714 
Fishery Bulletin 95(4), 1997 
(35-150 mm SL) preyed on shrimp and fishes. Most 
fish in the stomachs were well digested and difficult 
to identify, but the sagittae closely resembled those 
of white croaker, Genyonemus lineatus, and queen- 
fish, Seriphus politus. One case of cannibalism was 
observed; a 7-mm larva was eaten by a 35-mm juve- 
nile. Most shrimp were well digested, but at least 
two individuals were identified as belonging to the 
genus Crangon. Other items found in the stomachs 
included nematodes, bits of algae and surf grass, 
portions of crustaceans, and sand. The stomachs of 
six white seabass (5%) were empty or contained only 
nonfood items such as sand. 
Age and growth 
Otolith increments formed daily in sagittae of labo- 
ratory-reared white seabass (Fig. 7). The slope of the 
regression of observed number of increments on age 
was 0.96 and did not differ from unity (r 2 =0.96, n= 25, 
P<0.01, 95% confidence limits on slope: 0.89 and 1.04 
increments/d). The first increment formed 3-4 d after 
hatching, a period that corresponds to yolk absorption 
and onset of feeding (Kim, 1987; Orhun, 1989). 
Age was estimated for 50 wild white seabass rang- 
ing from 6.2 to 104 mm SL. The ten smallest fish 
that were aged ranged from 6.2 to 9.2 mm SL and 
were estimated to be 26-32 d old (Fig. 8). The age of 
a fish that was 15 mm SL, the length at metamor- 
phosis, was estimated to be about 40 d. The largest 
juvenile aged (104 mm SL) was estimated to be 108 
d old. The range of estimated ages suggests that 
white seabass remain in the nursery for 2-3 months 
after settlement. It should be noted that the oldest 
validated age was 76 d. A 149-mm-SL juvenile was 
not aged because it was probably much older than 
the oldest validated age. 
Growth of these fish was rapid in terms of length 
and weight. The parameters of the Gompertz model 
relating age and length were estimated as L 0 = 0.202, 
G=6.64, and g=0. 0273, where L 0 is length at time t 0 , 
G is the instantaneous growth rate at time t Q , and g 
is the rate of decrease of G (Fig. 8A). This equates to 
a maximum growth rate of 1.57 mm/d at 70 d. Weight 
also increased rapidly. The parameters relating 
weight and age were W 0 =2.72 x 10~ 7 , G=17.58, and 
g=0.0256, where W 0 is weight at time £ 0 (Fig. 8B). 
Wild fish between 6 and 104 mm SL grew at rates 
similar to those of laboratory-reared fish. Wild fish 
grew at a linear rate of 1.31 mm/d, compared with 
laboratory rates of 1.15, 1.32, and 1.04 mm/d for 
groups spawned in May, June, and September 1989 
(Fig. 9). Linear growth models were used to facili- 
tate statistical analysis by ANCOVA. Linear models 
fitted the data well, with coefficients of determina- 
tion (r 2 ) of 0.94-0.99 for the four groups. The rate of 
growth (slopes) did not differ among the four groups 
(ANCOVA, F=1.40, P=0.25). However, the June labo- 
ratory group was significantly larger at a given age 
than the wild fish (ANCOVA, F=5.05, P < 0.01; Tukey 
pairwise comparison), but the remaining groups did 
not differ in their length-at-age. 
The distribution of spawning dates, based on 
counts of otolith increments, indicated that spawn- 
ing occurred from March to July in 1987, and from 
