906 
Fishery Bulletin 96(4), 1 998 
smaller than the sizes in 1991. The difference in 
growth rates between 1990 and 1991 could not be 
explained by SST because in our study area, slower 
growth was recorded in 1990, when SST was higher, 
than in 1991. Relative abundance of sardine shirasu 
in the coastal waters of central Japan in 1991 de- 
creased to about 50% of the 1990 (Kishida et al., 
1994). The catch of sardine larvae in 1991 declined 
to 41% of that in 1990 (Fig. 2). Funakoshi (1996) dem- 
onstrated that abundance of macrozooplankton was 
negatively correlated with the biomass of young-of- 
the-year sardines in Ise Bay (Fig. 1). He considered 
that large biomasses of young-of-the-year sardines 
resulted in a reduced abundance of macrozoo- 
plankton, followed by a decline in growth of the sar- 
dines through density-dependent processes. The 
greater population of sardine larvae in the shirasu 
fishing grounds in 1990 than in 1991 may have re- 
sulted in the slower growth of the sardine larvae in 
1990. We need to study the density of food items for 
sardine larvae in the shirasu fishing grounds to ex- 
amine if sardine larval growth is limited by food 
availability. We also need to know interspecific com- 
petition for food between sardine and anchovy lar- 
vae in a coastal ecosystem, because they coincided 
in April and May in our study area (Fig. 4). 
The offshore Kuroshio frontal waters provide sar- 
dine larvae with sufficient foods. Nakata et al. (1995) 
calculated that the total food requirement of carnivo- 
rous macrozooplankters and sardine larvae was 
about 11% of the total production of small copepods 
( < 1 .0 mm prosome length) in Kuroshio frontal wa- 
ters. This resulted in a higher feeding incidence in 
the early larval stage ofS. melanostictus in the fron- 
tal waters compared with the inshore and offshore 
waters of the frontal area (Nakata, 1995). Juvenile 
S. melanostictus (32-48 mm fork length) collected in 
30 
20 
10 
March 1990 
/ 
0 
0 
0 
0 
0 
T 1 1 1 1 1 
0 10 20 30 
Age in days 
Figure 8 
Backcalculated growth trajectories of sardine larvae 
hatched in March and April in 1990 and 1991. Small dia- 
mond represents mean TL at age and vertical bar one SD 
of the m 
the Kuroshio frontal waters had stomach contents 
(mostly copepods and larvaceans) of 7-10% of wet 
body weight and were backcalculated to have grown 
at 0.8-0. 9 mm/d in the larval stage (Watanabe and 
Saito, 1998). Perhaps the coastal waters in and 
around the shirasu fishing grounds are a less favor- 
able feeding area for sardine larvae than the 
Kuroshio frontal waters. 
Acknowledgments 
We thank Shigeo Funakoshi for comments on ecol- 
ogy of larval sardine and anchovy in the study area. 
Masao Bando sampled the sardine larvae in the fish- 
ing ports. Hisae Furukawa prepared and read otolith 
specimens. This work was supported in part by 
Grants-in-Aid from the Ministry of Agriculture, For- 
estry, and Fisheries (Biocosmos project, BCP-98-IV- 
A-8) and from the Ministry of Education, Science and 
Culture. 
Literature cited 
Anderson, J. T. 
1988. A review of size dependent survival during pre-re- 
cruit stages of fishes in relation to recruitment. J. North- 
west Atl. Fish. Sci. 8:55-66. 
