FISHERY BULLETIN: VOL. 71, NO. 2 



{Benthosema fibulatum, Diaphus adenomus, 

 D. chrnjsorhynchus, and Lainpade)ia urolampa). 

 Diogenichthys atlanticus and Diaphus fragilis 

 are more abundant in Equatorial Pacific waters 

 (Hartmann, 1971). The principal range of 

 Lampanyctus tenuiformis is not clearly known. 



Nineteen of the rarer species — Protomyc- 

 tophum heckeri, Myctophum spp., Loweina 

 spp., Centrobranchus spp., Diaphus bertelseni, 

 D. metopoclompus, D. "glandulifer," Lampadena 

 luminosa, L. ayiomala, Taaningichthys spp., 

 Bolinichthys supralateralis, and Notoscopelus 

 caudispi)iosus — appear to be typical of central 

 or equatorial-central waters but are nowhere 

 commonly collected (Bekker, 1966; Nafpaktitis, 

 1968; Nafpaktitis and Nafpaktitis, 1969; Gibbs 

 et al., 1971; Wisner, 1971; Davey, 1972). Five 

 of these species — Myctophum selenoides, Dia- 

 phus "glandulifer," Lampadena luminosa, 

 Taaningichthys minimus, and Bolinichthys 

 supralateralis — appear to spawn in the study 

 area. Lampadena urophaos, although designated 

 as a transitional water mass species by Paxton 

 (1967), would appear to belong with the above 

 group. It appears to spawn in this area, and a 

 rough comparison of the present study's and 

 Paxton 's catch per effort indicates that it is 

 more abundant near Hawaii. 



Avoidance 



Although it was not possible to set confidence 

 limits for the calculated totals and size-frequency 

 curves, in most species the differences between 

 day and night totals were not large enough to 

 suggest that there was substantially greater 

 avoidance at either time. The exceptions, 

 Lampanyctus steinbecki and the Hygophum 

 spp., were sampled better during the day than 

 at new moon at night. They also apparently 

 avoid the IK better at new moon than at full 

 and may be sampled equally well by night tows 

 at full moon and day tows. 



These results contrast with those of Pearcy 

 and Laurs (1966) off Oregon. They noted sub- 

 stantial differences in day-night avoidance 

 among mesopelagic fishes. The differences were 

 in the opposite direction of the few noted in this 

 study; most of their species avoided the net 

 better during the day. Pearcy and Laurs used 



a 6-ft IK, and perhaps the larger trawl used in 

 the present study sampled fishes relatively 

 better during the day. 



The IK-CT comparisons show that, as Har- 

 risson (1967) and others have suggested, no 

 single net will adequately sample all species. 

 The IK gave higher estimates of overall abun- 

 dance than the CT for many species analyzed, 

 but greatly underestimated the abundance of 

 two species, Symbolophonis evermainii and 

 Lampanyctus iiobilis. The CT sampled the 

 Hygophum species better at night during new 

 moon, but the IK probably samples them as 

 well or better during the day or during full 

 moon at night. Even allowing for escapement 

 of small fishes through the CT meshes, the CT 

 caught higher proportions of larger fishes of 

 some species, e.g., Ceratoscopelus warmingi 

 and Lampanyctus steinbecki. Neither net ef- 

 fectively sampled the mature sizes of Lam- 

 paiiyctus )iobilis, Lampadena lumi)iosa, and 

 Bolinichthys supralateralis. 



There was evidence for some species that 

 one trawl sampled better at some depths only. 

 The IK appeared to sample Hygophum rein- 

 hardti and Bolinichthys longipes as well or 

 better than the CT at the deeper end of their 

 night depth ranges, but the CT sampled as well 

 or better at the shallower end. For Diaphus 

 schmidti and Diaphus species A, the CT appeared 

 to underestimate abundance at the shallower 

 end of the depth range and caught relatively 

 more fish at the deeper end. Obviously, more 

 data are needed since one or two very high or 

 low catches could have produced these results. 



Harrisson (1967) using only roughly com- 

 parable pairs of samples, suggested that a 

 larger otter-type trawl towed at 1 m/sec caught 

 more larger fish than an IK towed at 1.5 m/sec. 

 Aron and Collard (1969), however, have shown 

 that towing speed had an important effect on 

 an IK's estimates of abundance and size com- 

 position for one of the more abundant species 

 collected in their study. Kuba (1970) found 

 that in the Equatorial Pacific, the IK at 2 m/sec 

 sampled most species' abundance and size 

 composition as well or better than the CT at 

 1.5 m/sec. It would appear that the higher 

 speed used for the IK tows in this study gave 



424 



