CLARKE and WAGNER: VERTICAL DISTRIBUTION OF MESOPELAGIC FISHES 



grate. Diplospinus mnlfistriatiis exhibited a 

 pattern opposite to that of P. crassiceps; the small 

 fish either remain in the upper layers or descend 

 only slightly during the day while the larger 

 juveniles (>ca. 60 mm) and adults undertake a 

 substantial migration. 



Avoidance 



With the exception of the gempylid-trichiurid 

 species, there were few obvious indications of 

 sampling error due to avoidance, but in most cases 

 data were too few to even discuss the subject. The 

 failure to capture mature specimens of the two 

 large Poroniitra spp. indicates avoidance by these 

 and probably a fraction of the populations of 

 other large melamphaids. Bregmaceros japonicus 

 was apparently undersampled during the day, and 

 the large B. macclellandi were sampled better by 

 the CT than by the IK. It is not unexpected that 

 avoidance was indicated for the larger, more 

 "solidly built" species rather than for small species 

 such as M. danae and the Scopeloheryx spp. or 



species such as Bafhi/h'pti(s lime and Euryphar- 

 ipix pek'canoiiles which do not appear "designed" 

 for swimming ability. The most puzzling indica- 

 tion of avoidance was that suggested for OpintJio- 

 proftus mleatus. This species not only has few 

 characteristics indicating swimming prowess, but 

 was undersampled at night rather than during the 

 day as one might expect if vision were involved. 



Sexual Dimorphism and Sex Ratio 



In several species, the males appeared to be 

 smaller than females (Table 2). The extreme case 

 was Bafhyleptus lisae where the largest female 

 was about 2.5 times longer than the largest male. 

 In Scopelogadus mizolepis, the females mature at 

 about the size of the largest males observed and 

 reach somewhat larger maximum size. A similar 

 trend is suggested by the data for Anoplogasfer 

 cornuta and two other large melamphaids, Po- 

 romitra crassiceps and P. oscitans, but the 

 numbers involved are too small to confirm it here. 

 In two smaller species of the same family, P. 



Table 2.-Summary of data on sex ratio and sexual differences in size for 10 species of fishes. 

 Under Population sex ratio and left hand column gives the number and size ranges of all males 

 in the population with 95"? confidence limits (read to the nearest 0.01 from Chart 3 in Tate and 

 Clelland 19.57). Sex ratio was considered significantly different from 1:1 if these limits did not 

 cross 0.50. Under Size Difference similar figures are given for only those specimens larger than 

 the smallest mature female (since all Melamphaex danae were as large or larger than the 

 smallest mature female, the data are the same for both pairs of columns). For Bathylepiut^ linae, 

 where the smallest mature female was much larger than the largest male, and for three 

 melamphaids, where no mature females were taken, we have given only the number and size 

 range of females larger than the largest male. 



643 



