506 



Fishery Bulletin 89(3), 1991 



Reproduction 



The sex ratios for both species were skewed slightly 

 toward females. The deviation from unity was not 

 highly significant, and the difference may be merely 

 a sampling artifact. Off the coast of east Africa, 

 Williams (1965) reported overall sex ratios significantly 

 biased toward males for both C. melampygus (M:F 

 1.68:1, N 78) and C. ignobilis (M:F 2.01:1, AT 323). Males 

 also predominated in an earlier series of collections 

 there (Williams 1956). Off Zanzibar Island (east Africa) 

 during the summer spawning season, he reported en- 

 countering shoals of "ripe and running" C. ignobilis 

 composed almost entirely of males. He also sampled 

 shoals where only females were caught and suggested 

 that the sexes segregate in the prespawning period. 

 Our catch data indicate that disproportionate numbers 

 of gravid females were taken from large schools dur- 

 ing summer. This behavior suggests that spawning by 

 C. ignobilis is not a random process, but is coordinated 

 in space and time. Whether or not large numbers of 

 males and females aggregate to spawn is unknown, but 

 seems likely. 



Based on the data in Figure 5, it appears that both 

 species spawn primarily in summer. However, the 

 paucity of quantitative gonad data (e.g., GSI's) makes 

 it difficult to present a complete picture of the annual 

 reproductive cycle for either species. A review of the 

 spawning patterns of Hawaiian fishes by Walsh (1987) 

 suggested that most Hawaiian reef fishes spawn 

 primarily during summer. 



Based on the condition of gonads examined in the 

 field, Williams (1956, 1965) concluded that spawning 

 of C. ignobilis off east Africa (1°S-10°S lat.) occurs 

 from July to March, peaking in November to March (the 

 austral summer). He reached no conclusions about 

 seasonality of spawning in C. melampygus, but he 

 documented ripe gonads throughout the austral spring 

 and summer (September to March). 



Based on visual underwater observations of spawn- 

 ing behavior between Cebu and Bohol in the Philippines 

 (10°N lat.), von Westernhagen (1974) concluded that 

 the main spawning season of C. ignobilis was December 

 and January, with a lesser peak in June. "Limited but 

 noticeable" behavioral spawning activity was reported 

 throughout the year for a group of carangid species 

 that included C. ignobilis, but it is not clear in what 

 months C. ignobilis specifically was observed spawn- 

 ing. This seasonal trend based on observed behavior 

 at low northern latitudes differs from the trend in- 

 dicated at higher latitudes by our examination of 

 gonads. Von Westernhagen also found that gonads of 

 dissected specimens were always only partly ripe, sug- 

 gesting serial spawning over a prolonged seasonal 

 period. 



In our study, female C. melampygus became sexual- 

 ly mature at about 350mm SL, which is about 39% of 

 L ro . The calculated age of a 350 mm C. melampygus is 

 close to 2 years. Female C. ignobilis mature at about 

 600mm SL, about 33% of L^, at an estimated age of 

 about 3V2 years. 



Williams (1965) estimated that the onset of sexual 

 maturity in C. melampygus in east Africa occurred at 

 about 30-40 cm (apparently SL). The basis of this 

 estimate is not clear, but it agrees well with our results 

 for the NWHI. In southern Africa, van der Elst (1981) 

 indicated that maturity in this species occurred at 

 40 cm FL (source of information not stated). From plots 

 of standard length and weight frequency of more than 

 330 C. ignobilis, Williams (1965) surmised that matur- 

 ity was reached between 54 and 61cm (apparently 

 SL), and between about 3 and 5 kg. Both this length 

 and weight estimate are consistent with our results. 

 Van der Elst (1981) stated that "sexual maturity coin- 

 cides with a fork length of 60 cm and an age of approx- 

 imately 3 years." 



Fecundity in our study was estimated only for 

 C. melampygus. The equation relating fecundity to 

 weight suggests a strongly nonlinear increase in fecun- 

 dity with increasing body weight. Caranx melampygus 

 appears to be a highly fecund species with a profligate 

 reproductive pattern common to many pelagic marine 

 fishes. 



Diet 



The diet of C. ignobilis contained a considerable diver- 

 sity of fish types and invertebrates (Table 3). The 

 perciform and tetraodontiform fishes were very impor- 

 tant, but eels, crustaceans, and cephalopods were also 

 significant in numbers and volume. The large number 

 of reef fishes suggests that C. ignobilis spends much 

 of its time foraging over shallow-water reef habitats, 

 but the presence of squid and the schooling carangid 

 Decapterus macarellus indicates exploitation of more 

 open-water habitats as well. Diet results of Williams 

 (1965) from east Africa also indicate feeding at all 

 levels in the water column and probably over a variety 

 of substrates. The presence in the diet of nocturnally 

 active species such as eels, lobsters, and octopus sug- 

 gests at least partly nocturnal feeding habits. Okamoto 

 and Kawamoto (1980) concluded that C. ignobilis 

 "appeared to be primarily a night feeder" in the 

 NWHI. Van der Elst (1981) characterized this species 

 in southern Africa as "more active during the day, 

 especially at dusk and dawn." 



C. melampygus appears to be more dependent on 

 diurnally active, shallow- water reef fishes (Table 4). 

 The labrids, mullids, pomacentrids, scarids, and mona- 

 canthids, which made up over 80% of its diet (percent 



