Hulson et al.: Distribution of sampling effort for age composition of multiple species 
331 
Table 2 
Minimum and maximum estimated age sample sizes across sampling goals under simple random sampling 
(SRS), proportional allocation (PA), and fixed allocation (FA) 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). Estimated sample sizes without aging error are shown on the left of the “|” symbol 
and sample sizes including aging error are shown on the right (species acronyms are provided in Table 1). 
The top row for each species type contains the average of the minimum and maximum. Species acronyms are 
explained in Table 1. 
Species Region SRS PA FA 
Flatfish 
Avg. 
Min-Max 
269-730 
301-737 
205-659 
244-706 
447-1354 
523-1374 
AP 
BS 
286-687 
300-716 
224-666 
259-676 
530-1330 
616-1298 
F8 
BS 
278-640 
318-604 
197-565 
236-574 
356-1079 
389-1066 
NRS 
BS 
192-570 
208-727 
136-380 
152-406 
370-882 
446-890 
YS 
BS 
232-662 
249-676 
150-614 
178-630 
314-1358 
338-1365 
AF 
GOA 
240-458 
254-615 
151-410 
190-548 
369-836 
445-941 
DS 
GOA 
460-1686 
560-1515 
377-1614 
492-1785 
646-3058 
820-3124 
FS 
GOA 
268-616 
297-594 
222-556 
258-570 
600-1416 
718-1432 
NRS 
GOA 
210-579 
226-564 
174-526 
188-530 
400-1028 
430-1025 
RS 
GOA 
217-888 
257-852 
187-801 
224-843 
478-1699 
576-1700 
SRS 
GOA 
310-516 
342-504 
236-462 
265-498 
410-853 
447-898 
Rockfish 
Avg. 
Min-Max 
406-1823 
516-2967 
348-1977 
467-3002 
805-4805 
1072-6494 
NR 
AI 
360-1433 
427-1533 
335-1400 
402-1516 
566-3082 
1024-3386 
POP 
AI 
319-2000 
416-6319 
261-2702 
366-6218 
650-4882 
782-12224 
RB 
AI 
678-2522 
989-2523 
661-2822 
980-2662 
1928-8584 
2871-7710 
LDR 
GOA 
258-1030 
290-1172 
210-1044 
206-1146 
524-2726 
516-3040 
NR 
GOA 
406-1423 
464-1456 
342-1371 
421-1433 
660-4139 
860-3918 
POP 
GOA 
346-1670 
370-5015 
231-1718 
249-5300 
473-3886 
338-9380 
RB 
GOA 
477-2684 
655-2751 
396-2781 
642-2740 
837-6333 
1116-5801 
Roundfish 
Avg. 
Min-Max 
126-350 
133-329 
55-255 
62-256 
147-509 
200-576 
AM 
AI 
110-174 
112-180 
72-150 
75-155 
155-342 
178-379 
WP 
AI 
170-499 
191-427 
110-446 
135-374 
238-849 
334-876 
PC 
BS 
109-262 
117-254 
20-137 
40-188 
134-299 
187-389 
WP 
BS 
108-490 
109-510 
50-404 
8-412 
92-827 
105-942 
PC 
GOA 
138-294 
143-242 
69-133 
105-157 
191-323 
318-427 
WP 
GOA 
118-383 
125-359 
6-260 
8-250 
70-413 
75-440 
and biomass from survey data. Observation error in 
survey age-composition data from the operating model 
was generated with the multinomial distribution. To 
evaluate the influence of age-composition sample size 
in a fishery-independent survey on SCAA model results, 
13 sample sizes were used to generate trawl survey age 
composition data that ranged from 10 to 100,000 (by 
multiples of 2.5 and 2, e.g., 10, 25, 50, 100, 250, 500...). 
The influence of survey biomass uncertainty was 
evaluated concurrent with age composition uncertainty 
with 4 index uncertainty cases. These cases focused on 
the CV used to generate observation error in the log- 
normal survey biomass time series. Index case E0 gen- 
erated log-normal survey biomass data with a CV set 
at the average obtained by the AFSC bottom trawl sur- 
vey in the GOA (CV=9% for arrowtooth flounder, 25% 
for Pacific ocean perch, and 18% for walleye pollock). 
Index case El multiplied the CV in case E0 by 2. Index 
case E2 set the CV at 10% for all species types, and E3 
set the CV at 25% for all species types. Although some 
of the index cases may not occur in reality (for exam- 
ple, setting the CVs equal across species), our goal was 
to investigate the relationship with survey index CV 
as well as age composition based on survey data and 
a range of values is needed. Unlike the actual AFSC 
bottom trawl survey time series in the GOA (which is 
triennial from 1984 tol999 and biannual from 1999 to 
2011) this simulation analysis generated annual trawl 
survey biomass and age-composition data, so that vari- 
ability in model estimation results was not sensitive to 
gaps in the time series based on data from the trawl 
surveys. 
In the estimation models, the same number of pa- 
rameters was estimated for each species type so that 
resulting uncertainty was more directly comparable 
and was not sensitive to parameter differences. Estima- 
