FISHERY BULLETIN: VOL. 78, NO. 3 



tions, and other aspects of the ecology of the 

 species considered have been reported for the same 

 Study area (Clarke 1973; Clarke and Wagner 1976) 

 and other studies there summarized by Maynard 

 etal. (1975). 



Fishes were collected with a 3 m (lO-ft) Isaacs- 

 Kidd mid water trawl. To minimize the probability 

 of fishes' feeding while in the net, the terminal 

 section of the net was of ca. 3 mm knotless nylon 

 mesh instead of the commonly used, finer 

 plankton netting. 



The trawl was launched and towed at ca. 2 m/s 

 and the ship was slowed to ca. 1 m/s for retrieval. 

 Total time for descent to and ascent from towing 

 depth was 12-20 min. The trawl was towed at the 

 desired depth for ca. 2 h. Zooplankton were sam- 

 pled with 70 cm diameter, opening-closing bongo 

 nets of 505 /xm mesh. Ship's speed of ca. 1 m/s was 

 maintained for the entire tow; the nets were open 

 at the desired depth for 30-33 min. Time-depth 

 recorders attached to the nets indicated that the 

 depths of the "horizontal" (2 h) portions of the 

 trawl tows and the open part of the bongo net tows 

 were within 5 m of each other and of the desired 

 depth for each set of samples. All collections were 

 preserved in ca. 5% formaldehyde in seawater so- 

 lution immediately after the nets were on deck. 



Four different depths (70, 90, 110, 170 m) were 

 sampled (Table 1). In September 1973, two plank- 

 ton tows followed by two trawl tows were made on 



Table l. — Sampling information for trawl and plankton collec- 

 tions at four different depths off Oahu, Hawaii. D -(- R = total 

 time for descent and retrieval of trawl. 



the same night at each of three depths ( 70, 110, and 

 170 m). For the 170 m collections the bongo nets 

 failed to open and close properly on one of the tows. 

 A single trawl sample from 90 m was taken in 

 November 1974, and a single plankton sample 

 taken at the same depth two nights later. All tows 

 were taken between last light at dusk and first 

 light at dawn and within 2 d of new moon. Thus 



ambient light was essentially constant for all 

 samples taken at a given depth, and there were 

 probably no between-sample differences in verti- 

 cal distribution of either the fishes or their prey at 

 a given depth. Consequently, except for possible 

 captures in transit to and from towing depth (see 

 below), the fishes captured at a given depth were 

 assumed to have been feeding on the same prey 

 population sampled by the appropriate plankton 

 tows. 



Laboratory Procedures 



All nonlarval fishes from the trawls were iden- 

 tified and standard length (SL) measured to the 

 nearest millimeter. The fishes from each depth 

 were grouped by species and arbitrary size classes: 

 16-25 mm, 26-35 mm, 36-45 mm, 45-60 mm, and 

 >61 mm. Certain species or size classes from each 

 depth were eliminated from consideration be- 

 cause, based on previous evidence of depth-size 

 distributions (Clarke 1973; Clarke and Wagner 

 1976), they were almost certainly taken in transit 

 to and from towing depth. Among the size classes 

 that were considered, a few possibly, included 

 specimens that were captured above towing depth 

 and thus were not exposed to the same array of 

 prey as sampled by the plankton nets; these 

 groups are noted specifically in subsequent sec- 

 tions. 



For each specimen examined, standard length 

 was recorded and the stomach (anterior end of the 

 esophagus to the pyloric constriction) removed. 

 Prey items with bodies intact were noted sepa- 

 rately and measured to the nearest 0.1 mm with an 

 ocular micrometer. For the commonly occurring 

 crustacean prey, the following measurements 

 were used: copepods — prosome length, ostracods 

 — maximum carapace length, malacostracans — 

 the distance from the anteriormost point exclu- 

 sive of the antennae to the base of the telson. (The 

 telson of malacostracans was too frequently sepa- 

 rated to routinely include it in the length.) The 

 dimensions measured for other intact prey were 

 standard length for fishes, maximum diameter for 

 nearly spherical items such as gastropod veligers, 

 and total length for all others. Most intact 

 copepods and euphausiids could be indentified to 

 genus and most copepodite VI stages of the former 

 and juveniles and adults of the latter to species. 

 Ostracods were almost all Conchoecia spp., but 

 were not identified further. Other prey types were 

 identified only to major taxa. Identifiable frag- 



620 



