508 



Fishery Bulletin 89(3), 1991 



population was initially observed in visual censuses. 

 Census counts of P. spilosoma remained high to the 

 end of the study. This temporal pattern of predation 

 and prey abundance appears to offer an example of op- 

 portunistic feeding on sporadically abundant prey. The 

 much higher incidence of P. spilosoma in C. melam- 

 pygus appears to reflect the specific locations and 

 timing of sampling relative to the filefish population 

 surge. 



Potts (1980) analyzed the predatory behavior of 

 C. melampygus at Aldabra Atoll (Indian Ocean). His 

 observations indicated that this species sometimes 

 forages in small groups, but that it is most often found 

 singly or in pairs. Major (1978), who conducted an in- 

 vestigation of the predator-prey interactions of a cap- 

 tive group of juvenile C. ignobilis in Kaneohe Bay, 

 Hawaii, showed that there are advantages to group 

 hunting of schooling fish. Caranx species are known 

 to form interspecific feeding associations with large 

 labrids and other predatory species such as sharks, 

 barracuda, sting rays, and eagle rays (Potts 1980). 

 C. melampygus sometimes follow species that disturb 

 the substrate, causing small fishes and crustaceans to 

 be flushed out and exposed (Hobson 1974; pers. 

 observ.). Some of the small crustaceans and gastropods 

 found in the guts of C. melampygus may have been 

 taken in this way. 



Potts described C. melampygus as a diurnally active 

 predator that increased hunting activity at dawn and 

 dusk, when small midwater planktivorous fishes are 

 moving to or from the shelter of the reef. These obser- 

 vations are generally consistent with our analysis of 

 the feeding habits of this species in Hawaii. However, 

 nocturnal fishes and crustaceans occurred in some of 

 our samples, and planktivorous fishes did not dominate 

 the diet. 



The total incidence of benthic invertebrate prey in 

 our specimens was considerable: 43% (frequency) for 

 C. ignobilis and 22% for C. melampygus (including all 

 cephalopods). Caranx ignobilis is one of the few fishes 

 known to consume large, fully adult lobsters (Gooding 

 1985). Lobsters occurred in about 4% of the specimens, 

 composing over 10% of the diet by volume in our study 

 (Table 3). Okamoto and Kawamoto (1980) also reported 

 this interaction, and Gooding (1985) observed such 

 predation in the field. Lobsters are large, long-lived in- 

 vertebrates that appear to be situated high in the ben- 

 thic trophic system. This is one of several links that 

 suggests that these large, abundant Caranx species, 

 with their high rates of food consumption, are par- 

 ticularly important apex predators (Table 6). 



Large crustaceans occur rather widely in the diets 

 reported for C. ignobilis and C. melampygus (Williams 

 1956 and 1965, Hobson 1974, Okamoto and Kawamoto 

 1980, Parrish unpubl. data). In our study, the frequency 



of occurrence and numerical percent of crustaceans in 

 the diet were sizable, but the low volumes for groups 

 other than lobsters suggest that their dietary role is 

 quantitatively minor (Tables 3 and 4). Many crusta- 

 ceans could not be identified to low taxonomic level, 

 but the major groups found in these Caranx species ap- 

 pear to be among those most common in the diets of 

 a wide range of other carnivorous, demersal fish spe- 

 cies in the NWHI, e.g., shrimp, crabs, and stomatopods 

 (Parrish et al. 1985). All these major groups have also 

 been reported in other diet studies of these jacks. 



The higher incidence of crustaceans in smaller in- 

 dividuals of C. melampygus in our study was not ob- 

 viously related to characteristics of the habitat or social 

 structure. Fish of many sizes (larger and smaller than 

 350 mm SL) overlapped broadly among habitats, both 

 in our collections of specimens and in many field obser- 

 vations. Benthic sampling in the NWHI suggested that 

 large crustaceans were widely available in most of 

 these types of habitats. No direct evidence is available 

 to confirm the basis of this size difference in diet of 

 C. melampygus; it may reflect changing food prefer- 

 ence with age and the ability of larger individuals to 

 capture larger fish as prey. (All the crustaceans found 

 in C. melampygus specimens of all sizes could probably 

 have been captured easily by even the smallest of the 

 specimens.) 



In our study, C. ignobilis consumed both the benthic/ 

 demersal octopus and the more pelagic squid in quan- 

 tities (Table 3) that suggest that this trophic inter- 

 action is ecologically important to predator and prey 

 populations. Only two specimens of C. melampygus 

 contained cephalopods. There have been occasional 

 previous reports of cephalopods in these jack species 

 (Williams 1956, Okamoto and Kawamoto 1980, Ran- 

 dall 1980). 



There was a relatively low overlap of specific diet 

 items between C. ignobilis and C. melampygus (index 

 A yz = 0.42), indicating some separation of feeding 

 niches of these two sympatric congeners. Morpholog- 

 ically, they are very similar, but a mature C. ignobilis 

 is considerably larger than a mature C. melampygus. 

 This difference in body size probably accounts for the 

 occurrence or higher incidence of larger prey individ- 

 uals and some larger species (e.g., lobsters, octopus) 

 in C. ignobilis. 



Energy budget and 

 population consumption 



Preliminary energy budget calculations and estimates 

 of food consumption were made for jack populations 

 primarily for their ecological interest. Clearly, the esti- 

 mates of energy budget terms are crude and may be 

 of limited value from a physiological perspective. 



