FROST and McCRONE: MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC 



ences in the vertical distribution of S. leucupsarus 

 (Figure 3A) indicate that the majority of individu- 

 als do perform a diel vertical migration. 



The occurrence of some S. leucopsarus in the 

 deepest samples at night (Figure 3A) indicates 

 that the entire population was not participating in 

 the vertical migration described above. Differ- 

 ences in migratory pattern appear to be largely a 

 function of size or age of individual fish. The 

 length-frequency histogram for our entire catch of 

 S. leucopsarus at Station P in 1973 (Figure 4A) 

 indicates that several size classes of fish were 

 sampled. Since S. leucopsarus metamorphoses to 

 the juvenile stage at 18 mm (Smoker and Pearcy 

 1970), the abundant 19-35 mm size class (Figure 

 4A) probably represented the youngest juvenile 

 fish. The largest specimens caught at Station P 

 attain the maximum size expected for S. leucop- 

 sarus, 85-111 mm (Kulikova 1957; Smoker and 

 Pearcy 1970). 



To determine the effect of size on vertical migra- 

 tion, we examined the three obvious size classes of 

 S. leucopsarus: 19-35 mm, 38-82 mm, and 90-112 

 mm. The smallest size class ( 19-35 mm) performed 

 a clear diel vertical migration from 275-330 m in 

 the daytime to 0-55 m at night (Figure 5A). The 

 anomalously low density of fish on the third day 

 must be attributed to horizontal patchiness offish. 

 Note especially that only on one night ( N3, Figure 

 5A) was one small-sized Stenobrachius captured 

 below 275 m. The medium size class (38-82 mm) 

 shows a similar migration (Figure 5B), though 

 these fish seemed to be more dispersed vertically, 

 both at night and in the daytime, than the small- 

 est size class. The high density of medium-sized 

 fish at 275-330 m on the first night, not apparent 

 on the other three nights, probably reflected the 

 fact that this sample was collected between 0554 

 and 0613 h, a time period when the migratory 

 sonic scatterers, and presumably myctophids, 

 were descending. Some of the medium-sized fish 

 probably had already descended into the 275-330 

 m layer at the time this sample was collected. The 

 largest size class of S. leucopsarus (90-112 mm) 

 had a pattern of vertical distribution totally dif- 

 ferent from those of the two smaller size classes. 

 Individuals of the largest size class were not cap- 

 tured at all in the first two daytime series and were 

 caught only in the two deepest samples in the third 

 daytime series ( Figure 5C ). They were captured in 

 all four night series, but never in the surface layer 

 (0-55 m), and in three of the four night series, the 

 greatest density of large-sized fish occurred be- 



tween 330 and 440 m. It is tempting to conclude 

 from these data that the individuals of the largest 

 size class also perform a diel vertical migration, 

 moving from daytime depths below our lower limit 

 of sampling (440 m) into our sampling range at 

 night. Of course, a similar vertical distribution 

 pattern could be obtained if the largest fish avoid 

 the trawl in the daytime, although it seems un- 

 likely that all fish of this size class could effec- 

 tively do so. Nevertheless, with the data from 

 Station P (1973), it is impossible to discriminate 

 between these two possibilities for the largest 

 fish. 



Stenobrachius leucopsarus had a very similar 

 pattern of distribution and diel vertical migration 

 at Station Q (Figure 3B). The length-frequency 

 distribution of the species was strongly skewed to 

 juvenile fish 1 19-31 mm), which made up 88. 79^ of 

 the total catch of the species. There was only one 

 relatively distinct secondary mode, consisting of 

 very large fish (81-108 mm), which composed 3.9% 

 of the total catch (Figure 4B). Fish in the smallest 

 mode and also the rarer intermediate sizes offish 

 (32-79 mm) were clear vertical migrants, closely 

 following the pattern described above for Station 

 P, and there was no difference in vertical distribu- 

 tion between the small- and medium-sized fish. 

 Also, as at Station P, representatives of the largest 

 size class offish were captured, with the exception 

 of 1 fish ( out of 45 caught ) in the deepest sample on 

 day 4, only in the night hauls and almost always 

 (43 out of 44 fish) below 50 m. 



At Station P in 1975, the same patterns of diel 

 vertical migration ( Figure 3C ) and size-dependent 

 variation in vertical distribution of S. leucopsarus 

 were evident, though far fewer fish were collected, 

 both because of the fewer vertical series taken and 

 an apparent decrease in abundance of the species 

 compared with the previous 2 yr (Table 3). This 

 decrease appears due partly to reduced abundance 

 of the smallest size class (17-32 mm) which made 

 up only 47.2% of the total catch in 1975 (Figure 

 4C ), compared with 62.6% at Station P in 1973 and 

 88.7% at Station Q. In the one deep daytime verti- 

 cal series (782-440 mm) at Station P, large S. 

 leucopsarus were captured between 440 and 740 m 

 (Table 4), thus supporting our earlier hypothesis 

 that the largest fish caught at night above 440 m 

 migrated in the daytime below our usual range of 

 sampling. However, extensive day and night sam- 

 pling over the entire vertical range of the large 

 fish is required to completely rule out daytime 

 avoidance of the trawl. 



757 



