HOBSON: FEEDING RELATIONSHIPS OF FISHES 



diet volume and ranking index: 41.7). There also 

 was much unidentified debris, including sand and 

 foraminiferans (mean percent of diet volume and 

 ranking index: 42.3) that all five apparently had 

 scraped from the substratum, and which probably 

 included substantial nourishment in the form of 

 organic detritus. The other components of the diet, 

 all minor, were: diatoms in all five (mean percent 

 of diet volume and ranking index: 3.3), sponges in 

 all five (mean percent of diet volume and ranking 

 index: 2.3), and harpacticoid copepods in one 

 (mean percent of diet volume: 0.3; ranking index: 

 0.06). 



CONCLUSION.— Cen^rop3'^e potteri is a 

 diurnal species that feeds on benthic algae and 

 probably on organic detritus. 



Forcipiger flavissimus Jordan and 

 McGregor — Ian wiliwili nukiinitkii oi oi 



This long-snouted species (Figure 25a), numer- 

 ous in Kona, and widespread throughout the 

 Indo-Pacific region, was long called F. longiros- 

 tris. Only recently has the distinction between F. 

 flavissimus and the true F. longirostris (Figure 

 25b and c; treated below) been recognized 

 (Wheeler, 1964; Randall and Caldwell, 1970). 

 Forcipiger flavissimus occurs singly, or, more 

 often, in groups of two or three. It is active 

 throughout the day, especially over coral-rich 

 reefs, where it picks at objects on a variety of 

 surfaces. At night it is alert close among rock and 

 coral cover but apparently inactive. 



Twenty-seven specimens (116: 94-137 mm) 

 were speared during day and night. Of 11 that 

 were taken either at night (later than 4 h after 

 sunset)orduringearly morning twilight, the stom- 

 achs of 9 were empty, and those of 2 (collected 

 between 4 and 5 h after sunset) contained only a 

 few well-digested fragments. In contrast, all 16 

 specimens taken at various times of the day had 

 full stomachs, including relatively fresh material, 

 as listed in Table 28. 



Most of the unidentified fragments among the 

 gut contents were relatively fresh pieces that this 

 fish apparently had recently torn from the bodies 

 of larger animals. The similarity of its elongated 

 snout and mouth to a pair of needle-nosed pliers 

 (Figure 26, lower) underscores the adaptiveness of 

 its feeding apparatus to this habit. Even the grip- 

 ping surfaces on the pliers are paralleled in the 

 snout of F. flavissimus by expanded contact- 



surfaces in both upper and lower jaws — both of 

 which carry multiple rows of short, inwardly curv- 

 ing teeth (Figure 27b). 



CONCLUSION. — Forcipiger flavissimus is a 

 diurnal predator that tears pieces off larger 

 benthic animals. 



Forcipiger longirostris (Broussonet) — 

 lau wiliwili nukunuku oi oi 



This species is relatively numerous in Kona, 

 although it appears to be rare elsewhere in 

 Hawaii. Both color varieties — the yellow form 

 (Figure 25b), which is essentially identical to F. 

 flavissimus, discussed above, and the dark brown 

 form (Figure 25c) — were observed regularly 

 throughout the study. Like F. flavissimus, F. lon- 

 girostris occurs typically over coral-rich reefs, and 

 the two species overlap extensively; however, in 

 areas where one is numerous, the other occurs less 

 frequently. Despite this, I was unable to relate 

 observed differences in relative numbers to 

 specific habitat differences. Forcipiger longiros- 

 tris is generally larger, but the most obvious mor- 

 phological distinction between the two lies in the 

 relative lengths of their snouts and in their differ- 

 ent mouth structures (Figure 26). Less noticeable, 

 but probably also related to feeding, F. longiros- 

 tris has relatively larger eyes. Like its congener, 

 F. longirostris is active on the reef by day, swim- 

 ming singly or in groups of two or three, and prob- 

 ing with its long snout in cracks and crevices. At 

 night it is close among cover of rocks or coral 

 — alert, but apparently inactive. 



Of the 26 specimens (136: 98-162 mm) col- 

 lected, all 4 that were speared at night (later than 

 4 h after sunset and before first light in the morn- 

 ing) had empty stomachs, whereas the stomachs of 

 all 22 collected at various times during the after- 

 noon were full (including relatively fresh items). 

 Decapod shrimps were the major prey, occurring 

 in all 22 individuals that contained food (mean 

 percent of diet volume and ranking index: 88.4). 

 Other food items were: pagurid crabs, without the 

 mollusk shell, in two individuals (mean percent of 

 diet volume: 1.9; ranking index: 0.17), fish frag- 

 ments in one (mean percent of diet volume: 0.5; 

 ranking index: 0.02), and crustacean fragments in 

 nine (mean percent of diet volume: 9.2; ranking 

 index: 3.76). 



In contrast to the omnivorous F. flavissimus, F. 

 longirostris has a restricted diet. It does not tear 



965 



