Shi et a I.: Growth and survival of Pleuronectes vetulus 
167 
Table 1 
Monthly population-weighted mean date (PWMD), mean modal length (MML), population-weighted mean bottom temperature 
(PWMBT) and overall mean densities of 0 + English sole, 1985-88. 
Year 
Settlement time 
Month 
PWMD (days) 
MML (mm) 
PWMBT (°C) 
Density' (No. /ha) 
1985 
Early 
May 
21.66 
60.56 
13.07 
55.03 
Jun 
55.34 
68.08 
14.66 
130.95 
Jul 
83.75 
79.95 
12.94 
200.60 
Aug 
112.39 
84.55 
14.34 
132.69 
Sep 
135.72 
99.96 
12.24 
96.42 
1986 
Early 
May 
17.05 
67.54 
13.61 
118.25 
Jun 
44.71 
72.97 
16.24 
77.06 
Jul 
77.54 
88.58 
14.66 
96.75 
Aug 
106.18 
99.17 
12.07 
72.55 
Sep 
141.32 
112.98 
12.49 
65.80 
1986 
Late 
May 
22.51 
25.99 
12.33 
118.25 
Jun 
48.11 
36.09 
15.31 
77.06 
Jul 
75.63 
55.01 
15.43 
96.75 
Aug 
102.64 
66.10 
14.63 
72.55 
Sep 
144.51 
84.67 
14.81 
65.80 
1987 
Late 
May 
21.47 
25.17 
10.48 
188.77 
Jun 
45.58 
36.64 
13.70 
219.35 
Jul 
76.00 
48.83 
15.28 
346.38 
Aug 
106.47 
63.19 
12.48 
186.98 
Sep 
130.45 
71.49 
13.65 
200.48 
1988 
Early 
May 
12.48 
45.01 
13.21 
269.44 
Jun 
57.61 
61.24 
15.20 
193.18 
Jul 
86.10 
79.77 
14.91 
182.15 
Aug 
113.75 
91.58 
14.78 
116.13 
Sep 
149.10 
103.72 
12.26 
90.31 
1 The estimated densities are combined densities of early and late recruits. 
coefficients during July, August, and September, respec- 
tively. Model selection was based on the Bayesian In- 
formation Criterion (BIC) proposed by Schwarz ( 1976). 
Results 
The mean modal length (MML), population- weigh ted 
mean date (PWMD), population-weighted mean bot- 
tom temperature (PWMBT), and overall mean den- 
sities of 0 + English sole from May through Septem- 
ber are shown in Table 1. 
Growth 
Growth of 0 + English sole was linear over time (Fig. 
4). A general linear model pooled all data together 
and considered the effects of year, time of settlement, 
density, and temperature on growth. The final, best- 
fitted (R 2 = 0.99, P < 0.001) model was 
l fi - 8.44 + 43.48c + 6.76jq +7.79y 2 - 13.06y 3 
+ 0.43/,- 0.11c/, + 0 . 063 /^ + 0.12y 3 /, + . 
The results of partial / -tests (used in all compari- 
sons unless specified otherwise), indicated there was 
a significant settlement time effect on growth 
(PcO.Ol), late-settling cohorts growing the fastest. 
The year effect was also significant (multiple partial 
E-test P<0.Q5, Table 2). Multiple-partial P-tests in- 
dicated that there were no density (P=0.80) or tem- 
perature (P- 0.37) effects on the growth of 0 + English 
sole. 
The date of settlement was estimated by fitting 
separate regression equations to each cohort in Fig- 
ure 4, then by backcalculating to a length of 20 mm 
TL (Table 3), or by inverse prediction (Neter et al., 
1985). We estimated that settlement of the 1985 and 
1986 group-1 cohorts peaked in January, with 95% 
prediction intervals ranging from 27 November to 
22 March. Settlement of the 1986 group-2 and 1987 
cohorts peaked in May (with 95% prediction inter- 
vals ranging from 21 April to 29 May), and that of 
the 1988 cohort peaked in March (ranging from 17 
February to 16 April). 
Mortality The following equations were obtained 
from nonlinear least-square regression: 
