DIETS OF FOURTEEN SPECIES OF VERTICALLY 

 MIGRATING MESOPELAGIC FISHES IN HAWAIIAN WATERS 



Thomas A. Clarke ^ 



ABSTRACT 



Fishes and zooplankton were sampled at four depths (70, 90, 110, 170 m) at night in the upper layers 

 near Hawaii. Diets of the fishes were determined from stomach contents and preferences estimated by 

 comparison with prey densities at the appropriate depth. Generally, the fishes fed on relatively large, 

 pigmented or opaque crustaceans; other taxa and very small or translucent prey were rarely eaten. 

 There were, however, differences in diet and preference between species; these were frequently corre- 

 lated with morphological features, especially lens size and gill raker spacing. One group of four fishes 

 which were very similar in both diet and morphology were separated by depth distribution and size. 

 Comparison with other studies indicates that tropical species are perhaps more specialized and 

 ecologically separated in diet than their counterparts in high latitudes. 



Vertically migrating mesopelagic fishes are im- 

 portant components of oceanic ecosytems. In the 

 tropical open ocean, abundance of larvae 

 (Ahlstrom 1969) and estimates of biomass (Clarke 

 1973; Maynard et al. 1975) indicate that they are 

 the dominant group of micronekton and greatly 

 exceed the abundance of epipelagic forms. Stand- 

 ing crops are even higher in oceanic situations at 

 higher latitudes (Frost and McCrone 1979) and 

 coastal upwelling areas (Pearcy and Laurs 1966). 

 Tropical oceanic faunas are much more diverse. At 

 high latitudes and in quasi-neritic situations, one 

 to three species typically make up the great major- 

 ity of the standing crop (Pearcy and Laurs 1966; 

 Zahuranec and Pugh 1971; Baird et al. 1975; Frost 

 and McCrone 1979), while in the tropical open 

 ocean the abundances of the dozens of cooccurring 

 species are more evenly distributed (Clarke 1973, 

 1974). 



The diets of these fishes are of interest both to 

 assess their impact on lower trophic levels in 

 oceanic ecosystems and to determine the degree to 

 which cooccurring species are specialized with re- 

 spect to their feeding habits; however, previous 

 studies do not allow serious consideration of these 

 aspects. Few have presented extensive data on 

 more than one to three species. For the most part, 

 prey have not been identified adequately enough 

 to seriously discuss preference or dietary overlap, 

 and there has been no consideration of bias due to 



'University of Hawaii, Department of Oceanography and 

 Hawaii Institute of Marine Biology, PO. Box 1346, Kaneohe, HI 

 96744. 



Manuscript accepted January 1980. 

 FISHERY BULLETIN: VOL! 78, NO. 3, 1980. 



differing rates of digestibility and, therefore, abil- 

 ity to identify different prey types (Gannon 1976). 

 Few studies have compared stomach contents of 

 fishes with appropriate samples of the prey avail- 

 able; those that have done so have simply com- 

 pared percentages of different prey types and have 

 not considered biases or errors inherent in the 

 samples taken for prey abundance. 



This paper considers diets of 14 species of verti- 

 cally migrating mesopelagic fishes based on data 

 from collections taken near Hawaii in the central 

 North Pacific Ocean. All species are primarily zoo- 

 planktivorous and are known (Clarke 1978) or 

 suspected to feed principally in the upper 250 m at 

 night. The diets of each species are compared with 

 densities of zooplankton at each of the depths 

 sampled. While problems in feeding studies men- 

 tioned above have by no means been completely 

 eliminated, the methodology recognizes and at 

 least qualitatively attempts to account for major 

 sources of error. The results allow consideration of 

 biases of the fishes as "samplers" of the potentially 

 available prey and of dietary overlap between 

 species or sizes cooccurring at the same depths in 

 the water column. 



METHODS 



Field Collections 



All specimens for this study were collected ca. 20 

 km off the coast of Oahu, Hawaii, (ca. lat. 21°10'- 

 30 ' N, long. 158°10 '-30 ' W) over bottom depths of 

 2,000-4,000 m. The depth ranges, vertical migra- 



619 ^^ 



