376 



Fishery Bulletin 103(2) 



Figure 6 



Horizontal distribution of S. niphonius larvae in the Sea of Hiuchi in 1995 and 1996. 



that of clupeid fishes in the central Seto Inland Sea. 

 Piscivorous fishes tend to spawn earlier than other fishes 

 in freshwater ecosystems so that they attain sufficient 

 size to enable consumption of other young fishes by 

 the onset of piscivory (Keast, 1985). Because S. nipho- 

 nius larvae are piscivorous from the first feeding stage, 

 spawning that is synchronized with the seasonal peak 

 in abundance of clupeid larvae would be advantageous 

 for survival of S. niphonius larvae. 



Larvae of S. niphonius were abundant in the southern 

 part of the Sea of Hiuchi in late May 1995 and 1996 

 while eggs were abundant in the northwestern waters 

 during the same season. This difference in horizontal 

 distribution patterns of eggs and larvae seems to be as- 

 sociated with the drift by a residual flow (current) from 

 northern to southern waters. In the central part of the 

 Sea of Hiuchi, a residual flow in the middle (5-15 m) 

 layers proceeds southward at a speed of about 5 cm/s 

 (=4.32 km/d; Yanagi et al., 1995). Yolksac larvae of S. 

 niphonius are abundant in the 5- to 10-m layers in the 

 Sea of Hiuchi (Kishida, 1988) and do not exhibit diel 

 vertical migration (Shoji et al., 1999). The horizontal 

 distance between the stations with the highest egg and 

 larval abundance in late May was approximately 15 

 km in 1995 and 20 km in 1996. Given that the yolksac 

 stage is five days for mackerel larvae under 19°C (Shoji 

 et al., 2001), drift distance while larvae are entrained 

 in the southward residual flow during the yolksac stage 

 would be estimated to be approximately 20 km. The 

 estimate for the drift distance during the yolksac stage 



Table 1 



Feeding incidence (percentage of stomachs with prey) and 

 stomach contents of S. niphonius larvae collected in late 

 May of 1995 and 1996 in the Sea of Hiuchi. 



No. of larvae examined 

 No. of larvae feeding 

 Feeding incidence {.% ) 

 Size range (SL, mm) 



Stomach contents 



Sardinops melanostictus 



Konosirus punctatus 



Unidentified clupeids 



Engra ulis japon icus 



Unidentified Clupeiformes 



Mugiliidae 



Gobiidae 



Total 



1995 



1996 



approximates the horizontal distance between the sta- 

 tions of egg and larval highest abundance. It is there- 

 fore plausible that the larvae were transported by the 

 southward residual flow to the southern part of the Sea 



