RICHARDSON andPEARCY: COASTAL AND OCEANIC FISH LARVAE 



m) during a 24-h period in May 1972. Essentially, 

 the entire water column was sampled. The volume 

 of water filtered by each type of tow was about the 

 same and the number of day and night tows in 

 each stratum was equal. Because the nets had no 

 opening-closing device, samples from all but the 0- 

 to 10-m stratum were contaminated with catches 

 from overlying waters. However, the maximum 

 tow time spent outside the desired stratum was 

 20% for the deepest tows and was usually <10% 

 for the intermediate depths. Therefore, no cor- 

 rection factor was applied to the data. 



The greatest number of larvae and taxa was 

 taken near the surface both day and night (Table 

 4). The 51- to 100-m stratum yielded the fewest 

 larvae and taxa while the 11- to 50- and 101- to 

 150-m strata were intermediate. More larvae were 

 taken at night, primarily in the 0- to 10-m stratum 

 where avoidance during the day would be expected 

 to be greatest. Mean larval length in this stratum 

 was much greater at night which also indicated 

 daytime avoidance by large larvae in surface 

 waters. Mean larval length was also high in the 

 101- to 150-m stratum day and night, primarily 

 because of the abundance of large osmerids there. 



Of the 22 taxa taken, those represented by more 

 than 10 larvae were examined for trends in dis- 

 tribution (Table 4). Clupea harengus pallasi 

 (25-31 mm, x 28), Ammodytes hexapterus (17-37 

 mm,x33), and Ronquilus jordani (6-21 mm,f 13) 

 were concentrated in the upper 10 m at night and 

 were completely absent in daytime collections 

 from all depths. They exhibited strong daytime 

 avoidance, indicated by night/day ratios. Large 

 Sebastes spp. larvae (9-11 mm, x 10) were only 

 taken at night and perhaps avoided by day, 

 whereas small larvae (3-4 mm, x 4) were taken 

 both day and night in the upper two strata. 

 Stenobrachius leucopsarus (5-11 mm, x 8) and 



Isopsetta isolepis (14-23 mm, x 20) occurred 

 predominantly in the upper two strata but showed 

 no evidence of daytime avoidance. Mean larval 

 lengths were about the same by day and night. 



Of the remaining taxa, Radulinus asprellus 

 (9-15 mm, x 12) appeared to occur throughout the 

 water column in similar numbers and lengths 

 during both day and night. Cyclopteridae spp. 1 

 (4-8 mm,* 5) occurred mainly near the surface in 

 daytime but only in the 51- to 100-m stratum at 

 night, possibly a result of patchiness or con- 

 tamination of the deeper hauled net in the surface 

 stratum. Only osmerids occurred primarily near 

 the bottom (101-150 m), by day and night. Some 

 were taken near the surface at night which may 

 indicate vertical migration by some individuals or 

 avoidance by day. Preliminary examination of 

 specimens did not reveal the surface- and bottom- 

 occurring osrrierid larvae to be different species. 

 Mean lengths for deep- and surface-caught os- 

 merids were about the same, 21 and 23 mm. 



ASSEMBLAGES 



Two separate assemblages of fish larvae were 

 distinguished, using a similarity coefficient ma- 

 trix based on Sander's (1960) dominance-affinity 

 index (J lowest percent of all larvae in common 

 between two stations). In 1971 a coastal as- 

 semblage occurred at stations 2 to 28 km offshore, 

 which was distinct from another assemblage 

 occurring at stations farther offshore (Figure 4). A 

 similar pattern was found in 1972 during the 6 mo 

 for which data were available. In 1971, the mean 

 affinity value among stations 2, 6, 9, 18, and 28 

 was 65.81 and among stations 46, 56, 65, 74, 93, 

 and 111 it was 60.61. In 1972, the mean affinity 

 values for these same sets of stations were 43.21 

 and 56.61, respectively. Sebastes spp. were 



TABLE 4. — Number/l,000m 3 , number of taxa, and mean length offish larvae by day, night, and depth strata taken 

 during a 24-h period 18 km off the mid-Oregon coast (lat. 45°04.0'N) in May 1972. N/D = night to day ratio. Each 

 number is the sum of four replicate samples. 



131 



