CLARKE: DIEL FEEDING PATTERNS OF MESOPELAGIC FISHES 



or esophagus. In all of these specimens, the 

 stomach also contained food and was not everted; 

 the remains from the esophagus or mouth were 

 picked out as carefully as possible and added to the 

 stomach contents. (No specimens with everted 

 stomachs were included.) Use of these 19 speci- 

 mens had the desirable effect of increasing the 

 numbers upon which statistical estimates were 

 based, but, as pointed out earlier, possibly biased 

 the data because some of the regurgitated food 

 may not have been recovered. The data were 

 treated with and without these 19 specimens. In- 

 clusion of the latter either had no effect or in- 

 creased the significance of differences indicated 

 without them. Thus bias, if any, that was intro- 

 duced was insufficient to obscure between-period 

 differences in stomach fullness. 



Diel differences in stomach fullness were sig- 

 nificant {P<0.05) and resembled the trends of 

 other myctophids. There was, however, no clear 

 indication of a peak value at dawn; no specimens 

 from SR were available and the medians for N3, 

 N4, and Dl were similar to each other. The per- 

 centage of empty stomachs was low throughout 

 most of the night and increased steadily between 

 Dl and SS. 



The size-frequency distribution of the speci- 

 mens was bimodal; 45% were <40 mm SL and 47% 

 >60 mm. The small specimens had eaten mostly 

 copepods and amphipods 1-3 mm and Euphausia 

 spp. <10 mm, while the large ones had taken large 

 copepods ( >3 mm PL) and euphausiids, mysids, 

 sergestiids, and fishes 10-30 mm long. Intact prey 

 were found frequently in night specimens and oc- 

 casionally in those caught by day. The latter were, 

 with the exception of a single Lophothrix humili- 

 frons (apparently a deep-living copepod), migrat- 

 ing species that could have been taken at night. 

 One specimen from N4 contained; among the re- 

 mains of a euphausiid, crab megalopa, and 

 copepods; a partially digested insect (probably a 

 hymenopteran). 



Lampanyctus steinbecki (Figure 2) 



Stomach fullness values for L. steinbecki dif- 

 fered significantly (P<0.005) over the diel cycle. 

 The medians generally increased from SS to SR 

 and thereafter stayed at about 1% until a sharp 

 decrease between D3 and Nl. The percentage of 

 fish with empty stomachs was low for all periods. 

 The principal prey ofL. steinbecki were copepods > 

 ca. 2 mm PL — mostly aetideids, Pleuromamma, 



and Candacia — and euphausiids. A few intact 

 items were found in specimens from Dl and D2 but 

 all were shallow-living or migrating species that 

 could have been taken the previous night. With 

 the exception of a s\r\g\ePareuchaeta sp. (probably 

 a deep-living nonmigrator), the prey from D3 and 

 SS were all well digested. 



Notolychnus valdiviae (Figure 2) 



The //-test indicated highly significant 

 (P<0.005) diel differences in stomach fullness for 

 N . valdiviae. Median values were low early in the 

 night and increased to a peak at N4. The minimum 

 value at Dl was slightly below the early night 

 values. Stomach fullness increased slightly until 

 SS and then decreased at SS-N 1 . The percentage of 

 fish with empty stomachs was low at all periods. 

 The positions of the 75th percentiles indicated 

 higher percentages of fish with relatively full 

 stomachs at N3, N4, SS, and D3. 



Notolychnus valdiviae had taken a wide variety 

 of sizes (ca. 0.5-4.0 mm PL) and species of 

 copepods, but the bulk of the food in terms of 

 weight was made up by large (relative to the 

 weight of N. valdiviae) items such as 

 Pleuromamma xiphias, Candacia longimana, and 

 2-4 mm aetideids. Intact prey were more fre- 

 quently noted in specimens from N3 and N4 than 

 in those from the apparent "secondary peak" in 

 stomach fullness at D3 and SS. Considering only 

 those specimens with stomach fullness >2% 

 (whose numbers distinguish the peak periods from 

 others), only three of the nine from D3 and SS 

 contained intact or partially intact Pleuromam- 

 ma. The other six contained remains that were 

 either unrecognizable or barely so. In contrast, of 

 the 15 specimens from N3 and N4, 12 contained 

 1-3 intact items, while only 3 contained unrecog- 

 nizable remains. This plus the absence of any ap- 

 parent significant differences associated with the 

 D3/SS peak indicate that the latter was due to a 

 chance collection of a few more specimens that had 

 taken large meals the previous night rather than 

 to extensive daytime feeding. 



Triphoturus nigrescens (Figure 2) 



Overall diel differences in stomach fullness 

 were highly significant (P<0.005). Both medians 

 and means rose from low values at SS and Nl to a 

 peak at N4 and then, except for a slight increase at 

 D3, declined until SS. Due to the broad overlap in 



503 



