HOBSON: FEEDING RELATIONSHIPS OF FISHES 



Table 13. — Food of Holocentrus diadema. 



Twenty-eight specimens ( 109: 85-127 mm) were 

 speared as they swam in exposed locations on the 

 reef at various times during the night. Only two 

 had empty stomachs: in one of these, taken shortly 

 after nightfall, the entire gut was empty, which 

 indicated it had not as yet hunted successfully 

 that night; the other, taken with an empty 

 stomach just before daybreak, had a full intestine, 

 suggesting that it had fed early but not late during 

 the night. The other 26 specimens all contained 

 food in varying stages of digestion, most of it 

 identifiable, as listed in Table 13. 



CONCLUSION.— //o/oce/?^ri/s diadema is a 

 nocturnal predator that feeds mostly on benthic 

 crustaceans, although it also takes free- 

 swimming forms close to the bottom. 



Holocentrus lacteoguttatiim Cuvier 



This small squirrelfish is similar toH. xanthery- 

 thrus and H. diadema, but frequents shallower 

 water than the other two, being most numerous 

 during the day in rocky crevices along surge- 

 swept shores, often where the water is only 1 to 4 

 m deep. It aggregates in these crevices, and after 

 nightfall ranges out over coral, rock, or pockets of 

 sand on the surrounding reef. Gosline and Brock 

 (1960) also noted the shallowwater habits of this 

 species, but in at least some situations it occurs in 



depths below 30 m (Gosline, 1965). These habitat 

 distinctions are clearest in daylight, when the 

 three species have retired to their shelters. The 

 differences are less clear at night, when their ac- 

 tivity ranges overlap. Holocentrus lacteoguttatum 

 does not seem to have prominent nocturnal color 

 features, as do certain other species of 

 Holocentrus, treated above; however, several in- 

 dividuals after having been speared at night 

 showed faint traces of essentially the same white 

 markings characteristic of nocturnally active in- 

 dividuals of H. xantherythrus, H. diadema, and 

 H. tiere (see Figure 12b). 



Twenty-one specimens (88: 52-104 mm) were 

 collected at various times of day and night. All but 

 1 of 13 active individuals that were speared in the 

 open at night (more than 4 h after sunset and 

 before they had returned to shelter at daybreak) 

 had food in their stomachs; the lone exception, 

 collected 4 h after sunset, had a completely empty 

 gut, indicating it had not yet hunted successfully 

 that night. In comparison, only one of five col- 

 lected from aggregations under shelter during 

 midmorning had material in its stomach, and this 

 was extensively digested (all had full intestines, 

 however). Finally, all three that were collected 

 from aggregations under shelter during late af- 

 ternoon had completely empty guts, except for a 

 few well-digested fragments posteriorly. Items in 

 the 13 specimens containing identifiable material 

 are listed in Table 14. 



937 



