392 
Fishery Bulletin 118(4) 
9S 
wo 
oS 
DN) 
S) 
i 
— 
9 
Catch variability 
i) 
Catch variability 
Catch variability 
Catch variability 
lh 
3 11 19 27 35 43 51 
Scenario 
3 11 19 27 35 43 51 
Scenario 
2 
ro) 
ln Al 
3 11 19 27 35 43 51 
Scenario 
Mill 
3 11 19 27 35 43 51 
Scenario 
|E HR10 F HR15 HR15U HR30 
© 
w 
\H 
‘= 
a 
0.24 
—) 
De) 
—] 
De) 
1 
oS 
—_ 
Catch variability 
Catch variability 
Catch variability 
S 
— 
Catch variability 
3 11 19 27 35 43 51 
Scenario 
3 11 19 27 35 43 51 
Scenario 
sill 
rene ly 
3 11 19 27 35 43 51 
Scenario 
"lal dl 
sleweals = lL. l=. 
3 11 19 27 35 43 51 
Scenario 
0.0 0.0 
Figure 6 
Median catch variability in the directed pot fishery for golden king crab (Lithodes aequispinus) in the Aleutian Islands 
during the last 10 years of a 30-year projection period, which begins in 2018, for 53 scenarios of an operating model used to 
evaluate harvest control rules (HCRs). The stock is projected from 2 initial levels of abundance, measured in mature male 
biomass (MMB). Values are given in numbers without a unit for HCRs (A) HR10, (B) HR15, (C) HR15U, and (D) HR30, with 
the stock set initially at a healthy state (.e., MMB.0;:/MMB3,=1.55, where MMBoo1g is MMB in 2018 and MMB,, is 35% of 
the unfished level of MMB), and for HCRs (E) HR10, (F) HR15, (G) HR15U, and (H) HR380, with the stock set initially at an 
overfished state (i.e., MMB. ;./MMB,-=0.50). In the model used in this analysis, a linear relationship between catch per unit 
of effort and selected abundance is assumed. Data used in the model are for golden king crab in 1981-2018. For details about 
the HCRs, see Table 1. 
rebuilding time is short when values for steepness are 
higher (e.g., scenario 9) and is longer for HR30 than for 
HR10, HR15, and HR15U for most scenarios. The rebuild- 
ing times necessary to achieve MMA,,, are longer than 
those necessary to achieve MMB,, for all non-zero exploita- 
tion rates. The results for the nonlinear choice are similar 
(results not shown). 
The MMA and CPUE distributions for scenario 1 (with 
parameters estimated through the use of the assess- 
ment model) indicate that HR10, HR15, and HR15U per- 
formed better than HR30. Fewer instances of MMA being 
lower than MMA,,,. and more frequent values of higher 
CPUE were observed for HR10, HR15, and HR15U than 
for HR30 (Fig. 10). Catch distributions were similar for 
HR15 and HR380 and for the equilibrium state; size com- 
positions of crab in total catch as well as in retained catch 
did not differ significantly from those for HR15 and those 
for HR30 (Suppl. Fig. 8) (online only). This result occurred 
because the stock was projected from a healthy state 
(MMB>MMB,;;) and because the effective F under both 
HR15 and HR30 was dampened because of reduction in 
abundance over the years. 
The results for the conservation and economic perfor- 
mance criteria of HR15 and HR15U for the last 10 years of 
the 30-year projection period were very similar for the lin- 
ear and the nonlinear relationships of CPUE to abundance 
(Tables 4-6, Suppl. Tables 7—9 [online only]). When compared 
with HR10 and HR30, HR15 may optimize the balance 
between conservation and economic criteria (Tables 4-6). 
Conservation metrics were similar among all HCRs except 
for HR30 (nonlinear choice; Suppl. Table 7 [online only]). 
Probabilities of fishery closures were zero, and catch vari- 
ability was similar for all HCRs, yet relative stock status 
was lower under HR30. Relative to HR15, HR10 required 
substantially lower effort (~48% lower) to achieve the TAC, 
yet the TAC itself was lower (~15% lower); in contrast, 
HR30 required higher effort (~14% higher) to achieve only 
a marginally higher (~2% higher) TAC (Table 5). 
