FISHERY BULLETIN: VOL. 74, NO. 4 



One important systematic character, the third 

 maxilliped, appears directly related to feeding. 

 Many species of Sergestes have greatly enlarged 

 third maxillipeds, armed with stout spines and 

 varying in length and development among the 

 different species. Sergestes pectinatus in par- 

 ticular has highly modified third maxillipeds, with 

 a series of short, comblike setae between the 

 longer spines. The division of Hawaiian sergestids 

 into a long-maxilliped group and a short-maxil- 

 liped group would seem logically to indicate a 

 difference in diet between the two groups. 



The results of the DSB III study (Table 5) are 

 rather unexpected. All the species captured fed 

 largely on zooplankton-sized Crustacea in the 1- to 

 3-mm size range, chiefly calanoid copepods, 

 myodocopid ostracods, and hyperiid amphipods. 

 Some species also ate smaller zooplankton in the 

 0.4- to 0.6-mm size range, chiefly larval bivalves, 

 foraminifera, and cyclopoid copepods. Ability to 

 utilize prey in the small size range appeared to 

 depend not on the length of the third maxillipeds 

 but on the degree of setation of the first three 

 pairs of pereiopods and (when not enlarged) the 

 third maxillipeds. Species feeding on small zoo- 

 plankton all have long setae spaced about 0.3-0.4 

 mm apart. All well-sampled species in the short- 

 maxilliped group except Sergestes erectus fed on 

 the small zooplankton. Within the long-maxilliped 

 group, there is a gradation in degree of setation of 

 the pereiopods from S. annatus, which has very 

 short, sparse setae, through the S. orientalis 

 group, which have somewhat longer, more numer- 

 ous, but still rather sparse setae, to S. sargassi and 

 S. 'pectinatus, which have rather long setae spaced 

 about 0.5-0.6 mm apart. Of this group only S. 

 armatus was captured in quantity during DSB 

 III; its diet definitely lacked small zooplankton. A 

 few specimens of 5. sargassi and S. pectinatus 

 were captured; none contained small zooplankton, 

 but with the small sample size their status remains 

 in doubt. 



The dietary specializations of Hawaiian serges- 

 tids thus appear more related to size than to type 

 of prey. The variety of copepods, amphipods, and 

 ostracods that compose the large zooplankton 

 fraction all seem to be equally acceptable as prey. 

 The various modifications of the third maxilliped 

 may reflect specialized methods of capturing prey 

 rather than a specialized diet. In particular, the 

 diet of S. pectinatus lacks any distinctive charac- 

 teristics which can be associated with its unusual 

 maxillipeds. While large zooplankters are prob- 



ably seized individually, the small zooplankton 

 appear to be sieved from the water onto the long 

 setae, spaced so as to retain zooplankton and pass 

 water, a process akin to filter feeding. 



The small zooplankton probably represent a 

 supplementary rather than a primary resource for 

 sergestids. Larval bivalves, as meroplankton, are 

 unlikely to be abundant all year around (they were 

 abundant in the zooplankton during the December 

 1973 cruise; I have not examined other zooplankton 

 samples) and are unlikely to be abundant far from 

 land. Many of the individuals containing small 

 zooplankton also contain masses of an unidentified 

 greenish, fibrous material. Judkins and Fleminger 

 (1972) reported similar material in Sergestes 

 similis, and Foxton and Roe (1974) reported 

 similar material in a number of Atlantic species. If 

 this material is detritus and not the digested 

 remains of some unidentified organism, it would 

 represent another resource available to sieving 

 sergestids, potentially very important when small 

 zooplankton is sparse. The inefficient-looking 

 sieving mechanism of the Hawaiian sergestids are 

 a reminder that none of these species feed solely, 

 or even primarily, on the small zooplankton. Any 

 modifications for increased sieving ability must 

 not hamper the animal's ability to seize large 

 zooplankton. 



The results from DSB III are quite different 

 from those reported by Donaldson (1975). He 

 found a much larger proportion of large prey, such 

 as euphausiids and fishes, and also many more 

 chaetognaths. Part of the difference is due to his 

 much larger sample, where infrequently eaten 

 prey are more likely to turn up. Large sergestids 

 captured in very small numbers during the DSB 

 III cruise, particularly Sergia bisulcata, S. in- 

 equalis, and perhaps 5. tenuiremis, are likely to 

 eat larger prey than is reported here. However, 

 some of the difference between Donaldson's re- 

 sults and the DSB III results may be due to a 

 higher degree of feeding in the trawl during 

 Donaldson's study. The abnormal conditions in the 

 cod end of a mid-water trawl are apt to lead to 

 abnormal feeding. Judkins and Fleminger found a 

 much lower proportion of euphausiids in the 

 stomachs of sergestids caught by albacore than in 

 trawl-caught shrimp. They also found fish scales 

 only in trawl-caught shrimp, an unlikely food item 

 under natural conditions. These results emphasize 

 the need for future feeding studies to take what- 

 ever steps are necessary to minimize or eliminate 

 feeding in the trawl. 



830 



