Huison et al: Distribution of sampling effort for age composition of multiple species 
335 
A aeo 
A 
□ Roundfish 
□ Flatfish 
■ Rockfish 
3 S' 9 a s 
■ c 
AE1 
ii" 
n 
■ 
^ o S B 
Combined 
5j'8§c*Oco«’fHSQ 
Species (region) 
Sample sizes without aging error (AEO) 
Figure 3 
Proportion of total age-composition sample size for the (A, C) combined species types and (B, D) 
individual species investigated without aging error (AEO) and with aging error (AE1) and (E) 
direct comparison of estimated age-composition sample sizes with and without aging error by 
using data from the NOAA Alaska Fisheries Science Center bottom trawl surveys for the Gulf 
of Alaska (GOA, 1984-2011), Aleutian Islands (AI, 1980-2010), and Bering Sea (BS, 1982-2011). 
Proportional sampling and sampling goal 4 (SG4) are used for illustration. Species acronyms 
are explained in Table 1. rc E1 =age-composition sample size with aging error; n E o=age-composition 
sample size without aging error; E 2 =coefficient of multiple determination; CI=confidence inter- 
val. Note the different scale values on the y axis: the left side designates the proportion of total 
sample size combined across species types; the right side designates values for individual species. 
than the sample size of 500 samples for case El (verti- 
cal lines with arrows in each plot in Fig. 6). Likewise, 
the sample sizes at the point of diminishing returns 
were greater for E2 (5000) compared with E3 (500). 
Correlation analysis between changes in the esti- 
mates of the CV in the biomass of the final year, re- 
cruitment variability, and survey index uncertainty 
from the estimation model emphasized the relationship 
between these quantities and sample size. The percent 
change in the CV of the total biomass of the final year 
(relative difference between largest CV compared with 
baseline CV, or, the minimum CV obtained) was posi- 
tively correlated with recruitment variability (0.54; 
Fig. 7A). Conversely, the uncertainty in the survey in- 
dex was negatively correlated with the percent change 
in the CV of biomass in the final year (-0.57; Fig. 7B). 
Taken alone, however, neither of these correlations was 
as strong as when the percent change in the CV of to- 
