reinhardti is more slender and has higher num- 

 bers of AO photophores, gill rakers, and fin rays. 

 There was, however, a great deal of overlap in 

 these features, and except for adult males 

 separation of the two species was the most 

 difficult of all species collected. (Indeed, com- 

 parison of Hawaiian specimens with H. proxi- 

 mum from equatorial waters, where H. rei)i- 

 hardti does not occur, indicated that the 

 specimens from Hawaii were more like H. 

 reinhardti in body proportions than those from 

 the equator.) 



The seeming overlap between these two 

 species may be anomalous in that the study 

 area is at the edge of the principal ranges of 

 both. According to Bekker (1965), H. proximum 

 occurs principally below lat. 20 °N and H. 

 reinhardti above lat. 20 °N in the central part 

 of the Pacific. Over most of their ranges these 

 two species do not appear to occur together in 

 abundance, and Hawaii appears to lie in the 

 transition zone between one species and the 

 other. 



Importance in the Ecosystem 



The estimated total numbers of individuals 

 per 103 ni2 for nine abundant species is given 

 in Table 5 for each of the quarterly series and 

 the CT series in March 1971. These numbers 

 are the calculated totals from the best series 

 (day or night) from each quarter. Using the 

 same weighting factors, the biomass per 10^ m- 



Table 5. — Estimated number (individuals/lO^ m'-) for 

 nine species of myctophids at different seasons. Numbers 

 are totals calculated (see text) from samples taken 

 throughout the water column by Isaacs-Kidd trawl (IK) 

 during September 1970, December 1970, March 1971, 

 and June 1971 and by Cobb trawl (CT) during March 

 1971. 



FISHERY BULLETIN: VOL. 71, NO. 2 



Table 6. — Estimated biomass (wet weight in g/lO^ m^) 

 for nine species of myctophids at different seasons. Figures 

 were calculated by the same method as those in Table 5. 



for each species was calculated and is given in 

 Table 6. 



The figures are probably low for several 

 reasons. Neither IK nor CT filtered with 100% 

 efficiency; Pearcy and Laurs (1966) estimated 

 that filtering efficiency was 85% for a 6-ft IK 

 with coarser mesh than used here. There is a 

 negative bias due to avoidance. Substantial 

 avoidance of either the IK or CT was demon- 

 strated for many species, and some likely 

 avoided both. This probably affected the biomass 

 estimates more since the larger fish are more 

 likely to avoid the net. (The CT figures for 

 Lampanyctus niger are deceptively low since 

 substantial numbers were caught by IK below 

 the deepest depth sampled by the CT.) For 

 March 1971, the estimated total number is 

 about 7% higher and that for biomass about 

 20% higher if CT estimates for Hygophum 

 reinhardti a.nd Lampanyctus )iobiUs are included 

 and the higher of the two estimates (IK or CT) 

 is used for the other species. It was not possible 

 to estimate reliably numbers and biomass for 

 the less abundant species. They amounted to 

 about 10% of the total numbers caught by the 

 four quarterly series and would probably in- 

 crease the estimates of total numbers and 

 biomass by a similar factor. 



The low totals in June reflect the fact that 

 for most species few adults were present and 

 the juveniles of the next generation were not 

 fully recruited to the population. Other dif- 

 ferences were largely due to changes in one 

 or two important species. The peak in total 

 numbers in December was due principally to 



430 



