prostaglandin
Treatment
(Experiments
1, 2, and 3)
Each female was removed from its tank, injected, and immediately returned. For the intraperitoneal injection, subjects were restrained in a fish net, a 26-gauge needle was inserted under the scales and through the body cavity wall 2- to 3-mm dorsal of the ovipore. A volume of 10 ~1 was delivered using a Hamilton microliter syringe. Hormone-treated subjects received a total of 500 ng of PGFzu (Tris salt, Sigma) dissolved in 0.1 M sodium carbonate. Vehicle controls received an equal volume of the sodium carbonate solution. Handling controls were removed from the tank and restrained. All subjects were immediately returned to the same tank from which they had been removed and observed at 15, 45, 75, and 105 min postinjection.
PROSTAGLANDINS
Procedure
for Unspawned
AND
FISH
REPRODUCTIVE
BEHAVIOR
25
Females
One-hour adaptation (experiment 2). Following testing and within 48 hr postspawning, the spawned females from experiment 1 were removed from the test aquaria. For experiment 2, unspawned females from the holding tank were placed with the nestbuilding, parenting males. These unspawned females were probably virgin since holding conditions are not conducive to spawning. The females were left undisturbed in the new environment for a period of 1 hr, then injected with PGF2, (N = 6), or with vehicle (N = 6), and observed. During the hour adaptation period, all females had begun foraging around the test environment. Six-day adaptation (experiment 3). For experiment 3, females were placed with unspawned males and left undisturbed for 6 days. Females which had not spawned after 6 days were injected with PGFz, (N = 6) or with vehicle (N = 5). Six spawned females were injected with PGF*, 1 day postspawning as an additional comparison group, and as a replication of experiment 1. Indomethacin
Treatment
(Experiment
4)
Following pairing, subjects were observed at approximately hourly intervals during the light period for early indications of the onset of spawning. Pairs which are observed to curve and invert can reliably be expected to complete the 2- to 5-hr spawning sequence (personal observation). Following the first inversion females were first observed for 10 min, then were removed from the tank and injected with a 35 (N = 3) or 70 (N = 6) pg dose of indomethacin in a IO-p1 volume. Control subjects (N = 6) were injected with the sodium carbonate vehicle. Subjects were continuously observed for the occurrence of sexual acts and egg release. Ten-minute behavioral observations were performed at 1 and 2 hr postinjection. Behavioral
Observation
Procedure
Behavioral observation periods were 10 min in duration. For all experiments, pairs were continuously observed for the occurrence of curves and inversions. In experiments 1, 2, and 3, subjects were observed in more detail at 15, 45, 75, and 105 min postinjection. In experiment 4, formal observations occurred at 1 and 2 hr postinjection. In all PG treatment experiments, subjects were observed for 10 min prior to injection (baseline). In experiments 1 and 2, the total number of curves, inversions, and long approaches was recorded for each subject during the formal observation periods. A curve was recorded when the female positioned herself and the male wrapped his body around her. An inversion was recorded when the male rolled his body under the female and the pair
26
VILLARS,
HALE,
AND
CHAPNICK
inverted. Curves and inversions are the primary behavior patterns Seen during the spawning sequence and they occur repeatedly through the 3- to 5-hr reproductive sequence. During normal spawning, egg release begins approximately 1 hr after the first inversion (personal observations) and occurs concurrently with inversions. A long approach was recorded when the female approached the nestbuilding male and remained under the nest for longer than 5 sec. Observers were uninformed (“blind”) regarding the treatment of subjects. In experiments 3 and 4, the total time the female spent under the nest was recorded along with several additional behavioral measures. The female was considered to be under the nest if she was within three body lengths of the nestbuilding male and under the nest he had built. This measure is seldom ambiguous, since females which are not under the nest with the nestbuilding male usually spend their time in the comer most distant from the male. In experiment 4, latency to egg release was measured from the first inversion, which had also signaled the start of the observation. That latency had been observed in earlier work (Villars and DeNeff, 1979) to be quite consistent among spawning pairs. In addition to time under the nest, subjects in experiments 3 and 4 were also scored for female orient/partial approach, female lateral display/ frontal display, female tailwaglbite-butt, male nestbuilding, and maleinitiated interactions. A female orient/partial approach (O/A) was scored when the female’s nose was pointed toward the male and she moved, with fins folded, directly toward him for at least two body lengths. This behavior was scored only when the female was not under the nest. Female lateral display/frontal display (LD/FD) was scored if the female, during an interaction with the male, flared her gill covers (frontal display) or completely extended her dorsal and caudal fins (lateral display). Tailwagging (TW) was recorded when a female assumed a vertical posture, spread her fins, and waved her caudal fin. Tailwagging was seen and recorded in response to male approach and following female approaches to the male. However, it was not recorded following a male attack on the female, since the tailwag might be construed as part of an escape or avoidance response. A bite-butt (BB) was recorded when the female approached a male and rammed him firmly on the flanks. Though this action has the appearance similar to an attack by the female, it differs in that the butts are directed more caudally than are attacks and males respond sexually rather than agonistically with lateral display, a partial or complete curve of the male around the female or by increased vigor of male nestbuilding. Female O/A, LD/FD, and TW/BB were all considered indictions of female sexual “motivation” or “proceptivity.” Male nestbuilding was scored using an informal time sampling procedure; at regular intervals during an observation period each tank was scored if the male was nestbuilding. Males were considered to be “nestbuilding
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
27
males” if they were observed nestbuilding at least once during the observation period. A male-initiated interaction was scored when the male left the nest area and approached the female. An approach was usually followed by a male LD or attack or both. The entire intraction was scored as one event. Male actions toward the female which occurred in response to a female orient/partial approach were scored as male-initiated behaviors. Statistics In experiments 1, 2, and 4, statistical significance was determined with the test for differences between proportions (Daniel, 1978). In experiment 3, the magnitude of the difference between the three groups in percent responding was examined using the chi-square test for independence (Sokal and Rohlf, 1969). Since the expected values in some cells were sometimes less than 5, the probability estimates are provided for their descriptive value, and 95% confidante intervals (Daniel, 1978) were calculated to allow comparisons between groups. RESULTS
Experiment
1: Effects of PGF,,
on Prespawned Females
Recently spawned, prostaglandin-treated females showed a dramatic increase in sexual activity. Seven of the eight prespawned pairs which were PG treated were observed to curve and/or invert (spawning acts) during the observation periods following injection. The eighth prespawned, PGF,,-treated female was observed curving in the interval between observations. None of prespawned controls were observed to curve or invert either during the observation period or in the interval between observations. Curves and inversions appeared normal except that egg release did not occur; inversions without egg release are also seen in the early stages of normal spawning. Prostaglandin treatment reinstated sexual activities and stimulated female approach (Fig. 1) within 15 min of the injection for most females. Two hours after the treatment none of the treated females were engaging in sexual activity or approaching the parenting male. At 45 min following the injection the maximum number of subjects were responding. However, for those subjects showing a short latency response to PGFI, the peak rates of that response occurred shortly after the onset of the behavioral effect. Experiment
2: Effects of PGF,,
on Unspawned Females
Unspawned females placed with parenting males for 1 hr did not respond to PGF2, treatment. During the baseline period immediately prior to PG treatment none of the females showed long approaches to the nestbuilding male or signs of sexual activities. Prostaglandin treatment did not change
28
VILLARS,
t 1* hi.
45
HALE,
AND CHAPNlCK
75
105
135
165
Minutes
Post
Injection
195
225
1Effects of PG treatment on mating behaviors. The median number of long approaches, curves, and inversions is plotted as a function of time since injection in PGtreated (0-O) and control (O---O) subjects.
any of the recorded measures and there were no other behavioral changes apparent to the observers. All pairs in both control and PG-treated groups spawned the followed day. Experiment 3: PGF,, Treatment of Unspawned Females Familiarized with a Male In Experiment 2, unspawned females placed with parenting males for 1 hr showed no response to PGF2, treatment. Experiment 3 tested whether more extended familiarity with the male would expose a PG effect in females which were not prespawned. Sexually mature, unspawned females were placed with males; pairs which had not spawned after 6 days formed
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
29
PG-treated (N = 6) or vehicle control N = 5) groups. Pairs which had spawned 1 day previously (N = 6) formed a prespawned, PG-treated comparison group. As in experiment 1, PG treatment reinstated reproductive activity only in pairs which had completed spawning (PG-X) the previous day (Table 1, column 3). Half the prespawned pairs curved or inverted 45 or 75 min postinjection. The proportion of prespawned (PG-X) females spending time under the nest and the mean time (in seconds) which prespawned (PG-X) females spent under the nest increased following PGF2, injection (Table 1, columns 4 and 5). In contrast, only one PG-treated unspawned (PG-N) female spent time under the nest and that time was less than the time spent by the experienced, PG-treated females. None of the vehicletreated (VEH) females spent any time under the nest or engaged in any curves or inversions. The percent proceptivity and the mean proceptivity scores of both unspawned (PG-N, VEH) groups were below those of the prespawned (PG-X) group in both the baseline and treatment periods (Table 1, columns 6 and 7). Proceptive acts included orient/approach, lateral/frontal display, tailwag, and bite-butt. During the baseline period the groups did not differ in the proportion of males nest building (Table 1, column 8). That proportion fell slightly in the vehicle-treated group but not in the PG-treated groups. Experiment 4: Effects of PG Synthesis Inhibition Occurring Spawning
on Normally
Neither of the two doses of indomethacin blocked spawning behaviors when injected 10 min after the first inversion; all pairs completed spawning regardless of treatment. There was no change in the percentage of pairs showing spawning acts (curves or inversions) following vehicle (83%) or indomethacin (80%) treatments and there were no significant differences between groups in the percentage of pairs observed spawning. There was a slight, nonsignificant trend among the high dose group toward increased occurrence of spawning acts; whereas four of the six highdose subjects were observed curving/inverting during the baseline, 100% did so after the indomethacin treatment. There was also no effect of either dose of indomethacin on mean latency to egg release. The mean latency to egg release, measured from the first inversion, was 157 + 59 min for fish receiving the high dose and 157 + 67 min for fish receiving the vehicle. Animals receiving the low dose had a mean latency to egg release of 154 rf: 22 min. Thus indomethacin did not affect latency to first egg release. Moreover, there was no effect of indomethacin on any of the other recorded behaviors. DISCUSSION
Prostaglandin treatment produces a clear and dramatic reinstatement of sexual behavior in previously spawned females. These results parallel
6 6 5
6 6 5
75 min PG-X PG-N VEH
Overalld PG-X PG-N VEH
y b ’ d e
33 ( 6-73) 0 ( O-40) 0 ( O-50) NS
6 6 5
45 min PG-X PG-N VEH
(15-85) ( O-40) ( O-50) P < 0.05
33 ( 6-73) 0 ( O-40) 0 ( O-50) NS
Percent P under nest”
83 17 0 9.70,
67 0 0 9.59,
67 17 0 6.56,
(50-99) ( l-60) ( O-50) P < 0.01
(27-94) ( O-40) ( O-50) P < 0.01
(27-94) ( l-60) ( O-50) P < 0.05 305 -
476 17
280 111
271 -
Time under nestb
TABLE 1 Adapted, Unspawned Pairs (PC-N) PC&Treated, l-Day Prespawned
67 (27-94) 17 ( l-60) 0 ( O-50) 6.56, P < 0.05
on 6-Day
Column indicates the percent responding, with 95% confidence interval in parenthesis. The mean time in seconds which the female spent under the nest during the IO-min observation. Average frequency of acts per 10 minute observation for only those counted as responders. The “overall” category is the combined responding for the 15-, 4S-, and 75-min observations. Chi square was calculated as an index of the reliability of the difference between the three groups.
50 0 0 6.68,
(15-85) ( O-40) ( O-50) P < 0.05
0 ( O-40) 0 ( O-40) 0 ( O-50) NS
6 6 5
15 min PG-x PG-N VEH
50 0 0 6.68,
0 ( O-40) 0 ( O-40) 0 ( O-50) NS
6 6 5
PG-X PC-N VEH
Percent curve invert”
of PC&Treatment
N -
Effects
Baseline
The
94) 60) 66) 0.05’
(60-100) (27- 94) ( 8- 66) P < 0.10
(27( l( lP <
100 (60-100) 67 (27- 94) 67 (20- 92) NS
100 (60-100) 17 ( l- 60) 40 ( 8- 66) 8.84, P < 0.025
100 (60-100) 33 ( 6- 73) 20 ( 1- 66) 8.44, P < 0.05
100 67 40 4.80,
67 17 20 6.56,
Percent proceptive”
5.0 2.0 2.5
6.1 2.0 1.0
8.1 2.2 1.5
5.2 1.0 1.0
Mean proceptive’
Compared to Vehicle-Treated, Pairs (PC-X)
(VEH)
100 (60-100) 100 (60-100) 60 (20- 92) 5.44, P < 0.10
100 (60-100) 83 (40- 99) 40 ( 8- 66) 5.70, P < 0.10
100 (60-100) 100 (60-100) 20 (20- 92) 12.55, P < 0.05
100 (60-100) 100 (60-100) 60 (20- 92) 5.44, P < 0.10
100 (60-100) 83 (40- 99) 80 (34- 99) NS
Percent 6 NB”
Unspawned
2.5 2.5 A
2 5
1
3.5 3.5 3
2.5 2.5 A
Mdn d Init’
and to
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
31
those reported for goldfish (Stacey, 1976, 1981), where PGs reinstated sexual behavior in females which had not ovulated or which had been stripped of their eggs. These results also are consistent with reports in other vertebrates indicating that prostaglandins enhance receptivity and proceptivity in Rana pipiens (Daikow and Nemiroff, 1981), as well as in rats and hamsters (Hall et al., 1975; Buntin and Lisk, 1979; RodriguezSierra et al., 1977, 1978). The time course of the PG effect appears remarkably similar across the vertebrates tested. We have observed an effect of PGF2, approximately 15 min postinjection and a peak at approximately 45 min. A similar time course was reported by Stacey (1978) for the goldfish and by RodriguezSierra and Komisaruk (1977) for the rat. The effects of PGF*, on female M. operculuris might best be described as reinstatement of proceptivity rather than receptivity. Female M. opercufuris are largely responsible for the initiation of reproductive behaviors. While the nestbuilding male may engage in “courtship” behaviors such as “leading to the nest,” the onset of sexual behavior is signaled by the female approach to the nestbuilding male (unpublished observations). Moreover, the approach of the female to the nestbuilding male is apparently not contingent on prior male courtship or solicitation; a steadily nestbuilding male is a suflicient stimulus. The onset of persistent female approaches toward the male is the first clear indication that spawning acts will be initiated. PG-treated females court/solicit the nestbuilding male, tailwagging and butting his flanks to initiate the sex acts of curving and inverting. Though those sex acts are often described as male actions, the PG induction of full mating seen here implicates the female hf. operculuris as the major initiator of mating behavior including sex acts. For the period of its effectiveness, PG-stimulated behavior looked indistinguishable from the normal pattern of spawning behavior. It should be noted that egg release did not occur during the inversions; however, inversions without egg release are also normally seen during the early stages of the spawning cycle. The details of the female interactions with the male appeared essentially the same as that seen at the height of spawning activity; the female hovered under the nest, assisting with egg retrieval and nestbuilding, and periodically butting the male in the flanks, thereby eliciting a lateral display, curve, or possibly an inversion. In naturally occurring spawning the female leaves the nest area for 1 or 2 min following an inversion; the PG-treated females behaved likewise. Thus, the behavior of both PG-treated and naturally spawning females shows a characteristic approach-retreat rhythm. The behavior change of one PG-treated, prespawned female was particularly striking in this regard. During the pretreatment period she had remained at the periphery of the nest, foraging and occasionally retrieving an egg which had drifted from the nest, returning the eggs closer to the male. Such behavior is
32
VILLARS,
HALE,
AND
CHAPNICK
characteristic of many females in the hours foliowing spawning, though they have usually moved to the opposite side of the tank by the day following spawning. However, following PG treatment, this same female began to show the approaches and retreats characteristic of normal spawning. Her approaches were directed to the male and to the center of the nestbuilding activity. Her retreats were to the opposite end of the tank. This was a striking change in the quality of the behavior pattern. Pollack, Thompson, Stabler, and Keenen (1981) have demonstrated that female Trichogaster trichopterus, an anabantoid closely related to M. opercularis, will engage in spawning acts with more than one male during the same spawning cycle. They suggest that the retreat following a successful inversion may facilitate location of other appropriate mates. Thus, the degree to which PG-elicited behavior mimics the normal pattern of reproductive behavior is striking and suggests that endogenous PGs are a factor in the onset of normal female proceptivity. Quantitative data are presently being gathered to allow more detailed comparisons between naturally occurring and PG-induced behavior. Sensitivity to the effects of prostaglandin treatment on M. opercularis appears to depend on the prior spawning of the female. This limited sensitivity contrasts with the reports in goldfish (Stacey, 1981) and rats (Rodriguez-Sierra and Komisaruk, 1978). Prostaglandin treatment stimulates proceptivity in female goldfish which are partially ovariectomized and PG treatment even induces female behavior patterns in male goldfish (Stacey, 1981). Likewise the effects of PG on rat sexual behavior do not require the presence of the ovaries or of hormonal priming (RodriguezSierra and Komisaruk, 1977). In contrast, we have only been able to demonstrate a PG effect in fish which have recently spawned. In experiment 2, unspawned females placed with males for 1 hr prior to PG injection show no behavioral effects of PG treatment. The 1 hr adaptation period preceding PG injection appeared adequate, since the previously spawned males had returned to nestbuildjng and parenting while the females were foraging, showing no signs of distress. The PG dose should have been adequate, since the same dose had completely reinstated sexual behavior in the prespawned females of experiment 1. Moreover the females were physiologically prepared for spawning, since all subjects in both control and PG treatment groups were observed to spawn the following day. It was possible that the I-hr adaptation period used in experiment 2 was insufficient. However in experiment 3, greater familiarity with the male and with the spawning environment did not unmask a PG effect in unspawned females; prostaglandin treatment produced no clearly apparent effect in unspawned subjects when compared to vehicle-treated controls. It is possible that 6-day unspawned females, though familiar with the male, were for some other nonspecific reason unable to demonstrate a response to PG. However, the most obvious reasons are ruled out; both
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
33
males and females appeared healthy and were eating well. We took careful records of nestbuilding and the males were nestbuilding and seemed otherwise superficially like adequate social stimuli. It remains to be determined whether prespawning sensitizes the females to later PG treatment via a behavioral or an endocrine mechanism. In experiment 3 the proportion of males nestbuilding declined in the vehicle-treated group, while males in both PG-treated groups showed continued high levels of nestbuilding. It may be that males are sensitive to the endocrine state of the female. Indeed, male goldfish have been shown to preferentially court PG-treated females, leading Partridge, Liley, and Stacey (1976) to suggest that PG treatment may induce production of a pheromone which stimulates male behaviors. Such a pheromone in M. operculuris could serve to stimulate male nestbuilding. Further study with more sensitive measures of nestbuilding activity are necessary. It is clear that in M. opercularis indomethacin does not interfere with maintenance of ongoing sexual behavior. The low dose of indomethacin and the time course of testing which we used were comparable to those used by Stacey (1976) and by Diakow and Nemiroff (1981) to block sex behavior, in goldfish and Ram pipiens, respectively. Moreover, the higher dose which we used produced increases in frequency of spawning acts, which, though statistically insignificant, are clearly opposite to our predicted effect. At the higher dose of indomethacin there was some suggestion that egg release and time spent by the female under the nest may have been reduced slightly. The high dose may also have altered the ratio of curves to inversions. However, these effects were not reliable with the sample sizes we used and were minimal in comparison to the absence of a more general effect. Further study with indomethacin is in progress. On the one hand, prostaglandins may initiate sexual behavior without being necessary for maintenance of the cycle; indomethacin treatment earlier in the spawning cycle may reveal an effect of PG synthetase inhibition. Use of PG synthesis inhibitors such as ibuprofen or acetaminophen might also yield different results than found here. Alternatively, PGs may not be directly involved in the maintenance or initiation of sexual behavior. Exogenous PG reinstatement of sexual behavior may be operating indirectly. Inhibition of naturally occurring behavior following PG inhibition is essential to demonstrate a direct role of PGs in sexual behavior. To date, no such effect has been demonstrated in any of the species studies. In summary, PGFZO(reinstates sexual activity in female M. opercularis which have recently completed spawning. This drug-induced behavior appears remarkably similar to normally occurring behavior. However, unlike in the goldfish, sensitivity to PG in M. opercularis is restricted, such that unspawned females are largely unresponsive to PGF,, treatment.
34
VILLARS,
HALE, AND CHAPNICK
The nature and extent of the restrictions are not yet clear; endocrine and/or behavioral factors may account for the unresponsiveness of unspawned females. Finally M. operucularis do not show reductions in ongoing sexual behavior following treatment with a prostaglandin synthetase inhibitor. ACKNOWLEDGMENTS The authors thank Martha Burdick for her assistance with experiments 3 and 4. Thanks are also due to Nancy Shepardson for assistance with typing the manuscript.
REFERENCES Behrman, H. R. (1979). Prostaglandins in hypothalamo-pituitary Rev.
Physiol.
and ovarian function. Ann.
41, 685-700.
Buntin, J. D., and Lisk, R. D. (1979). Prostaglandin E,-induced lordosis in estrogen primed female hamsters: Relationship to progesterone action. Physiol. Behav. 23(j), 569-575. Cheal, M., and Davis, R. E. (1974). Sexual behavior: Social and ecological influences in the Anabantoid fish, Trichogaster trichopterus. Behav. Biol. 10, 435-445. Daniel, W. W. (1978). Applied Nonparametric Statistics. Houghton-Mifflin, Boston, Mass. Diakow, C., and Nemiroff, A. (1981). Vasotocin, prostaglandin, and female reproductive behavior in the frog, Rana pipiens. Horm. Behav. 15, 86-93. Dudley, C. A., and Moss, R. L. (1976). Facilitation of lordosis in the rat by prostaglandin E,. J. Endocrinol. 71, 457. Hall, D. D. (1966). An ethological study of three species of anabantoid fishes (Pisces, Belontiidae), Ph.D. dissertation, University of Oklahoma. Hall, N. R., Luttge, W. F., and Berry, R. B. (1975). Intracerebral prostaglandin E,: Effects upon sexual behavior, open field activity, and body temperature in ovariectomized female rats. Prostaglandins 10, 8784387. Marrone, B. L., Rodriguez-Sierra, J. F., and Feder, H. H. (1979). Differential effects of prostaglandins on lordosis behavior in female guinea pigs and rats. Biol. Reprod. 20 853-861. Partridge, B. L., Liley, N. R., and Stacey, N. E. (1981). The role of pheromones in the sexual behavior of the goldfish. Anim. Behav. 24, 291-299. Pollak, E. I., Thompson, T., Stabler, A. L., and Keener, D. (1981). Multiple matings in the Blue Gourami, Trichogaster Trichopterus. (Pisces, Belontiidae). Anim. Behav. 29 5563. Rodriguez-Sierra, J. F., and Komisaruk, B. R. (1977). Effects of prostaglandin Ez and indomethacin on sexual behavior in the female rat. Horm. Behav. 9, 281-289. Rodriguez-Sierra, J. F., and Komisaruk, B. R. (1978). Lordosis induction in the rat by prostaglandin E, systemically or intracranially in the absence of ovarian hormones. Prostaglandins 15, 513-525. Stacey, N. E. (1976). Effects of indomethacin and prostaglandins on the spawning behavior of female goldfish. Prostaglandins 12 113-126. Stacey, N. E. (1981). Hormonal regulation of female reproductive behavior in fish. Amer. Zool.
21, 305-316.
Stacey, N. E., and Liley, N. R. (1974). Regulation of spawning behaviour in the female goldfish. Nature (London) 247, 71-72. Stacey, N. E., and Peter, R. E. (1979). Central action of prostaglandins in spawning behavior of female goldfish. Physiol. Behav. 22, 1191-l 196. Sokal, R. R., and Rohlf, F. J. (1969). Biometry. Freeman, San Francisco, Calif. Tokarz, R. R., and Crews, D. (1981). Effects of prostaglandins on sexual receptivity in the female lizard, Anolis carolinensis. Endocrinology 109, 451-457.
PROSTAGLANDINS
AND FISH REPRODUCTIVE
BEHAVIOR
35
Turner, G. D., and Bagnara, J. T. (1976). General Endocrinology. Saunders, Philadelphia, Pa. Villars, T. A., and Davis, R. E. (1977). Castration and reproductive behavior in the Paradise Fish, Macropodus operculuris (L) (Osteichthyes: Belontiidae). Physiol. Behav. 19, 371-375. Wars, T. A., and DeNeff, S. J. (1979). Quantitative description of the patterns of reproductive behavior of Macropodus opercularis. Amer. Zool. 19, 935.
AND
BEHAVIOR
19, 21-35 (1985)
Prostaglandin-F,, Stimulates Reproductive Behavior of Female Paradise Fish (Macropodus opercuhris) TRUDY ANN VILLARS, Department
of Psychology,
NANCY HALE, Wheaton
College,
AND DEBORAH CHAPNICK Norton,
Massachusetts
02766
The function of prostaglandin-F,, (PGF,,) in reproductive behavior of adult female paradise fish (Macropodus opercularis) was evaluated. In experiment 1, females were allowed to spawn normally with a male and were left with that male until the following day when testing occurred. Subjects were then randomly assigned to one of three groups: PGF*, (N = 8), vehicle control (N = 3), or a handling control (N = 3); for testing, females were removed, injected, then returned to the same tank. Seven of the eight subjects injected with 500 ng PGFz, initiated complete reproductive behavior following injection. Vehicle and handling control subjects did not show any female reproductive behaviors. Prostaglandininduced increases in sexual behaviors were seen from 15 to 75 min after the injection and were maximal approximately 45 min postinjection. In experiment 2, mature but unspawned (and presumably sexually naive) females were placed for 1 hr with males which were actively nestbuilding and parenting, following which subjects were either injected with PGFZ, (N = 6) or given a vehicle injection (N = 6). Experimental females did not behave differently from control females. Neither groups was observed to approach the male or initiate spawning behaviors following treatment; all subjects in both groups spawned on the following day. In experiment 3, sexually mature females were given extended familiarity with the male and the test tank. Pairs which had not spawned after 6 days (unspawned) were randomly assigned to experimental (N = 6) or control (N = 5) groups. Pairs which had spawned 1 day previously (N = 6) formed a “prespawned” comparison group. Females from the experimental and prespawned groups were injected with PGF,, (500 ng), while the control females received a vehicle injection. Half of the “prespawned” females performed spawning acts while none of the unspawned or control females did so. Detailed behavioral analyses showed little or no effect of PG treatment, despite greater familiarity with the male. In experiment 4, females were placed with males and observed until the early signs of spawning, at which time they were injected with indomethacin (a PG synthesis inhibitor) or vehicle. The behavior of females treated with 35pg (N = 3) or 7Opg (N = 6) of indomethacin was unaffected by indomethacin and largely indistinguishable from controls (N = 6). In summary, exogenous prostaglandins reinstate sexual behavior in female paradise fish. However, responsiveness to prostaglandins is influenced by prior sexual experience. Moreover, inhibition of PG synthesis with indomethacin following onset of sexual behavior appears insufficient to block behavior or egg release. o 1985 Academic PKZSS, IIK.
21 0018/506X/85 $1.50 Copynghf 0 1985 by Academic Press. Inc. All rights of reproductmn in any form reserved.
22
VILLARS,
HALE,
AND
CHAPNICK
Prostaglandins (PGs) are involved at several points in vertebrate reproductive physiology (Turner and Bagnara, 1976), and appear also to modulate reproductive behavior. At the level of the ovary PGs have been implicated in several stages of ovarian function including ovum maturation, ovulation and breakdown of the corpus luteum. In the CNS exogenous PGs stimulate the adenylate cyclase system and can induce release of LRF as well as pituitary gonadotropins (Behrman, 1979). One of the earliest demonstrations for a role of PGs in reproductive behavior was seen with goldfish (Stacey, 1976; Stacey, 1981, for review). Early work had indicated that female goldfish sexual behavior is tied to the presence of eggs in the reproductive tract, since spawning behavior of female goldfish can be terminated by removal (“stripping”) of ovulated eggs, and reinstated by injecting ovulated eggs through the ovipore into the ovarian lumen (Stacey and Liley, 1974). Indomethacin, a PG inhibitor, blocks the onset of spawning behaviors following egg injection (Stacey, 1976), while injection of PGs induces spawning behaviors in both male and female goldfish and can overcome the indomethacin-induced blockade of reproductive behavior in females (Stacey, 1976; 1981). The behavioral effect of PGs is enhanced if the same dose of PG is introduced intraventricularly rather than via systemic injection (Stacey and Peter, 1979), suggesting a central site of action. Prostaglandins have also been shown to influence female proceptivity and receptivity in several tetrapod vertebrates. In Rana pipiens prostaglandins induce female receptivity to male clasps and inhibition of prostaglandin synthesis depresses AVT-induced receptivity (Diakow and Nemiroff, 1981), suggesting that AVT induces receptivity by stimulating PG synthesis. In contrast, receptivity of the female is blocked by prostaglandin treatment in the female lizard Anolis carolinesis (Tokarz and Crews, 1981) and in the guinea pig (Man-one, Rodriguez-Sierra, and Feder, 1979). Prostaglandins stimulate the occurrence of lordosis and solicitation in estrogen-primed, ovariectomized female rats and hamsters (Dudley and Moss, 1976; Hall, Luttge, and Berry, 1975; Rodriguez-Sierra and Komisaruk, 1978; Buntin and Lisk, 1979). Moreover, prostaglandins can induce lordosis responding within 20 min in ovariectomized female rats in the absence of estrogen priming. (RodriguezSierra and Komisaruk, 1977). Interestingly, the time course of the PG effect on lordosis reported by Rodriguez-Sierra and Komisaruk is similar to the time course seen in goldfish, suggesting that the PG effects on reproductive behavior may represent a general phenomenon in vertebrates. The anabantoid teleosts provide an appropriate model with which to test the generality of the effects of PGs on female teleost reproductive behavior, since the pattern of their reproductive behavior allows detailed analysis of hormone-behavior interactions. The elaborate reproductive sequence shown by M. opercularis (cf., Hall, 1966; Villars and DeNeff,
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
23
1979) includes a courtship of 1 to 4 days, a spawning sequence extending over 3 to 5 hr, involving tight synchronization between male and female behavior, and a parental phase of approximately 3 days in which the male defends and cares for the young. During the Courtship phase the male builds a nest of bubbles clustered at the water surface. He may frequently approach the female to “lead” her to the nest or to attack if she is unresponsive. This time period apparently is a function of developing female reproductive readiness. A study by Cheal and Davis (1974) suggested that female reproductive readiness of Trichogaster trichopterus, a species closely related to M. opercularis, is crucial to the initiation of the spawning sequence. If the female from a pair of T. trichopterus is replaced daily, spawning does not occur, while spawning occurs within the normal time if the male is replaced daily. The female M. operculuris also appears to play a major role in the initiation and termination of spawning (personal observation). The female initiates spawning by approaching a nestbuilding male, and she terminates the behavioral sequence when she has apparently depleted the store of ovulated eggs available for egg release. The Spawning phase begins when the female approaches the nestbuilding male. It continues for 3 to 5 hr and consists of repeated bouts in which the male curves his body around the female and the pair inverts under the nest (Villars and DeNeff, 1979). These curves and inversions are analogous to the mounts and intromissions of the rodent-mating sequence. Following each inversion a stream of approximately 10 to 50 eggs are released and fertilized in the stream of milt released concurrently by the male; the first inversions do not result in egg release, and earlier bouts of egg release produce fewer eggs. The male places the eggs in the bubble nest and cares for them by continually rearranging them and adding further bubbles to the nest. The male’s Parental behavior continues for several days after spawning is complete until the young are free swimming. The 300 to 500 young hatch in approximately 3 to 5 days. In a laboratory setting pairs may spawn repeatedly at intervals of 1 to 3 weeks (Hall, 1966; personal observation). The present studies were designed to determine whether prostaglandins play a role in the regulation of female reproductive behavior of M. opercularis comparable to the effect seen in goldfish. Experiment 1 tested whether exogenous PGF2, could induce sexual activity in females which had recently terminated spawning. Experiments 2 and 3 examined the effects of PGs on unspawned females which were placed with actively nestbuilding males. In Experiment 2, females were adapted to the test situation for 1 hr before testing. In experiment 3 females were adapted for 6 days prior to testing. Experiment 4 examined the ability of indomethacin, a PG synthetase inhibitor, to interfere with ongoing sexual behavior.
24
VILLARS,
HALE,
AND
CHAPNICK
METHODS Subjects
Subjects were sexually mature Mucropodus opercularis obtained from a local supplier. Experimental females weighed from 1.4 to 2.9 g (x: 1.9 + 0.4) and were 4.5 to 5.7 in. (x 5.0 + 0.3) in experiments 1 and 2. In experiment 4, mean body weight was 3.1 + 0.67. Subjects in experiment 3 were not weighed, but were from the same population as those in experiment 4. Fish of both sexes were housed together, in groups of 3050, in 80- or 120-liter aquaria, and fed once or twice daily with freezedried worms or Tetramin flakes. Under these holding conditions, males do not establish territories or nests, and mating does not occur. A photoperiod of 14L-10D was used in all experiments and was supplemented by natural light in experiments 1 and 2. Temperatures in the holding and test aquaria ranged from 21 to 24°C and pH ranged from 6.7 to 7.2. Spawning
Procedure
Pairs of fish were placed in 40-liter aquaria half-filled with distilled water into which 4 mg of sodium bicarbonate had been added (experiments 1 and 2) or with filtered well water (experiments 3 and 4) and were allowed to spawn normally according to the procedure of Villars and Davis (1977). Females were each placed in an aquarium first and the males were introduced 24 hr later to minimize male aggression against the female. Pairs generally spawned 2 to 10 days after the introduction of the male, somewhat more slowly than in earlier work (Villars and Davis, 1977). Occurrence of spawning is indicated by direct observation of spawning acts or the presence of eggs in the nest. On the day following spawning, “prespawned” females in experiment 1 were randomly assigned to one of three groups: PG treated (N = 8) vehicle control (N = 3) or handling control (N = 3). Prostaglandin
Treatment
(Experiments
1, 2, and 3)
Each female was removed from its tank, injected, and immediately returned. For the intraperitoneal injection, subjects were restrained in a fish net, a 26-gauge needle was inserted under the scales and through the body cavity wall 2- to 3-mm dorsal of the ovipore. A volume of 10 ~1 was delivered using a Hamilton microliter syringe. Hormone-treated subjects received a total of 500 ng of PGFzu (Tris salt, Sigma) dissolved in 0.1 M sodium carbonate. Vehicle controls received an equal volume of the sodium carbonate solution. Handling controls were removed from the tank and restrained. All subjects were immediately returned to the same tank from which they had been removed and observed at 15, 45, 75, and 105 min postinjection.
PROSTAGLANDINS
Procedure
for Unspawned
AND
FISH
REPRODUCTIVE
BEHAVIOR
25
Females
One-hour adaptation (experiment 2). Following testing and within 48 hr postspawning, the spawned females from experiment 1 were removed from the test aquaria. For experiment 2, unspawned females from the holding tank were placed with the nestbuilding, parenting males. These unspawned females were probably virgin since holding conditions are not conducive to spawning. The females were left undisturbed in the new environment for a period of 1 hr, then injected with PGF2, (N = 6), or with vehicle (N = 6), and observed. During the hour adaptation period, all females had begun foraging around the test environment. Six-day adaptation (experiment 3). For experiment 3, females were placed with unspawned males and left undisturbed for 6 days. Females which had not spawned after 6 days were injected with PGFz, (N = 6) or with vehicle (N = 5). Six spawned females were injected with PGF*, 1 day postspawning as an additional comparison group, and as a replication of experiment 1. Indomethacin
Treatment
(Experiment
4)
Following pairing, subjects were observed at approximately hourly intervals during the light period for early indications of the onset of spawning. Pairs which are observed to curve and invert can reliably be expected to complete the 2- to 5-hr spawning sequence (personal observation). Following the first inversion females were first observed for 10 min, then were removed from the tank and injected with a 35 (N = 3) or 70 (N = 6) pg dose of indomethacin in a IO-p1 volume. Control subjects (N = 6) were injected with the sodium carbonate vehicle. Subjects were continuously observed for the occurrence of sexual acts and egg release. Ten-minute behavioral observations were performed at 1 and 2 hr postinjection. Behavioral
Observation
Procedure
Behavioral observation periods were 10 min in duration. For all experiments, pairs were continuously observed for the occurrence of curves and inversions. In experiments 1, 2, and 3, subjects were observed in more detail at 15, 45, 75, and 105 min postinjection. In experiment 4, formal observations occurred at 1 and 2 hr postinjection. In all PG treatment experiments, subjects were observed for 10 min prior to injection (baseline). In experiments 1 and 2, the total number of curves, inversions, and long approaches was recorded for each subject during the formal observation periods. A curve was recorded when the female positioned herself and the male wrapped his body around her. An inversion was recorded when the male rolled his body under the female and the pair
26
VILLARS,
HALE,
AND
CHAPNICK
inverted. Curves and inversions are the primary behavior patterns Seen during the spawning sequence and they occur repeatedly through the 3- to 5-hr reproductive sequence. During normal spawning, egg release begins approximately 1 hr after the first inversion (personal observations) and occurs concurrently with inversions. A long approach was recorded when the female approached the nestbuilding male and remained under the nest for longer than 5 sec. Observers were uninformed (“blind”) regarding the treatment of subjects. In experiments 3 and 4, the total time the female spent under the nest was recorded along with several additional behavioral measures. The female was considered to be under the nest if she was within three body lengths of the nestbuilding male and under the nest he had built. This measure is seldom ambiguous, since females which are not under the nest with the nestbuilding male usually spend their time in the comer most distant from the male. In experiment 4, latency to egg release was measured from the first inversion, which had also signaled the start of the observation. That latency had been observed in earlier work (Villars and DeNeff, 1979) to be quite consistent among spawning pairs. In addition to time under the nest, subjects in experiments 3 and 4 were also scored for female orient/partial approach, female lateral display/ frontal display, female tailwaglbite-butt, male nestbuilding, and maleinitiated interactions. A female orient/partial approach (O/A) was scored when the female’s nose was pointed toward the male and she moved, with fins folded, directly toward him for at least two body lengths. This behavior was scored only when the female was not under the nest. Female lateral display/frontal display (LD/FD) was scored if the female, during an interaction with the male, flared her gill covers (frontal display) or completely extended her dorsal and caudal fins (lateral display). Tailwagging (TW) was recorded when a female assumed a vertical posture, spread her fins, and waved her caudal fin. Tailwagging was seen and recorded in response to male approach and following female approaches to the male. However, it was not recorded following a male attack on the female, since the tailwag might be construed as part of an escape or avoidance response. A bite-butt (BB) was recorded when the female approached a male and rammed him firmly on the flanks. Though this action has the appearance similar to an attack by the female, it differs in that the butts are directed more caudally than are attacks and males respond sexually rather than agonistically with lateral display, a partial or complete curve of the male around the female or by increased vigor of male nestbuilding. Female O/A, LD/FD, and TW/BB were all considered indictions of female sexual “motivation” or “proceptivity.” Male nestbuilding was scored using an informal time sampling procedure; at regular intervals during an observation period each tank was scored if the male was nestbuilding. Males were considered to be “nestbuilding
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
27
males” if they were observed nestbuilding at least once during the observation period. A male-initiated interaction was scored when the male left the nest area and approached the female. An approach was usually followed by a male LD or attack or both. The entire intraction was scored as one event. Male actions toward the female which occurred in response to a female orient/partial approach were scored as male-initiated behaviors. Statistics In experiments 1, 2, and 4, statistical significance was determined with the test for differences between proportions (Daniel, 1978). In experiment 3, the magnitude of the difference between the three groups in percent responding was examined using the chi-square test for independence (Sokal and Rohlf, 1969). Since the expected values in some cells were sometimes less than 5, the probability estimates are provided for their descriptive value, and 95% confidante intervals (Daniel, 1978) were calculated to allow comparisons between groups. RESULTS
Experiment
1: Effects of PGF,,
on Prespawned Females
Recently spawned, prostaglandin-treated females showed a dramatic increase in sexual activity. Seven of the eight prespawned pairs which were PG treated were observed to curve and/or invert (spawning acts) during the observation periods following injection. The eighth prespawned, PGF,,-treated female was observed curving in the interval between observations. None of prespawned controls were observed to curve or invert either during the observation period or in the interval between observations. Curves and inversions appeared normal except that egg release did not occur; inversions without egg release are also seen in the early stages of normal spawning. Prostaglandin treatment reinstated sexual activities and stimulated female approach (Fig. 1) within 15 min of the injection for most females. Two hours after the treatment none of the treated females were engaging in sexual activity or approaching the parenting male. At 45 min following the injection the maximum number of subjects were responding. However, for those subjects showing a short latency response to PGFI, the peak rates of that response occurred shortly after the onset of the behavioral effect. Experiment
2: Effects of PGF,,
on Unspawned Females
Unspawned females placed with parenting males for 1 hr did not respond to PGF2, treatment. During the baseline period immediately prior to PG treatment none of the females showed long approaches to the nestbuilding male or signs of sexual activities. Prostaglandin treatment did not change
28
VILLARS,
t 1* hi.
45
HALE,
AND CHAPNlCK
75
105
135
165
Minutes
Post
Injection
195
225
1Effects of PG treatment on mating behaviors. The median number of long approaches, curves, and inversions is plotted as a function of time since injection in PGtreated (0-O) and control (O---O) subjects.
any of the recorded measures and there were no other behavioral changes apparent to the observers. All pairs in both control and PG-treated groups spawned the followed day. Experiment 3: PGF,, Treatment of Unspawned Females Familiarized with a Male In Experiment 2, unspawned females placed with parenting males for 1 hr showed no response to PGF2, treatment. Experiment 3 tested whether more extended familiarity with the male would expose a PG effect in females which were not prespawned. Sexually mature, unspawned females were placed with males; pairs which had not spawned after 6 days formed
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
29
PG-treated (N = 6) or vehicle control N = 5) groups. Pairs which had spawned 1 day previously (N = 6) formed a prespawned, PG-treated comparison group. As in experiment 1, PG treatment reinstated reproductive activity only in pairs which had completed spawning (PG-X) the previous day (Table 1, column 3). Half the prespawned pairs curved or inverted 45 or 75 min postinjection. The proportion of prespawned (PG-X) females spending time under the nest and the mean time (in seconds) which prespawned (PG-X) females spent under the nest increased following PGF2, injection (Table 1, columns 4 and 5). In contrast, only one PG-treated unspawned (PG-N) female spent time under the nest and that time was less than the time spent by the experienced, PG-treated females. None of the vehicletreated (VEH) females spent any time under the nest or engaged in any curves or inversions. The percent proceptivity and the mean proceptivity scores of both unspawned (PG-N, VEH) groups were below those of the prespawned (PG-X) group in both the baseline and treatment periods (Table 1, columns 6 and 7). Proceptive acts included orient/approach, lateral/frontal display, tailwag, and bite-butt. During the baseline period the groups did not differ in the proportion of males nest building (Table 1, column 8). That proportion fell slightly in the vehicle-treated group but not in the PG-treated groups. Experiment 4: Effects of PG Synthesis Inhibition Occurring Spawning
on Normally
Neither of the two doses of indomethacin blocked spawning behaviors when injected 10 min after the first inversion; all pairs completed spawning regardless of treatment. There was no change in the percentage of pairs showing spawning acts (curves or inversions) following vehicle (83%) or indomethacin (80%) treatments and there were no significant differences between groups in the percentage of pairs observed spawning. There was a slight, nonsignificant trend among the high dose group toward increased occurrence of spawning acts; whereas four of the six highdose subjects were observed curving/inverting during the baseline, 100% did so after the indomethacin treatment. There was also no effect of either dose of indomethacin on mean latency to egg release. The mean latency to egg release, measured from the first inversion, was 157 + 59 min for fish receiving the high dose and 157 + 67 min for fish receiving the vehicle. Animals receiving the low dose had a mean latency to egg release of 154 rf: 22 min. Thus indomethacin did not affect latency to first egg release. Moreover, there was no effect of indomethacin on any of the other recorded behaviors. DISCUSSION
Prostaglandin treatment produces a clear and dramatic reinstatement of sexual behavior in previously spawned females. These results parallel
6 6 5
6 6 5
75 min PG-X PG-N VEH
Overalld PG-X PG-N VEH
y b ’ d e
33 ( 6-73) 0 ( O-40) 0 ( O-50) NS
6 6 5
45 min PG-X PG-N VEH
(15-85) ( O-40) ( O-50) P < 0.05
33 ( 6-73) 0 ( O-40) 0 ( O-50) NS
Percent P under nest”
83 17 0 9.70,
67 0 0 9.59,
67 17 0 6.56,
(50-99) ( l-60) ( O-50) P < 0.01
(27-94) ( O-40) ( O-50) P < 0.01
(27-94) ( l-60) ( O-50) P < 0.05 305 -
476 17
280 111
271 -
Time under nestb
TABLE 1 Adapted, Unspawned Pairs (PC-N) PC&Treated, l-Day Prespawned
67 (27-94) 17 ( l-60) 0 ( O-50) 6.56, P < 0.05
on 6-Day
Column indicates the percent responding, with 95% confidence interval in parenthesis. The mean time in seconds which the female spent under the nest during the IO-min observation. Average frequency of acts per 10 minute observation for only those counted as responders. The “overall” category is the combined responding for the 15-, 4S-, and 75-min observations. Chi square was calculated as an index of the reliability of the difference between the three groups.
50 0 0 6.68,
(15-85) ( O-40) ( O-50) P < 0.05
0 ( O-40) 0 ( O-40) 0 ( O-50) NS
6 6 5
15 min PG-x PG-N VEH
50 0 0 6.68,
0 ( O-40) 0 ( O-40) 0 ( O-50) NS
6 6 5
PG-X PC-N VEH
Percent curve invert”
of PC&Treatment
N -
Effects
Baseline
The
94) 60) 66) 0.05’
(60-100) (27- 94) ( 8- 66) P < 0.10
(27( l( lP <
100 (60-100) 67 (27- 94) 67 (20- 92) NS
100 (60-100) 17 ( l- 60) 40 ( 8- 66) 8.84, P < 0.025
100 (60-100) 33 ( 6- 73) 20 ( 1- 66) 8.44, P < 0.05
100 67 40 4.80,
67 17 20 6.56,
Percent proceptive”
5.0 2.0 2.5
6.1 2.0 1.0
8.1 2.2 1.5
5.2 1.0 1.0
Mean proceptive’
Compared to Vehicle-Treated, Pairs (PC-X)
(VEH)
100 (60-100) 100 (60-100) 60 (20- 92) 5.44, P < 0.10
100 (60-100) 83 (40- 99) 40 ( 8- 66) 5.70, P < 0.10
100 (60-100) 100 (60-100) 20 (20- 92) 12.55, P < 0.05
100 (60-100) 100 (60-100) 60 (20- 92) 5.44, P < 0.10
100 (60-100) 83 (40- 99) 80 (34- 99) NS
Percent 6 NB”
Unspawned
2.5 2.5 A
2 5
1
3.5 3.5 3
2.5 2.5 A
Mdn d Init’
and to
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
31
those reported for goldfish (Stacey, 1976, 1981), where PGs reinstated sexual behavior in females which had not ovulated or which had been stripped of their eggs. These results also are consistent with reports in other vertebrates indicating that prostaglandins enhance receptivity and proceptivity in Rana pipiens (Daikow and Nemiroff, 1981), as well as in rats and hamsters (Hall et al., 1975; Buntin and Lisk, 1979; RodriguezSierra et al., 1977, 1978). The time course of the PG effect appears remarkably similar across the vertebrates tested. We have observed an effect of PGF2, approximately 15 min postinjection and a peak at approximately 45 min. A similar time course was reported by Stacey (1978) for the goldfish and by RodriguezSierra and Komisaruk (1977) for the rat. The effects of PGF*, on female M. operculuris might best be described as reinstatement of proceptivity rather than receptivity. Female M. opercufuris are largely responsible for the initiation of reproductive behaviors. While the nestbuilding male may engage in “courtship” behaviors such as “leading to the nest,” the onset of sexual behavior is signaled by the female approach to the nestbuilding male (unpublished observations). Moreover, the approach of the female to the nestbuilding male is apparently not contingent on prior male courtship or solicitation; a steadily nestbuilding male is a suflicient stimulus. The onset of persistent female approaches toward the male is the first clear indication that spawning acts will be initiated. PG-treated females court/solicit the nestbuilding male, tailwagging and butting his flanks to initiate the sex acts of curving and inverting. Though those sex acts are often described as male actions, the PG induction of full mating seen here implicates the female hf. operculuris as the major initiator of mating behavior including sex acts. For the period of its effectiveness, PG-stimulated behavior looked indistinguishable from the normal pattern of spawning behavior. It should be noted that egg release did not occur during the inversions; however, inversions without egg release are also normally seen during the early stages of the spawning cycle. The details of the female interactions with the male appeared essentially the same as that seen at the height of spawning activity; the female hovered under the nest, assisting with egg retrieval and nestbuilding, and periodically butting the male in the flanks, thereby eliciting a lateral display, curve, or possibly an inversion. In naturally occurring spawning the female leaves the nest area for 1 or 2 min following an inversion; the PG-treated females behaved likewise. Thus, the behavior of both PG-treated and naturally spawning females shows a characteristic approach-retreat rhythm. The behavior change of one PG-treated, prespawned female was particularly striking in this regard. During the pretreatment period she had remained at the periphery of the nest, foraging and occasionally retrieving an egg which had drifted from the nest, returning the eggs closer to the male. Such behavior is
32
VILLARS,
HALE,
AND
CHAPNICK
characteristic of many females in the hours foliowing spawning, though they have usually moved to the opposite side of the tank by the day following spawning. However, following PG treatment, this same female began to show the approaches and retreats characteristic of normal spawning. Her approaches were directed to the male and to the center of the nestbuilding activity. Her retreats were to the opposite end of the tank. This was a striking change in the quality of the behavior pattern. Pollack, Thompson, Stabler, and Keenen (1981) have demonstrated that female Trichogaster trichopterus, an anabantoid closely related to M. opercularis, will engage in spawning acts with more than one male during the same spawning cycle. They suggest that the retreat following a successful inversion may facilitate location of other appropriate mates. Thus, the degree to which PG-elicited behavior mimics the normal pattern of reproductive behavior is striking and suggests that endogenous PGs are a factor in the onset of normal female proceptivity. Quantitative data are presently being gathered to allow more detailed comparisons between naturally occurring and PG-induced behavior. Sensitivity to the effects of prostaglandin treatment on M. opercularis appears to depend on the prior spawning of the female. This limited sensitivity contrasts with the reports in goldfish (Stacey, 1981) and rats (Rodriguez-Sierra and Komisaruk, 1978). Prostaglandin treatment stimulates proceptivity in female goldfish which are partially ovariectomized and PG treatment even induces female behavior patterns in male goldfish (Stacey, 1981). Likewise the effects of PG on rat sexual behavior do not require the presence of the ovaries or of hormonal priming (RodriguezSierra and Komisaruk, 1977). In contrast, we have only been able to demonstrate a PG effect in fish which have recently spawned. In experiment 2, unspawned females placed with males for 1 hr prior to PG injection show no behavioral effects of PG treatment. The 1 hr adaptation period preceding PG injection appeared adequate, since the previously spawned males had returned to nestbuildjng and parenting while the females were foraging, showing no signs of distress. The PG dose should have been adequate, since the same dose had completely reinstated sexual behavior in the prespawned females of experiment 1. Moreover the females were physiologically prepared for spawning, since all subjects in both control and PG treatment groups were observed to spawn the following day. It was possible that the I-hr adaptation period used in experiment 2 was insufficient. However in experiment 3, greater familiarity with the male and with the spawning environment did not unmask a PG effect in unspawned females; prostaglandin treatment produced no clearly apparent effect in unspawned subjects when compared to vehicle-treated controls. It is possible that 6-day unspawned females, though familiar with the male, were for some other nonspecific reason unable to demonstrate a response to PG. However, the most obvious reasons are ruled out; both
PROSTAGLANDINS
AND
FISH
REPRODUCTIVE
BEHAVIOR
33
males and females appeared healthy and were eating well. We took careful records of nestbuilding and the males were nestbuilding and seemed otherwise superficially like adequate social stimuli. It remains to be determined whether prespawning sensitizes the females to later PG treatment via a behavioral or an endocrine mechanism. In experiment 3 the proportion of males nestbuilding declined in the vehicle-treated group, while males in both PG-treated groups showed continued high levels of nestbuilding. It may be that males are sensitive to the endocrine state of the female. Indeed, male goldfish have been shown to preferentially court PG-treated females, leading Partridge, Liley, and Stacey (1976) to suggest that PG treatment may induce production of a pheromone which stimulates male behaviors. Such a pheromone in M. operculuris could serve to stimulate male nestbuilding. Further study with more sensitive measures of nestbuilding activity are necessary. It is clear that in M. opercularis indomethacin does not interfere with maintenance of ongoing sexual behavior. The low dose of indomethacin and the time course of testing which we used were comparable to those used by Stacey (1976) and by Diakow and Nemiroff (1981) to block sex behavior, in goldfish and Ram pipiens, respectively. Moreover, the higher dose which we used produced increases in frequency of spawning acts, which, though statistically insignificant, are clearly opposite to our predicted effect. At the higher dose of indomethacin there was some suggestion that egg release and time spent by the female under the nest may have been reduced slightly. The high dose may also have altered the ratio of curves to inversions. However, these effects were not reliable with the sample sizes we used and were minimal in comparison to the absence of a more general effect. Further study with indomethacin is in progress. On the one hand, prostaglandins may initiate sexual behavior without being necessary for maintenance of the cycle; indomethacin treatment earlier in the spawning cycle may reveal an effect of PG synthetase inhibition. Use of PG synthesis inhibitors such as ibuprofen or acetaminophen might also yield different results than found here. Alternatively, PGs may not be directly involved in the maintenance or initiation of sexual behavior. Exogenous PG reinstatement of sexual behavior may be operating indirectly. Inhibition of naturally occurring behavior following PG inhibition is essential to demonstrate a direct role of PGs in sexual behavior. To date, no such effect has been demonstrated in any of the species studies. In summary, PGFZO(reinstates sexual activity in female M. opercularis which have recently completed spawning. This drug-induced behavior appears remarkably similar to normally occurring behavior. However, unlike in the goldfish, sensitivity to PG in M. opercularis is restricted, such that unspawned females are largely unresponsive to PGF,, treatment.
34
VILLARS,
HALE, AND CHAPNICK
The nature and extent of the restrictions are not yet clear; endocrine and/or behavioral factors may account for the unresponsiveness of unspawned females. Finally M. operucularis do not show reductions in ongoing sexual behavior following treatment with a prostaglandin synthetase inhibitor. ACKNOWLEDGMENTS The authors thank Martha Burdick for her assistance with experiments 3 and 4. Thanks are also due to Nancy Shepardson for assistance with typing the manuscript.
REFERENCES Behrman, H. R. (1979). Prostaglandins in hypothalamo-pituitary Rev.
Physiol.
and ovarian function. Ann.
41, 685-700.
Buntin, J. D., and Lisk, R. D. (1979). Prostaglandin E,-induced lordosis in estrogen primed female hamsters: Relationship to progesterone action. Physiol. Behav. 23(j), 569-575. Cheal, M., and Davis, R. E. (1974). Sexual behavior: Social and ecological influences in the Anabantoid fish, Trichogaster trichopterus. Behav. Biol. 10, 435-445. Daniel, W. W. (1978). Applied Nonparametric Statistics. Houghton-Mifflin, Boston, Mass. Diakow, C., and Nemiroff, A. (1981). Vasotocin, prostaglandin, and female reproductive behavior in the frog, Rana pipiens. Horm. Behav. 15, 86-93. Dudley, C. A., and Moss, R. L. (1976). Facilitation of lordosis in the rat by prostaglandin E,. J. Endocrinol. 71, 457. Hall, D. D. (1966). An ethological study of three species of anabantoid fishes (Pisces, Belontiidae), Ph.D. dissertation, University of Oklahoma. Hall, N. R., Luttge, W. F., and Berry, R. B. (1975). Intracerebral prostaglandin E,: Effects upon sexual behavior, open field activity, and body temperature in ovariectomized female rats. Prostaglandins 10, 8784387. Marrone, B. L., Rodriguez-Sierra, J. F., and Feder, H. H. (1979). Differential effects of prostaglandins on lordosis behavior in female guinea pigs and rats. Biol. Reprod. 20 853-861. Partridge, B. L., Liley, N. R., and Stacey, N. E. (1981). The role of pheromones in the sexual behavior of the goldfish. Anim. Behav. 24, 291-299. Pollak, E. I., Thompson, T., Stabler, A. L., and Keener, D. (1981). Multiple matings in the Blue Gourami, Trichogaster Trichopterus. (Pisces, Belontiidae). Anim. Behav. 29 5563. Rodriguez-Sierra, J. F., and Komisaruk, B. R. (1977). Effects of prostaglandin Ez and indomethacin on sexual behavior in the female rat. Horm. Behav. 9, 281-289. Rodriguez-Sierra, J. F., and Komisaruk, B. R. (1978). Lordosis induction in the rat by prostaglandin E, systemically or intracranially in the absence of ovarian hormones. Prostaglandins 15, 513-525. Stacey, N. E. (1976). Effects of indomethacin and prostaglandins on the spawning behavior of female goldfish. Prostaglandins 12 113-126. Stacey, N. E. (1981). Hormonal regulation of female reproductive behavior in fish. Amer. Zool.
21, 305-316.
Stacey, N. E., and Liley, N. R. (1974). Regulation of spawning behaviour in the female goldfish. Nature (London) 247, 71-72. Stacey, N. E., and Peter, R. E. (1979). Central action of prostaglandins in spawning behavior of female goldfish. Physiol. Behav. 22, 1191-l 196. Sokal, R. R., and Rohlf, F. J. (1969). Biometry. Freeman, San Francisco, Calif. Tokarz, R. R., and Crews, D. (1981). Effects of prostaglandins on sexual receptivity in the female lizard, Anolis carolinensis. Endocrinology 109, 451-457.
PROSTAGLANDINS
AND FISH REPRODUCTIVE
BEHAVIOR
35
Turner, G. D., and Bagnara, J. T. (1976). General Endocrinology. Saunders, Philadelphia, Pa. Villars, T. A., and Davis, R. E. (1977). Castration and reproductive behavior in the Paradise Fish, Macropodus operculuris (L) (Osteichthyes: Belontiidae). Physiol. Behav. 19, 371-375. Wars, T. A., and DeNeff, S. J. (1979). Quantitative description of the patterns of reproductive behavior of Macropodus opercularis. Amer. Zool. 19, 935.