CLARKE: FEEDINC HABITS OF STOMIATOII) FISHES 



up to 10-12 assorted copepods and ostracods, all 

 apparently recently ingested — were found in 

 several stomachs of fishes that otherwise had 

 eaten only relatively large items. I also found 

 several apparently freshly ingested euphausiids 

 and sergestid shrimps in stomachs of fishes that 

 otherwise appeared to be strictly piscivorous. 

 Specimens from these slow tows were not in- 

 cluded in the data presented here.) 



Estimates of biomass and relative abundance 

 of vertically migrating fishes in the study area 

 and of feeding incidence of the nekton-eating 

 stomiatoids were made from catches of a series of 

 oblique 3 m Isaacs-Kidd trawl tows taken at 

 approximately monthly intervals between Au- 

 gust 1977 and November 1978. A time-depth re- 

 corder and a flowmeter were mounted on the 

 trawl for all tows. All fishes were identified, spe- 

 cies were grouped by taxa and known or prob- 

 able feeding habits, and wet weights of each 

 group determined for each sample. All nekton- 

 eating stomiatoids from the series were exam- 

 ined and are included in the results below. 



The 58 night tows in this series fished between 

 the surface and ca. 350 m and covered the night- 

 time depth range of all vertically migrating spe- 

 cies. The relative abundances of the different 

 taxonomic and trophic groups were calculated 

 based upon total numbers from all the night 

 tows. Biomass (wet weight) per unit area of each 

 group was calculated as in Maynard et al. (1975) 

 for each sample. The overall mean of all samples 

 and all seasons was used as the estimated aver- 

 age biomass. The 28 day tows covered the day 

 depth ranges of the vertically migrating species 

 (ca. 350-1,000 m), but for various reasons it was 

 not possible to reliably estimate volume filtered 

 (and therefore biomass per unit area) from these 

 tows. The numbers of nekton-eating stomiatoids 

 and of prey species in the catch and the numbers 

 of prey found in the stomachs of the stomiatoids 

 from both day and night tows were used to esti- 

 mate feeding incidence relative to the numerical 

 standing crop of prey species. 



RESULTS 



Photichthyidae 



Vinciguerria nimbaria (Table 3) from three 

 samples within its day depth range were divided 

 into two size groups (16-25 mm and over 25 mm 

 SL). Catches of three of the six size-depth groups 

 were high, and only fish with visually apparent 



full stomachs were selected for examination 

 (Table 3, columns 3, 5, 6). All fish of the other 

 groups were examined, but total numbers of in- 

 tact prey were still quite low for these. 



Overall the most frequent items in the stom- 

 achs were small copepods and ostracods. Oncaea 

 spp. were the dominant prey in most size-depth 

 groups, and in all samples Oncaea— especially 

 the small forms — were more frequent in the diets 

 of the smaller fish than in those of the large. Be- 

 yond this, however, the diet composition varied 

 between groups without much apparent relation 

 to size or depth, e.g., Clausocalanus spp. and 

 Pleuromamma gracilis were important fractions 

 of the prey of the small fish from 400 m and both 

 size groups from 450 to 500 m; candaciids and 

 Scolecithrix danae were taken by most groups, 

 but decidedly more frequently (as percentage of 

 prey items) by the large fish from 400 m; the fre- 

 quency of ostracods varied among the groups 

 from 3% to 42% of the total items. 



Some of this variability was undoubtedly a 

 consequence of small sample sizes from three 

 groups, but part resulted from large numbers of 

 certain prey types occurring in only one or a few 

 of the fish from a given size-depth group. Exam- 

 ples include (see appropriate column of Table 3): 

 All 7 P. gracilis from 1 of 6 fish with prey (col- 

 umn 1); all 5 amphipods from 1 of 3 fish (column 

 2); 4 of 5 Undinula spp. from 1, 4 of 5 Sapph irina 

 spp. from another, and all 11 pelecypod larvae 

 from another out of 18 fish (column 3); 13 of 41 

 Clausocalanus spp. in 1 and 34 of 73 P. gracilis in 

 3 others out of 20 fish (column 5); 14 of 15 Scole- 

 cithrix danae in 1 and 24 of 34 P. gracilis in 2 

 others out of 9 fish (column 6). The presence or 

 absence of only one or two fish such as these had 

 an important effect on percentages of certain 

 items in the estimated diet of a size-depth group. 



Vinciguerria nimbaria over 30 mm SL had 

 eaten considerably larger items more frequently 

 than smaller fish. The only large day-caught 

 specimen (37 mm) contained remains of another 

 fish, a 3.0 mm amphipod and two Nematobrach- 

 ion spp., each ca. 15 mm long. Ten specimens 

 over 30 mm were taken in a night tow at 70 m. 

 Most items in the stomachs of these fish were on 

 the borderline between "intact" and "remains" 

 and difficult to count similarly to those from the 

 day specimens, but it was clear that euphausi- 

 ids— mostly Stylocheiron spp.— were the most 

 frequent items and that small copepods were 

 much less important than in the smaller fish. Re- 

 mains of six to nine Stylocheiron each were found 



291 



