HOBSON: FEEDING RELATIONSHIPS OF FISHES 



capture prey during daylight. After gaining a posi- 

 tion close to its quarry, the attack is consum- 

 mated with a short dart forward, the victim being 

 literally sucked in with a sudden expansion of its 

 tubular snout. Hiatt and Strasburg (1960) did not 

 observe feeding, but speculated that this species in 

 the Marshall Islands probes with its long snout in 

 shallow holes and interstices of the reef and cap- 

 tures prey there by rapidly dilating its mouth. 

 They found a small atherinid fish in the gut on one 

 specimen. Sometimes trumpetfish accompany 

 schools of grazing surgeonfishes — usually mixed 

 groups of Acanthurus sandvicensis and A. ni- 

 groris, which frequently move across the reef. At 

 these times, small organisms probably are driven 

 out from algal cover by the grazing herbivores and 

 become available as prey to the trumpetfish. Occa- 

 sionally, the trumpetfish swims close beside large 

 herbivores, especially parrotfishes, apparently 

 using these large fishes as shields behind which to 

 get close to prey not threatened by the herbivore. 



Although 52 individuals (410: 220-621 mm) 

 were speared at different times of day and night, 

 no pattern was evident in the condition of the gut 

 contents from specimens taken at these different 

 times. Of 27 that contained food in their stomachs, 

 18 had captured fishes (mean percent of diet vol- 

 ume: 63; ranking index: 42), and 11 had taken 

 caridean shrimps (mean percent of diet volume: 

 37; ranking index: 15.07). 



It probably is significant that, with only two 

 exceptions, those sampled had preyed on either 

 fishes or shrimps — not both. The data cannot re- 

 late this selectivity to day or night activity or to 

 size of predator. The 16 individuals that had 

 preyed exclusively on fishes were within exactly 

 the same size range (241-528 mm) as the 9 indi- 

 viduals that had preyed exclusively on shrimps. 

 Furthermore, the mean sizes of the two groups 

 differed only slightly — 401 mm for the fish eaters, 

 396 mm for the shrimp eaters. The two individuals 

 that had taken both fishes and shrimps were 241 

 and 337 mm long. 



Aulostomus chinensis takes relatively large 

 prey: the 15 fish items (representing among others 

 Apogon snyderi, Acanthurus nigrofuscus, Can- 

 thigaster sp., and a labrid) that could be measured 

 accurately had a mean standard length of 58 mm 

 (range 25-88 mm) whereas the 11 shrimps that 

 could be measured accurately had a mean total 

 length of 55 mm (range 13-110 mm). Ten of the 11 

 shrimps were Saron marmoratus, the only shrimp 

 this large that was numerous in exposed locations 



on the reef. Although this shrimp occupies ex- 

 posed positions only at night, I have no evidence 

 that it is taken hy Aulostomus chinensis in greater 

 numbers after dark. 



Because the trumpetfish has an especially long, 

 attenuated body, and because it takes relatively 

 large prey, individuals that have recently in- 

 gested a meal often can be recognized by their 

 distended bellies. Such individuals were occasion- 

 ally seen during all periods of day and night, but 

 most often during, or shortly after, twilight. Con- 

 sistent with this, all three specimens that con- 

 tained fresh prey (little or no damage by digestion) 

 were collected during late twilight: in two of these 

 instances (one in the morning, one in the evening) 

 the prey were fishes; in the other instance (eve- 

 ning), the prey was a shrimp, S. marmoratus. 

 Beyond this, the gut contents were of little help in 

 establishing a pattern to feeding times; nor did the 

 incidence of individuals with empty stomachs in- 

 dicate a pattern, for they were collected during all 

 periods of day and night. 



CONCLUSION. — Aulostomus chinensis stalks 

 prey, mostly fishes and caridean shrimps, most 

 successfully during twilight, but also during the 

 day and perhaps also at night. 



General Remarks on Trumpetfishes 



The activity of Aulostomus chinensis in Kona 

 seems to be typical of the genus in other seas. 

 Randall (1967) reported only fishes and caridean 

 shrimps in 79 A. maculatus from the West Indies, 

 and also remarked on the large size of these prey, 

 as well as the way this trumpetfish sucks them 

 into its mouth by expanding its tubular snout. 

 Randall often observed A. maculatus hovering 

 vertically in the water over small fishes and sev- 

 eral times darting down on them (I did not see A. 

 chinensis feed this way). Collette and Talbot 

 (1972) judged A. maculatus in the Virgin Islands 

 to be primarily crepuscular. They were uncertain 

 about its nocturnal activity, but judged one they 

 saw in a gorgonian at 2330 h to be quiescent. 

 Eibl-Eibesfeldt (1955) described the way 

 trumpetfish in the Indian Ocean use other fishes 

 as cover behind which to approach small prey, and 

 this was also reported by Collette and Talbot 

 (1972) from the Virgin Islands. 



943 



