CLARKE: FEEDING HABITS OF STOMIATOII) FISHES 



is some indirect evidence for night feeding. 

 Most of the predators are found with their prey 

 both day and night; however, some prey, the 

 Bregmaceros and Melamphaes spp., occur well 

 below their predators during the day (Clarke 

 and Wagner 1976) and could only have been 

 eaten at night or during migration. The absence 

 of nonmigrating species in the diets also indi- 

 cates less or no feeding during the day. Finally, if 

 the lures of these predators are used to attract 

 fish which are themselves actively searching for 

 prey, the high frequency of species which feed in 

 the upper layers at night and the low frequency 

 of day-feeding, but vertically migrating, stomia- 

 toids also indicates night feeding by the preda- 

 tors. 



The average biomass of the nekton-eating 

 stomiatoids in the study area was a substantial 

 fraction of that of their prey (Table 9) and indi- 

 cates that they are probably an important source 

 of mortality to the prey species. An estimate of 

 the stomiatoids' impact on the prey populations 

 can be made from the catch and stomach content 

 data from the 1977-78 series of oblique trawl 

 tows and estimates of stomach evacuation time. 

 The entire series of tows caught 17,543 vertically 

 migrating planktivorous fish of the types eaten 

 by the stomiatoids: Myctophids, exclusive of the 

 Lampanyctus niger complex, plus other non- 

 stomiatoid planktivores. The same tows caught 

 822 nekton-eating stomiatoids, exclusive of Pho- 

 tostomias spp. A minimum of 111 fishes or fish 

 remains were found in the stomachs of these 

 predators. If the totals from this extensive series 

 of samples are taken as representative of the 

 average state in the study area, then on the aver- 

 age the nekton-eating stomiatoids consume 

 0.63% of the prey numbers over a period of time 

 equal to that required to evacuate the stomach. 



This estimate is likely to be low because both 

 feeding incidence of the predators and their 

 numbers relative to the prey are probably under- 

 estimated. There were 107 stomiatoids whose 

 stomachs were ruptured, and any prey they 

 might have contained are not included. In fact, it 

 is possible that individuals with distended stom- 

 achs were more susceptible to damage during 

 capture and consequently, that the incidence of 

 prey in the damaged specimens might have been 

 higher than in the undamaged ones. The num- 

 bers of both predators and prey caught by the 

 trawls are both negatively biased due to avoid- 

 ance of the net, but there is evidence that the 

 larger stomiatoids, especially Astronesthes spp., 



are much better avoiders than the small plankti- 

 vores (Clarke 1973, 1974). Unless there was a dif- 

 ference in bias between stomiatoids with full and 

 empty stomachs, this would also result in an 

 underestimate of the percent of the prey popula- 

 tion consumed. 



There are no available data to directly esti- 

 mate the time required to evacuate the stomach 

 for these fishes, but studies of other fishes fed 

 comparable sized meals indicate that evacuation 

 time is no longer than 4 d and probably less. 

 Evacuation times determined at temperatures 

 similar to those encountered by the stomiatoids 

 at night (15°-25°C) are mostly less than a day 

 (Pandian 1967; several studies summarized by 

 Magnuson 1969); at temperatures similar to those 

 of the day depths (4°-5°C) values are 2-4 d (Tyler 

 1970; Popova and Sytina 1977). If evacuation time 

 were 4 d, the annual consumption by stomiatoids 

 would be 57.5% of the average standing crop of 

 prey (0.63% X 365/4); if the time were 1 d, con- 

 sumption would be 2.3 times the standing crop. 



Although annual production by vertically mi- 

 grating planktivorous fishes probably exceeds 

 the average standing crop (Clarke 1973), the esti- 

 mated consumption by piscivorous stomiatoids 

 indicates that the latter account for a large and 

 possibly predominant share of the former's pro- 

 duction. Though stomiatoids appear to be the 

 most abundant piscivores, consumption of the 

 migrators by other, similar-sized predators, e.g., 

 scopelarchids, chiasmodontids, trichiuroids, 

 eels, and squids, is also likely to be substantial. 

 The migrating planktivorous fishes, in turn, 

 appear to be the dominant group of plankton con- 

 sumers in the tropical open ocean (Clarke 1973; 

 Maynard et al. 1975). Together, these indicate 

 that a large fraction of primary production is 

 eventually channeled into small predators, 

 smaller than the average planktivore in many 

 other parts of the ocean, rather than into large, 

 commercially harvestable species. 



ACKNOWLEDGMENTS 



I am indebted to many people who assisted in 

 collection and sorting of the samples from which 

 these fishes were taken and also to the crew of the 

 RV's Teritu, Moana Wave, and Kana Keoki. K. 

 Gopalakrishnan identified most of theeuphausi- 

 ids and decapods and capably taught me how to 

 identify the remainder. This research was sup- 

 ported by NSF GA-38423, NSF OCE 77-09202, 

 and the Hawaii Institute of Marine Biology. 



303 



