374 



Fishery Bulletin 103(2) 



late May where the surface temperature was between 

 17° and 19°C (Fig. 5). The larvae were abundant in the 

 middle to southern waters, where the surface tempera- 

 ture was between 18° and 20°C in late May (Fig. 6). 

 There was no significant difference in egg and larval 

 abundance in late May between the two years ( ANOVA: 

 F=0.03, P=0.87 for eggs; F=0.02, P=0.89 for larvae). 



Clupeid larvae 



Of the 107,252 larvae collected throughout the cruises, 

 clupeid larvae were most dominant, accounting for 57.2% 

 in number. Gizzard shad and Japanese sardine larvae 

 accounted for 76.4% and 23.6% of clupeid larvae, respec- 

 tively. A seasonal change in abundance of clupeid larvae 

 and a peak in abundance in late May in both years 

 were evident (Fig. 4C). Maximum abundance (no./m 2 ) 

 was more than 400 in late May in 1995 in the southern 



waters and there was no station where the abundance 

 exceeded 300/m 2 in 1996 (Fig. 7). The difference in 

 abundance of clupeid larvae in late May between the two 

 years was significant (ANOVA: F=8.12, P=0.005). 



Feeding 



Clupeid larvae (gizzard shad, Japanese sardine, and 

 unidentified clupeid larvae) were the most dominant 

 items in the stomachs of S. niphonius larvae (Table 1). 

 Feeding incidence (percentage of stomachs with food) 

 was significantly higher in 1995 than in 1996 (chi square 

 test; df=l, chi-square=8.538, P=0.0035). 



Growth 



Age of S. niphonius larvae collected in late May in 1995 

 and 1996 was estimated to be between 5 and 14 days 



Figure 3 



Contour plots of the surface water temperature (°C) of the Sea of Hiuchi during the three 

 cruises in 1995 and 1996. 



