Spawning Hawkfish

Environmental Biology of Fishes Vol. 24, No. 4, pp. 295-300, 1989 0 Kluwer Academic Publishers, Dordrecht.

Pelagic spawning of the hawkfish Oxycirrhites

typus

(Cirrhitidae)

Terry J. Donaldson 1,4& Patrick L. Colin2,3 ’ Section of Ichthyology, Museum of Natural Science and Department of Zoology and Physiology, 119 Foster Hall, Louisiana State University, Baton Rouge, LA 70803, U.S.A. 2Motupore Island Research Station, University of Papua New Guinea, Papua New Guinea 3 Present address: Caribbean Marine Research Center, cfo Florida State University Marine Laboratory, Box 456, Sopchoppy, FL 32358, U.S.A. 4Reprint requests. Received 30.3.1988

Rt. 1,

Accepted 22.7.1988

Key words: Behavior, Courtship, Eggs, Habitat availability, Site-attachment,

Social group, Spawning ascent

Synopsis Pelagic spawning of the deep slope coral-dwelling cirrhitid Oxycirrhites typus was observed for two social groups at Papua New Guinea. This species was previously reported to be a demersal spawner in an aquarium. Courtship in social groups consisting of a single male and one or two females commenced just prior to or after sunset among the branches of gorgonian or antipatharian corals. Males and females occupied separate corals; males either visited females at their corals or met them at an adjacent coral just prior to courtship. Courtship was sequential and consisted of two or more bouts with each female that culminated in a rapid ascent into the water column and the release of floating eggs. Fertilized eggs, taken from a third social group, were spherical and averaged 0.69 mm in diameter. Spawning pairs sought refuge in their resident corals or in the coral where courtship occurred immediately after spawning was completed.

Introduction The long-nosed hawkfish Oxycirrhites typus is a distinctive tropical to sub-tropical cirrhitid distributed from Baja California, Mexico west to the Red Sea. This species is typically found on deep slopes of coral and rocky reefs and inhabits gorgonian and antipatharian corals at depths of 25-100 m (Randall 1963). There is some published information concerning its reproductive behavior. Takashita (1975) and Tanaka et al. (1985) briefly described apparent courtship behavior in aquaria. Another observation suggested the possibility that this species might lay demersal eggs (reviewed in Thresher 1984). This suggestion has been cited elsewhere (i.e. Randall 1985).

Thresher (1984) questioned demersal spawning of this species as a result of field observations of pelagic spawning by another cirrhitid, Cirrhitichthys oxycephalus, in the Gulf of California. He believed the existence of two modes of spawning within a small, monophyletic family of 34-35 species was unlikely. Additional species of cirrhitids have been spawning pelagic eggs, including Cirrhitichthys falco (Donaldson 1986, 1987), C. aprinus (Donaldson, unpublished), C. aureus (Y. Yogo, personal communication), and Paracirrhites arcatus, P. forsteri and Cirrhituspinnulatus (Donaldson unpublished). Recently, we observed courtship and pelagic spawning of Oxycirrhites typus near Motupore Island, Papua New Guinea, where this species occurred in relatively shallow (20-26m) depths.

296 Methods and study sites Observations of courtship and spawning were made using scuba between 6-15 November, 1986 at times to include sunset and dusk (1730-1830 h). Additional observations of behavior and habitat use were made during occasional daylight dives within these dates. Data were recorded on plastic slates, and underwater photographs of courtship and spawning were taken. Time of sunset was determined from meteorological tables. Water visibility was estimated from known distances to subjects at approximately 20-25 m. Spawning was observed at two sites located about 16 km SE Port Moresby. The first site was located on the slope of a broad fringing reef ca. 500m south of Loloata Island at a depth of 22-23 m. This site had poor water visibility, ca. 3-6m and water temperatures between 2627” C. The second site was located at part of the Papuan Barrier Reef, locally called Horseshoe Reef, ca. 6 km south of Motupore Island. Water temperature was 285°C. Body sizes were estimated in the field by measuring the distance between two points on the substratum relative to the observed position of the tip of the snout and the edge of the caudal fin for a given individual. Fertile eggs were obtained by collecting adults with quinaldine-alcohol solution shortly before (within one hour) spawning. Fish were returned to the laboratory alive, their body sizes measured, and their gametes stripped by gentle hand pressure shortly after sunset (when they would have naturally spawned) the same evening. Gametes were mixed in a petri dish containing seawater and later embryos that hatched were measured using an ocular micrometer. Eggs, embryos and larvae were preserved in 3% formalin and specimens deposited in the Australian Museum, Sydney (AMS) .

Results Three social groups were studied. Courtship and spawning were observed from a group consisting of a male (72 mm TL) and two females (71 and 67 mm TL, respectively) at the Loloata Island site and

from a pair (male- 65 mm TL; female- 55 mm TL) at the Horseshoe Reef site. In the former group, the male and the largest female lived among the branches of separate antipatharian corals; the smaller female lived on a large Subergorgia sp. sea fan located less than 0.5m away from the larger female’s coral. In the latter group, both male and female were observed at 23-26m in two Subergorgia sp. sea fans, located 3m apart. Gametes were taken from a second pair collected from an antipatharian coral near the Loloata Island site (male- 99.5 mm TL; female- 92 mm TL). Courtship and spawning of 0. typw was observed off Loloata Island on five consecutive nights, 6-10 November, 1986 and at Horseshoe Reef, 15 November. Courtship began 23 minutes before to 22 minutes past sunset at Loloata Island and 13 minutes past sunset at Horseshoe Reef. The male at the Loloata Island site left his coral and made his way, either swimming between the coral branches or along the bottom, to the first female, located less than 0.5 m away. Once there, he waited, dorsal fin erect, until the female arrived from the upper branches of her coral. Occasionally, the male swam into the upper branches of the coral in search of the female. As the male approached, the female swam toward him, dorsal fin erect. Daily courtship consisted of several bouts of activity during a period which ranged from 12-19 (x = 5.3) minutes in duration. Generally, the male and female circled horizontally or vertically around one another among the branches of the coral and occasionally settled upon the branches to rest. The pair rested by aligning their bodies l-3 cm apart in parallel or anti-parallel position (see Donaldson 1986, 1987, for description of cirrhitid motor patterns). As courtship progressed, the female led the male in short swimming movements (‘lead and follow’; Donaldson 1987) through the branches of the coral, each movement separated by periods, 3-5 seconds in duration, of parallel or anti-parallel rest. During rest periods of longer duration the male moved his body close to the female and nudged her abdomen or flank below the dorsal fin with his snout (Fig. 1). Then the male often mounted the female by placing his body along her upper flank, adjacent to or touching her dorsal fin; balance was

297

Fig. 1. Nudging behavior by a male Oxycirrhites typus (the male is the lower fish).

Fig. 2. Descent after pelagic spawning by a pair of Oxycirrhires fypus.The female is partially obscured by the edge of the coral.

298 maintained by extending a pectoral fin against her body while he beat his caudal fin vigorously. The male also swam from his position parallel to the female, looped to pass in front of her and exposed his flank perpendicularly to her snout, before coming to rest parallel to her opposite flank. Just prior to spawning, the pair assumed parallel or nearparallel positions, lifted their snouts upward by raising their bodies with extended pelvic fins, raised their dorsal fins and fin spines, and quickly swam up or out from the coral to a point 20-25 cm away. There, they turned their snouts downward, flexed their bodies rapidly, and released a small cloud of gametes, which remained visible for 2-3 seconds in the fading light; no egg predation was observed although egg predators, usually pomacentrids, were present nearby. Immediately after spawning, the pair quickly swam back to the branches of the same coral (Fig. 2) and briefly assumed a position parallel to one another before the male departed. After spawning with the first female, the male swam to the second female on the upper edge of the Subergorgiu sea fan OSm away. Daily courtship bouts with the second female commenced 4-12 minutes past sunset and lasted 2.1-9.7 minutes before spawning occurred. On one occasion, the first female did not spawn with the male and in turn he did not join the second female, who remained on the edge of the sea fan until total darkness had fallen (ca. 1840 h). Courtship and spawning behavior was the same as with the first female, although fewer bouts of courtship (x = 3.3) were observed per period before spawning. On one exception (6 November), however, the male and the second female were observed at parallel rest, their bodies 1 cm apart, on the edge of the sea fan. There, they positioned their vents out past the edge over open water and both quivered together during three periods of l-2 seconds each over the space of one minute. Gametes may have been simultaneously released by each during these brief periods. Afterwards, the pair repositioned themselves a few cm away and repeated the behavior four more times in succession without ascending into the water column. Large predatory fishes were observed in close

proximity to the courting 0. typus, and it is possible that the pair spawned from the edge of the sea fan rather than making a spawning ascent above the coral and exposing themselves to the risk of predation. Alternately, these behaviors may merely have been a demonstration of male responses to an apparent lack of readiness on the part of the female, After spawning had been completed with the second female, the male slowly swam back to his coral by way of the first female’s coral; the second female sought shelter within the sea fan. Courtship and spawning behavior of 0. typus at Horseshoe Reed was similar, although the onset of courtship was after rather than before sunset. Prior to the onset of courtship, both male and female rested and fed in separate Subergorgia sea fans. Sometime just after 1730 h (32 minutes before sunset), the female moved slowly from her sea fan and swam upward along the reef slope, past the male’s sea fan and two adjacent but shallower sea fans, and eventually moved into the branches of a large antipatharian coral ca. 6 m from her starting position. This coral was located at the base of a steep reef face at 23 m depth. At 1800 h the male traced the female’s path to the large antipatharian coral where the female was sitting. However, the male’s progress was much slower as he moved from one point of cover to the next before arriving at 1812 h. Courtship commenced at 1815 h (13 minutes past sunset) when the male ascended into the upper branches of the coral and joined the female. Courtship consisted of two bouts, each in separate parts of the coral, and was brief (3 minutes and 4 seconds total duration) compared with courtship durations of the Loloata Island social group. After the spawning ascent, the male and female moved into shelter within the branches of the coral. Fertilized eggs taken from the third social group were spherical and averaged 0.69mm in diameter (+ 0.23; N = 17), slightly larger than that of ripe ova of C. oxycephalus reported in Thresher (1984). Eggs taken later from a fourth group, collected 5 May, 1987 and stripped at 2000 h, hatched after 15 hours incubation at 27°C. Attempts to rear the embryos were unsuccessful.

299

Discussion Courtship and spawning behavior of 0. typus was similar to that described from field studies for the cirrhitids (Thresher 1984, Donaldson 1986, 1987). Aspects in common included: (1) onset of courtship just before or after sunset with spawning after sunset, (2) successive courtship and spawning by the male in social groups containing one or more females, (3) courtship and spawning at a site common to both the male and female or at a site located within the female’s home area, (4) pelagic spawning with a rapid but relatively short ascent into the water column by both male and female, (5) the return of the male and female to resident corals after spawning, except at Horseshoe Reef where the pair moved into shelter in the same coral used for courtship. Pelagic spawning at or near dusk is found among many tropical reef fishes. Spawning ascents into the water column above the substratum allow the release of planktonic eggs some distance above the bottom, minimize immediate egg predation by benthic organisms, and facilitate transport of eggs away from the spawning site (see Thresher 1984 for review). Spawning adults are at risk from piscivores, but this threat may be reduced by rapid dashes to minimal heights where eggs can be released (Moyer 1987). Oxycirrhites typus spawning was rapid (ca. l-2 seconds in duration) and ascent lengths were relatively short (<0.25 m). Observed ascent lengths were low compared to other cirrhitids (C. oxycephafus, 0.3-l.Om, Thresher 1984, personal observation; C. fulco, 0.20.6m, Donaldson 1986,1987; and C. uprinus, 0.30.6m, Paracirrhites arcatus, 0.4l.Om, P. forsteri, 1.0-2.5 m, and Cirrhitus pinnulatus, 1.52.0m, personal observation), and may reflect the initial distance from the reef where the spawning ascent starts from a large gorgonian or black coral. Although planktivorous and piscivorous fishes were present at both spawning sites, no attempted predation on spawning adults or released eggs were observed. At the Loloata Island site water visibility was considerably less and this, because of reduced ability to detect predators, could have influenced the distance moved by fishes between resident corals and spawning sites. At Loloata the male trav-

elled only 0.5 m between corals while at Horseshoe Reef the male and female moved slightly over 10 m to the shallower spawning site although both resided in corals located less than 2 m apart. Movement between corals by 0. typus at Horseshoe Reef was slower compared with those at Loloata but the distances travelled were much greater. Movement of 0. typus at Horseshoe Reef consisted of quick dashes between patches of cover (corals, rocks and epifauna) alternating with periods of rest under cover for up to five minutes. Both male and female maintained erect dorsal spines while moving. This behavior has been observed in other cirrhitids, including C. f&o (Donaldson 1987) and C. aprinus (personal observation). Popular accounts of 0. typus suggest that this strongly site-attached species occurs in pairs, may be monogamous, and seldom, if ever, leaves the gorgonian or antipatharian coral where it resides. In Papua New Guinea, this species occurred in social groups consisting of a single male and one or two females. This species used one or more corals within a given area, moved freely between those corals, and formed pairs prior to courtship. Although the social organization and reproductive behavior of only a few social groups were described here, the patterns of organization and reproduction were consistent with those observed for other cirrhitids (Thresher 1984, Donaldson 1986, 1987). Social organization and reproductive behavior of 0. typus may depend upon habitat availability (i.e. suitable corals). The subsequent acquisition of this habitat is by: (1) emigrating adults, in search of potential mates, or by (2) newly-settled larvae, which may be incorporated into existing social groups when they reach maturity.

Acknowledgements We thank G.B. Constantino, J.M. Fitzsimons, J.T. Moyer, J.E. Randall, and R.E. Thresher for their comments on the manuscript or useful discussion. L.J. Bell, A.E. Trumble and M.C. Wilkins are thanked for their kind assistance. The comments of two anonymous reviewers are appreciated. Research by the senior author was generously sup-

ported by an award from the Coypu Foundation. Field operations were funded by the University Research Committee, University of Papua New Guinea.

References cited Donaldson, T.J. 1986. Courtship and spawning behavior of the hawkfish Cirrhitichthys f&o at Miyake-jima, Japan. Jap. J. Ichthyol. 33: 325L333. Donaldson, T.J. 1987. Social organization and reproductive behavior of the hawkfish Cirrhitichthys falco (Cirrhitidae). Bull. Mar. Sci. 41: 531-540. Moyer, J.T. 1987. Quantitative observations of predation dur-

ing spawning rushes of the labrid fish Thalassoma cupido at Miyake-jima, Japan. Jap. J. Ichthyol. 34: 76-81. Randall, J.E. 1963. Review of the hawkfishes (Family Cirrhitidae). Proc. U.S. Nat. Mus. 114: 389-450. Randall, J.E. 1985. Guide to Hawaiian reef fishes. Harrowood Books, Newton Square. 70 pp. Takashita, G.Y. 1975. Long-snouted hawkfish. Mar. Aquarist 6(6): 27-31. Tanaka, Y., Y. Shiobara, T. Ohyama, I. Ozaki & M. Moriya. 1985. Spawning behaviour, eggs and larvae of the hawkfish, Oxycirrhites typus, in an aquarium. Advance Abstracts for the 18th Annual Meeting of the Ichthyological Society of Japan, No. 58 (in Japanese). Thresher, R.E. 1984. Reproduction in reef fishes. TFH Publications, Neptune City. 339 pp.