Bigamy or m o n o g a m y with maternal egg care in the triggerfish, Suffiamenchrysopterus M a s a h i r o I s h i h a r a 1'3 & T e t s u o K u w a m u r a 2

1Department of Fisheries, Faculty of Agriculture, Kyoto University, Kitashirakawa-oiwakecho, Kyoto 606-01, Japan 2Biological Laboratory, Faculty of Liberal Arts, Chukyo University, 101-2 Yagoto-honmachi, Nagoya 466, Japan (Correspondingauthor) 3present address: 9-25-302 Nozatocho, Kadoma-shi 571, Japan (Received19 March 1996;in revisedform 31 May 1996;accepted1 July 1996)

Ichthyological Research 9 The ichthyological Society of Japan 1996

Ishihara, M. and T. Kuwamura. 1996. Bigamy or monogamy with maternal egg care in the triggerfish, Surf&menchrysopterus. Ichthyol. Res., 43 (3): 307-313. Abstract Reproductive behavior and mating systems of the triggerfish, Sufflamen chrysopterus (Balistidae), were studied on the fringing reef of Sesoko Is-

land, Okinawa. Both males and females maintained territories against consexual adults, feeding on benthic animals within their own territories. Each male territory overlapped one or two female territories, with mating occurring between the cohabitants. The monogamous males were smaller and foraged more frequently than the bigamous ones, suggesting that the former allocated more energy to growth rather than to improving reproductive success. Pair spawning occurred around sunrise, females only taking care of the demersal eggs until hatching, which occurred around sunset of the same day. On spawning days females foraged less frequently than usual, but as frequently as males. Females spawned at intervals of 5-7 days, usually shifting sites within their territories. Thus both feeding and spawning sites were available for females within their territories, providing males with the opportunity to monopolize females by defending their territories.

Key words. - - Balistidae; female care; monogamy; bigamy; territory.

-ating systems of animals may be affected

M by the distribution patterns of mates and _

ecological resources (Emlen and Oring, 1977; Krebs and Davies, 1993). Various mating systems have been reported among fish species, and even within a species (Kuwamura, 1984, 1988; Shapiro, 1991; Turner, 1993). For example, facultative monogamy occurs in polygynous fishes, depending on environmental conditions (Fricke, 1980a; Barlow, 1984). Among triggerfishes (Balistidae), which are common on coral reefs (Myers, 1989), polygynous mating systems seem to be common (Fricke, 1980b; Thresher, 1984; Kuwamura, 1991; Gladstone, 1994). The halfinoon triggerfish, Suffiamen chrysopterus, was also noted as polygynous by Fricke (1980b), although the

number of females in a harem was not reported. On the other hand, Kawase and Nakazono (1994) observed monogamous mating in a lowdensity population of this species. As in several haremic balistids (Fricke, 1980b; Thresher, 1984), females take care of demersal eggs in S. chrysopterus (Kawase and Nakazono, 1994). Among balistid fishes, however, spawning behavior has been described in detail only in Sufflamen fraenatus, observed in an aquarium (Kawabe, 1984). We investigated the reproductive behavior and mating system of S. chrysopterus on a coral reef of Okinawa. In this paper we describe spacing and foraging patterns of males and females as well as courtship, spawning and parental-care behavior. Then we discuss factors affecting the

308

M. Ishihara & T. K u w a m u r a

Fig. 1. Territories of each male (thick line) and female (thin line) Sufflamenchrysopterus. Solid circles and triangles indicate spawning sites and shelters, respectively, of each female (shelters of females a and b were not found). Reef flat and sandy floor are shown by shaded and white areas, respectively.

mating system of the species, comparing with that of other balistids.

Materials and Methods In August 1985, we conducted underwater observations using scuba, on the fringing reef of Sesoko Island (26~ 127~ Okinawa, southern Japan. The study area, about 55rex 40 m (part of Area A in the map given by Kuwamura, 1991), covered the reef flat, reef edge and sandy floor in depths of 2-5 m at mean sea level. Five males and 8 females of Surf&men chrysopterus found in this area were individually discriminated on the basis of body and fin scars. We estimated their total length (TL) to the nearest cm underwater, and determined their sex by sexual dichromatism (see Kawase and Naka-

zono, 1994) and from reproductive behavior. The water temperature ranged from 26.6 to 28.8~ during the study period. During 16-31 August, we surveyed the study area every day, and recorded the location of each individual and egg clutches. On 16 and 22-24 August, however, none or only a few of the fish could be observed due to the heavy waves and water turbidity caused by typhoons. To confirm home ranges and spacing patterns, we conducted 10-min observations in the daytime (except dawn and dusk), 6-8 times for each male and 57 times for each female. We recorded swimming routes, feeding sites and frequencies, and interactions with conspecifics and other fishes. We regarded the area covering all the feeding sites and swimming routes (except for invasions: see Results) of a fish as its territory. Sufftamen chrysopterus feeds mainly on small benthic animals such as crustaceans (Kuwamura, 1991). The number of picking per 10 rain was averaged for each fish and regarded as its feeding frequency. Preliminary observations in September 1984 suggested that spawning occurred around sunrise and hatching around sunset. In August 1985, we attempted to observe courtship and spawning behavior for about 1 h around sunrise on 10 mornings (sunrise time shifted from 06:03 to 06:09). We observed hatching and related behavior for about 1 h around sunset on 5 evenings (sunset time shifted from 19:03 to 18:54), using a light occasionally. We followed the subject fish until it entered its shelter (i.e., sleeping site: rock holes on the bottom). To examine diurnal changes in parental behavior, we conducted 10-min observations 6 times each in the early morning (06:05-07:40), around noon (11:30-14:15) and in the late afternoon (18:00-19:00). The 10-rain observations were made once in each period of a spawning day for 4 females, but for the remaining females data from some periods are lacking. We recorded the frequency and duration (sec) of fanning (aerating) eggs in the nest, frequency of attacks against other fishes, and feeding frequency and sites. Data from different individuals on different days were pooled for comparisons between the periods; the early morning period was further di-

309

R e p r o d u c t i v e Behavior of Triggerfish vided into 2 periods (before and after 06:40) in the analyses.

37.4 picks per 10min vs. 16.1-25.8; Table 1). The territory sizes of females were not significantly correlated with body size (Kendall's ~-= 0.31, n = 8, p >0.05; Table 1). Instead, the female territories established on the reef flat (1122 0 8 m 2, n = 3 ) were larger than those on the sandy floor (60-110m 2, n = 5 ; Mann-Whitney U=0, p<0.05). The feeding frequencies of females on non-spawning days were correlated neither with territory size (Kendall's r = - 0 . 2 1 , n=8, p>0.05), nor body size (Kendall's "r= -0.08, n=8, p>0.05; Table 1). The feeding frequencies of females (34.8-50.6 picks per 10 min, n = 8 ) were significantly higher than those of males (16.1-37.4, n = 5 ; Mann-Whitney U = I , p<0.01).

Results Territory arrangement and feeding frequencies Males and females maintained territories against consexuals (Fig. 1). Territorial border fights rarely occurred (6 times between males and 5 between females), as did invasion of the territories of other consexuals (once by male B into A's territory). Males and females foraged only, but very widely, within their own territories. Male territories overlapped one or two female territories, mating occurring only between the cohabitants. The territories of monogamous males (A and B) covered mainly the sandy floor, while those of bigamous males covered the sandy floor (D), reef flat (C) or both (E) (Fig. 1). Males (14-19 cm TL) were always larger than their mates (11-15 cm; Table 1). The two monogamous males (14-15 cm) were smaller than any of the three bigamous males (16-19 cm), the territory sizes of the former (106-118m 2) also being smaller than those of the latter (180284 m2). The monogamous males tended to feed more frequently than the bigamous ones (33.0-

Spawning behavior Spawning behavior was seen 6 times between 05:50-06:19, being from 16min before to 10 min after sunrise. Courtship and spawning sequences were as follows. In the early morning, the female began to pick at the substrate (sand and coral rubble) several minutes after leaving her shelter within the territory. The female attacked other species, such as labrids, approaching the picking site. The male visited the female some 4-10 min prior to spawning, conducting lateral displays with his unpaired fins flattened.

Table 1. Body size, territory size and feeding frequency of male and female Sufflamen chrysopterus Males

Females

TL (cm)

Territory size (m2)

Feeding per 10 min

TL (cm)

Habitat

Territory size (m2)

Feeding per 10 min on NSD

Feeding per 10 min on SPD

A(14) B (15) C (19)

118 106 284

33.0 37.4 25.8

D (16)

232

17.5

E (17)

180

16.1

a(12) b (13) c~ (15) c2 (14) d 1 (13) d2 (12) e~ (11) e2 (13)

sand sand reef reef sand sand sand reef

72 86 112 208 60 110 102 114

50.6 45.9 46.2 49.0 47.5 41.7 48.5 34.8

13.0 23.0 23.3 18.4 19.2 18.0 29.3 --

NSD, non-spawning days; SPD, spawning days; --, no data.

310

M. Ishihara & T. K u w a m u r a sneaking never occurred. The transparent eggs were attached to the sand within an area of about 6-7 cm diameter. Just after spawning, the male usually left the nest site to forage, or to visit the second female if bigamous, but the female remained to begin fanning the eggs in a head-down posture. Seven of 8 females spawned 2-3 times at intervals of 5-7 days during 17-31 August (Table 2). The female (e2) that spawned only once foraged in the lowest frequency among the females on non-spawning days (Table 1), although the correlation between the feeding and spawning frequencies was not significant (Kendall's "r= 0.56, n=8, p>0.05). Spawning was not related to the lunar cycle, but most females spawned synchronously, especially after the typhoon of 22-24 August (Table 2). The monogamous males (A and B) succeeded in spawning less often (2-3 times) than the bigamous males (4-5 times) during the period 17-31 August. Both females of a harem spawned on the same morning on 2 out o f 3 spawning days for male D, on 1 o f 3 days for male E, but on none o f 5 days for male C (Table 2). Females usually spawned at different sites within their territories on different days, such sites not always being situated near the females' shelters (Fig. 1). In the evening just prior to the spawning day, females sometimes (6 out of 17 observations) showed aggressive behavior against other fishes such as wrasses (Labridae) and monoclebreams (Nemipteridae). Only in one

Fig. 2. Spawning posture of the male (left) and female Suffiamen chrysopterus (Photo by M.I.).

The male's throat and belly were bluish-purple and the head and dorsal surface pale-brown, while the female's body color was dark brown as usual. The female occasionally exhibited dipping behavior, i.e., swimming upward and then rapidly downward about 3 0 c m with the body quivering and the unpaired fins flattened, and tried to lead the male toward the picking site (i.e., nest site). While the female continued picking at the bottom, the male began to nuzzle the former's caudal peduncle and belly. Within several minutes of nuzzling, the male and female settled on the nest side by side with their bellies touching, and then released gametes onto the nest for a few seconds (Fig. 2). During the courtship and spawning, no other males approached the pair;

Table 2. N Aug. 16 Female a b c1

S -

C2

--

dI

-

d 2

-

e1

-

e 2

--

17

18

19

20

Spawning intervals of each female

21

22

S

-

S S S

23

24

25

26

27

28

29

F 30

31

S S S

S

S S S

S S S S

S

-

S

-

S

-

N, new moon; F, full moon. S, spawning and/or egg care observed; -, no survey due to typhoons.

311

Reproductive Behavior of Triggerfish of the 6 cases, however, was the site where the aggression occurred used for spawning the following morning.

300 S

a @

2oo

/ /

9

9

M a t e r n a l e g g care

O) tO tU I.l.

In the early morning of the day following spawning, no eggs were found in any nests (n = 18). Evening observations (n=5) revealed that hatching began from 18:54-19:00, being from 4min before to 5 min after sunset and about 13 hrs after spawning. At that time, females began to blow the eggs, circling within the nest. Blowing, as well as fanning, continued for 921 min until the females left for their shelters at 19:06-19:20. Egg care was conducted only by the female parent, which began fanning just after spawning. The total time allocated to fanning (42-297 sec per 10min, n=18) tended to decrease after 06:30 but increased again toward the evening (Fig. 3a). Females often (6-23 times per 10min) left the nest for a time to attack or feed. Females attacked various fishes approaching or passing their nests. No predation on eggs was observed. In total (131 times), 26 fish species of 12 families were attacked and a starfish was removed from the nest by mouth in one case. The most often attacked were labrids (49 times on 5 species, including 25 Thalassoma lunare), followed by mullids (24 on 4 spp.), nemipterids (12 on 2 spp.), pomacentrids (11 on 3 spp.) and chaetodontids (8 on 3 spp.). The frequencies of attacks (0-21 per 10 min) did not differ significantly between the periods (Fig. 3b). Females foraged 0-41 times per 10rain (n= 18) in the intervals between fanning and attacks, but only within a few meters of the nest. Feeding frequencies decreased in the evening just before hatching (Fig. 3c). Female feeding frequencies on spawning days were significantly lower than those on non-spawning days (Wilcoxon signed-ranks test, T=0, n=7, p < 0.05), but did not differ from those of the males (Mann-Whitney U=13, n1=7, n2--5, p>0.05; Table 1). Males did not approach the nests during 6 of the 18 10-rain observations of female egg care;

100

i

i

i

r

b

20 CO 0

Q

40

"C j/

\

G

-o

x

20 \

9

i

i

i

)6:05- 06:5006:30 07:40

Time of

11:3014:15

o 18:00-

19:00

day

Fig. 3. Diurnal changes in fanning time (a), frequency of attacks (b) and frequency of feeding (c) in egg-caring females. Each dot designates data obtained per 10-min observation. Medians of each period are connected by broken lines; *p<0.05 (MannWhitney U=2).

no difference was found between monogamous and bigamous males (2/5 vs. 4/13; Fisher's exact probability test, p>0.05). In the remaining cases, males stayed within a few meters of the nest (2 cases of bigamous males when only one of the females spawned), or visited the egg-caring female at least once per 10min (n= 10). Even in these 12 cases, the males never entered the nest to conduct fanning, and only in 2 cases defended the nest against other fishes such as wrasses and monoclebreams. Males as well as females sometimes became aggressive against the observer when the latter approached their nests.

312

M. Ishihara & T. K u w a m u r a

Discussion The present study confirmed both monogamy and bigamy in Sufftamen chrysopterus. The distribution patterns of resources and females may affect the degree of polygyny or male monopolization of females (Emlen and Oring, 1977; Krebs and Davies, 1993). Low population density may result in facultative monogamy in haremic species, as has been reported for the balistids Pseudobalistes fuscus (Fricke, 1980b) and S. chrysopterus (Kawase and Nakazono, 1994). In the present study area, however, S. chrysopterus occurred in rather high densities compared with other balistids (Kuwamura, 1991). Since S. chrysopterus feeds on benthic animals, female territories appear to function principally for defense of food resources, although not all other benthos-feeding fishes are excluded (Kuwamura, 1991). If females established smaller, contiguous territories, males would have the opportunity to control more females. In the study area, the females inhabiting the sandy floor had smaller territories than those on the reef flat. However, 2 of the 3 males on the sandy floor were monogamous, while 2 males on the reef flat were bigamous, suggesting no relation between female territory size and male monopolization. Besides the distribution pattern of females, male-male competition may also affect monopolization of females. The bigamous males were larger than the monogamous ones, although the sample size was small in this study. The former had higher reproductive success, but foraged less frequently than the latter, suggesting the incurrence of costs in maintaining (patrolling) larger territories or monopolizing more than one female. In other words, it is probable that the monogamous males could have allocated more energy to growth through their higher feeding frequencies, instead of improving reproductive success. Long-term observations are needed to confirm whether monogamous males later become bigamous as they grow larger. Courtship and spawning behavior of S. chrysopterus observed in this study was similar to that of'its congener S. fraenatus, observed in

an aquarium (Kawabe, 1984), although female dipping behavior was not reported from the latter. In both species, females prepare nests and occasionally display leading behavior, while males actively court by nuzzling only just before spawning. By contrast, males prepare nests and actively court visiting females in a lek-like species, Pseudobalistes flavimarginatus (Gladstone, 1994). Gladstone suggested that the sites suitable for nesting by the latter occurred in a restricted area so that individuals necessarily migrated there to spawn in a lek-like system. In case of S. chrysopterus, however, sites suitable for nests were not limited, as females often shifted spawning sites within their territories. Thus by defending the female territories that included not only food but also suitable spawning sites, males were able to monopolize the females. Although P. flavimarginatus exhibits biparental care, with females aerating the eggs and males chasing away egg predators (Gladstone, 1994), maternal care seems to be the norm among haremic balistids (Fricke, 1980b; Thresher, 1984). In S. chrysopterus, males never conducted fanning and rarely chased other fishes around the nest (also see Kawase and Nakazono, 1994). The attitude of males, monogamous and bigamous, to the eggs did not appear to vary. One parent seems to be sufficient for egg care, since no predation on eggs occurred during the present study. Probably the cost of parental care may not be so high for female S. chrysopterus, because it continued only from dawn to dusk within a single day, as is known in some other balistids (Fricke, 1980b; Thresher, 1984). Moreover, female S. chrysopterus could still feed on egg-caring days (as frequently as males), although less frequently than on non-spawning days. The spawning interval of each female observed in this study was 5-7 days, which was similar to that (4-6 days) reported by Kawase and Nakazono (1994). Both studi~s suggested no relation to the lunar cycle. By contrast, semilunar spawniffg cycles have been reported for P. flavimarginatus (Lobel and Johannes, 1980; Gladstone, 1994) and suggested for Rhinecanthus aculeatus (Thresher, 1984) living in Pacific

313

R e p r o d u c t i v e Behavior of Triggerfish

coral reefs. The latter living around the present study area of Sesoko Island also spawns in a semi-lunar cycle (T. Kuwamura, unpubl.). It is unknown at present why the two coexisting species, R. aculeatus and S. chrysopterus, exhibit different spawning cycles. However, it seems possible that the water conditions caused by the typhoons may have affected the spawning cycle of S. chrysopterus in the present study, resuiting in synchronous spawning unrelated to the lunar cycle.

Acknowledgments.- We are grateful to the staff of Sesoko Station, Tropical Biosphere Research Center (formerly, Sesoko Marine Science Center), University of the Ryukyus for providing the facilities for field work. We also thank Izumi Akagawa and two anonymous reviewers for their critical reading of the manuscript. This work was partly supported by Grants-inAid (Nos. 59115004 & 60107002) to T.K. from the Japan Ministry of Education, Science and Culture. This is Contribution No. 290 from the Sesoko Station.

Literature Cited Barlow, G. W. 1984. Patterns of monogamy among teleost fishes. Arch. FischWiss., 35: 75-123. Emlen, S. T. and L. W. Oring. 1977. Ecology, sexual selection and the evolution of mating systems. Science, 197: 215-223. Fricke, H. W. 1980a. Control of different mating systems in a coral reef fish by one environmental factor. Anim. Behav., 28: 561-569. Fricke, H. W. 1980b. Mating systems, maternal and biparental care in triggerfish (Balistidae). Z. Tierpsychol., 53: 105-122. Gladstone, W. 1994. Lek-like spawning, parental care

and

mating

periodicity

of

the

triggerfish

Pseudobal&tes flavimarginatus (Balistidae). Env. Biol. Fish., 39: 249-257. Kawabe, R. 1984. Spawning behavior of the bridled triggerfish, Sufflamen fraenatus, in the aquarium. Japan. J. Ichthyol., 31: 193-197. (In Japanese with English abstract.) Kawase, H. and A. Nakazono. 1994. Reproductive behavior of the flagtail triggerfish, Suffiamen chrysopterus. Proc. 7th Int. Coral Reef Syrup., Guam, 2: 905-907. Krebs, J. R. and N. B. Davies. 1993. An introduction to behavioural ecology. Blaekwell Scientific Publications, Oxford. 420 pp. Kuwamura, T. 1984. Social structure of the protogynous fish Labroides dimidiatus. Publ. Seto Mar. Biol. Lab., 29:117-177. Kuwamura, T. 1988. Parental care and social systems in fishes. Kaimeisha, Tokyo. 136 pp. (In Japanese.) Kuwamura, T. 1991. Habitat segregation, coexistence or interspecific territoriality between two triggerfishes, Rhinecanthus aculeatus and Surf&men chrysopterus, with notes on distribution of other balistids at Sesoko Island, Okinawa. Galaxea, 10: 65-78. Lobel, R S. and R. E. Johannes. 1980. Nesting, eggs, and larvae of triggerfishes (Balistidae). Env. Biol. Fish., 5:251 252. Myers, R. E 1989. Micronesian reef fishes. Coral Graphics, Guam. 298 pp. Shapiro, D. Y. 199l. Intraspecific variability in social systems of coral reef fishes. Pages 331-355 in P. E Sale, ed. The ecology of fishes on coral reefs. Academic Press, San Diego. Thresher, R. E. 1984. Reproduction in reef fishes. T. E H. Publications, Neptune City. 399 pp. Turner, G. E 1993. Teleost mating behaviour. Pages 307-331 in T. J. Pitcher, ed. Behaviour of teleost fishes. 2nd edition. Chapman & Hall, London.