566 
Fishery Bulletin 96(3), 1998 
Table 1 
Total length composition (cm) of the biological sample of 
Pacific red snapper by sex. The column labeled “unsexed” 
groups fishes with undifferentiated gonads (almost all of 
those <30 cm) and those gutted and with no traces of go- 
nadal tissue (most of those between 30 and 40 cm and all 
of those > 40 cm). 
Size range 
Males 
Females 
Unsexed 
Total 
10.1-15.0 
0 
0 
24 
24 
15.1-20.0 
0 
0 
259 
259 
20.1-25.0 
4 
5 
366 
375 
25.1-30.0 
23 
23 
404 
450 
30.1-35.0 
67 
59 
305 
431 
35.1-40.0 
109 
76 
117 
302 
40.1-45.0 
66 
46 
59 
171 
45.1-50.0 
40 
28 
56 
124 
50.1-55.0 
32 
24 
52 
108 
55.1-60.0 
37 
17 
38 
92 
60.1-65.0 
26 
25 
26 
77 
65.1-70.0 
26 
23 
18 
67 
70.1-75.0 
21 
29 
11 
61 
75.1-80.0 
9 
30 
7 
46 
80.1-85.0 
5 
9 
0 
14 
85.1-90.0 
0 
3 
0 
3 
90.1-95.0 
0 
0 
0 
0 
95.1-100.0 
1 
0 
0 
1 
Total 
466 
397 
1742 
2605 
Male and female length compositions were signifi- 
cantly different (Kolmogorov-Smirnov two-sample 
test, D- 0.0836, 0.01<P<0.05). Although the largest 
specimen was a 99.2-cm-TL male (Table I), females 
predominated beyond 70 cm (sex ratio 1:1.97). 
Conversion equations for lengths (cm) and weights 
(kg) did not differ between sexes ( ANCOVA, P>0.05); 
therefore data were pooled: 
TL= 1.246SL + 0.104 
TW=1.207GW - 0.020 
GW= 1.763 x 10- 5 TL 2 877 
TW= 1.816 x 10“ 5 TL 2 905 
(rc=2603, r 2 = 0.997, 
P<0.001) 
(n=841, r 2 =0.988, 
P<0.001) 
(n=2350, P 2 =0.990, 
P<0.001) 
(n=844, P 2 =0.980, 
P<0.001). 
Exponents of the length-weight relationships were 
significantly different from three (f-test, P<0.05). 
Age determination 
Concentric annuli and circuli were observed in most 
otoliths and scales. Most scales were found to be in- 
terpretable, although circuli were never as clearly 
defined as otolith annuli. Most samples in-92) pre- 
Table 2 
Values of percent disagreement (%D) and percent average 
error (%E) for whole-otolith, sectioned-otolith, and scale age 
determinations for individual structures (precision) and between 
the structure and whole-otolith age (between structures) of 
Pacific red snapper. No. = number of determinations. Rejec- 
tions are otoliths not included in the analysis for being too 
large (size) or for lacking a clear marking pattern (other). 
Index Rejections 
Age 
%D 
%E 
No. 
Size 
Other 
Precision 
Whole-otoligh 
16.90 
3.89 
143 
7 
1 
Sectioned-otolith 
10.42 
2.13 
144 
1 
6 
Scale 
34.01 
4.19 
147 
2 
2 
Between structures 
Sectioned-otolith 
26.36 
2.82 
129 
Scale 
58.65 
8.86 
133 
sented less than 40% regenerated scales. A correla- 
tion was found between the percentage of regener- 
ated scales and fish length (r=0.310, PcO.001). 
Of the 1356 otoliths used for whole-otolith age de- 
terminations, 186 were discarded for the following 
reasons: large size prevented the enumeration of all 
the growth rings (2.2%), breakage or loss (1.4%), de- 
formities or abnormal calcification (2.3%), and non 
interpretable otoliths as defined in the “Methods” 
section (7.8%). Whole-otolith ages were available for 
1170 individuals ranging from 10.2 to 83.5 cm TL. 
A number of scales and otoliths were also rejected 
in the comparison of whole-otolith, sectioned-otolith, 
and scale ages (Table 2). Noninterpretable otoliths 
were included in the computation of precision indi- 
ces, but not in the between-structures indices. Most 
of the rejections of whole otoliths were due to size, 
but a similar proportion of sectioned otoliths lacked 
a clear marking pattern. Sectioned otoliths and scales 
were the most and least precise structures, respec- 
tively, for age determination (Table 2). The %D indi- 
cates a factor of three difference in the precision of 
sectioned otoliths and scales. However, the %E shows 
that such differences are less important (less than a 
factor of two) because this index incorporates the dif- 
ference in age estimations (Beamish and Fournier, 
1981). The difference between whole-otolith ages and 
sectioned-otolith or scale ages was much larger (Fig. 
4). Sectioned-otolith ages were found to be at least twice 
as similar to whole-otolith ages as scale ages according 
to %D, and more than three times as similar according 
to %E (Table 2). 
The magnitude of the discrepancies in age determi- 
nations among structures increased with both age and 
length of the fish (Fig. 4). Differences between scale 
