338 
Fishery Bulletin 115(3) 
A EO 
\ O 
• ' 
ACV(Flatfish) = 0.058 
ACV(Rockfish) = 0.129 
ACV(Roundfish) = 0.21 3 
O Flatfish 
© Rockfish 
O Roundfish 
'O 
w -O — <5 -o -O --O -O -O 
'O — 
0 -o -0--9 -o -0--0 -o -o 
n = 2500 
C E2 
ACV(Flatfish) = 0.059 
ACV(Rockfish) = 0.156 
ACV(Roundfish) = 0.218 
O-O— 0-0-0 
? E1 ACV(Flatfish) = 0.06 
ACV(Rockfish) = 0.094 
\ ACV(Roundfish) = 0.1 89 
0-O— o-O-O— 0-0-0 
0-6-0— O-O-O— 0-0-0 
D E3 
o 
• b 
ACV(Flatfish) = 0.056 
ACV(Rockfish) = 0.127 
ACV(Roundfish) = 0.204 
O. « 
O ''Q 
'0 *-0 -ft =0 ==o -o -o — o -o -o 
feg 10 (Surv#y age-cofiipdsition sarfiple size) 3 
Figure § 
Coefficient of variation (CV) for biomass during the final year from the estimation models 
for the species types (flatfish, rockfish, and roundfish) evaluated across survey age-com- 
position sample sizes (per species) and survey index uncertainty cases (A) EO, (B) El, (C) 
E2, and (D) E3. Text in the top right corners of each graph denotes the absolute change in 
CV for each species type, and the vertical lines with arrows indicate the age sample size 
beyond which the CV in the biomass during the final year changed less than 2.5% for all 3 
species types. 
tics, in particular growth. It was shown that the rela- 
tive proportion of sample size could be related to the 
growth rates of the species considered. The significant 
relationships between the proportion of age-composi- 
tion sample sizes by species and life-history statistics 
could be due partially to the patterns in the variance 
between- and within-length intervals that resulted 
with species types. 
On the basis of our results of this study, species that 
are relatively slower growing require more samples to 
determine age composition than those that are relative- 
ly faster growing. The overall idea being that for rela- 
tively faster growing species, there is more distinction 
between the lengths at a given age; thus, the length 
composition is relatively more informative regarding 
age than it is for slower growing species, and fewer 
age samples are need to determine the age composition 
when performing 2-stage sampling. These results are 
also generalizable beyond just the species sampled by 
the AFSC bottom trawl surveys. Any fisheries science 
organization around the world will be constrained by 
the total number of otoliths it can process in a given 
year when considering how to distribute age sample 
size in a fishery-independent survey or fishery. The 
guidance that this study provides is that the growth 
characteristics of the species being sampled can help 
determine the relative magnitude of the sample size 
that should be used for each species. However, the re- 
sults of this study should be taken in light of the cave- 
ats inherent to the method used to determine the dis- 
