CLARKE: FEEDING HABITS OF STOMIATOID FISHES 



undescribed or of uncertain status. Eustomias 

 bifilis was the only one for which large numbers 

 were available, and only 4 others were repre- 

 sented by more than 25 specimens (Table 6). The 

 remaining identifiable species were pooled 

 according to pectoral ray and photophore counts 

 along with specimens whose barbels had been 

 damaged and could not be identified to species. 

 Those designated "3, low" were all damaged spec- 

 imens with 3 pectoral rays and 15 or fewer VAL 

 and VAV photophores. Eustomias bifilis and E. 

 enbarbatus were the only other species from the 

 area with the same counts. Those designated "3, 

 hi" included 69 specimens of at least 6 unde- 

 scribed species and 65 damaged specimens, all 

 with 3 pectoral rays and over 15 VAL and VAV 

 photophores — the same counts as for E. longi- 

 barba and E. gibbsi. Those designated "2" in- 

 cluded 6 damaged specimens and 146 others of 

 about 20 species, which, like E. "silvescens," had 

 only two pectoral rays. The 2-rayed species have 

 shorter and generally more ornate barbels than 

 any of the 3-rayed species (cf. illustrations in 

 Morrow and Gibbs 1964). 



All prey items and remains from the 3-rayed 

 species with low counts were fish. Of 20 intact 

 items from E. bifilis, 11 were the myctophid 

 Bolinichthys loyigipes and 6 were myctophids of 

 at least 3 other genera {Benthosema, Diogenich- 

 thys, and Hygophum). One of the three unidenti- 

 fied items was definitely not a myctophid and 

 was probably a Howella sp. The range of relative 

 size of prey (15-34% of SL) was large, but there 

 was no trend with the size of the predator. One 

 and probably both of the intact fish found in E. 

 enbarbatus were Howella sp. The six intact items 

 from the damaged specimens (most of which 

 were probably the abundant E. bifilis) included 

 three Bolinichthys longipes, a Benthosema, an un- 

 identified myctophid, and an unidentified fish. 



The prey of E. longibarba, E. gibbsi, and the 

 other species with three pectoral rays and high 

 photophore counts were, with one exception, fish. 

 Of the 17 intact fish, 15 were myctophids includ- 

 ing 7 and probably 8 Bolinichthys longipes and at 

 least 2 other genera (Benthosema and Cerato- 

 scopelus). The median relative size of fish prey 

 for these Eustomias spp. (25%) was significantly 

 higher (P<0.05, Mann-Whitney test, one-tailed 

 probability) than that for the Eustomias spp. 

 with three rays and low photophore counts 

 (20.5%). One specimen of E. longibarba had eaten 

 a large euphausiid, Thysanopoda pectinata. 



One of the Eustomias spp. with two pectoral 



rays had eaten asergestid shrimp, but all other 

 prey of this group were fish. These Eustomias 

 spp. appeared to eat fewer and different mycto- 

 phids than did any of the 3-rayed species. Eu- 

 stomas "silvescens" (cf. fig. 106A in Morrow and 

 Gibbs 1964), the most commonly taken species of 

 this group, had eaten three Scopelosaurus spp., 

 three myctophids (two Bolinichthys longipes and 

 a Diaphus), and two unidentified fish. Stomachs 

 of the remaining species contained a total of 18 

 intact fish: 12 myctophids, 2 Howella sp., and 4 

 Scopelosaurus spp. (plus 2 more of the latter that 

 were too digested to measure). Five and probably 

 six of the myctophids were Diaphus spp., and 

 only three and probably four were B. longipes. In 

 the 3-rayed species of Eustomias, Diaphus was 

 found only once, and B. longipes was the most 

 common prey. Although data are too few to be 

 certain, some of the 2-rayed species appeared to 

 have diets that were restricted or included high 

 proportions of relatively rare fishes. For one un- 

 described form, all four items were Diaphus 

 spp.; for another, two out of four were Howella 

 sp.; and for a third and fourth, two out of two 

 items and two out of four remains, respectively, 

 were Scopelosaurus spp. The median relative 

 size of prey of the 2-rayed species (27%) was sig- 

 nificantly (P = 0.01) higher than that for the 3- 

 rayed species with low counts, but did not differ 

 from that for the 3-rayed species with high 

 counts. 



The two crustaceans recorded from Eustomias 

 spp. appear suspicious and indicative of postcap- 

 ture ingestion, especially since no digested crus- 

 tacean remains were found in any of the stom- 

 achs. The two items showed no obvious signs of 

 having been eaten after capture, but neither 

 were they much digested. The only indirect evi- 

 dence that these were actual prey items and not 

 eaten in the net is that I have found both crusta- 

 ceans and their remains in the stomachs of 

 several E. bulbornatus, a species which does not 

 occur in the study area. Since at least one species 

 of the genus appears to eat crustaceans, it is pos- 

 sible that others may do so occasionally. 



Based upon a limited amount of data (Table 6), 

 the remaining melanostomiatid genera, as well 

 as Stomias danae (Stomiatidae) and the Aristo- 

 stomias spp. (Malacosteidae), are piscivorous. 

 All the identifiable fish eaten by these species 

 were myctophids. All three items from Lepto- 

 stomias spp. were Notolychnus valdiviae. The 

 relative size of prey of the small Photonectes spp. 

 and several of the Bathophilus spp. was high— 



297 



