Zheng et al Catch-length analysis for crab populations 



583 



Kodiak Island red king crab 



As in the case of Bristol Bay RKC, the abundance in 

 the first year and annual recruitment of red king crab 

 off Kodiak Island were mainly affected by natural 

 mortality and the weighting factor (Table 2). For all 

 fits, the recruitment in 1965 was much higher than 

 those in other years. With A = 1, total RSS decreased 

 slightly with increasing natural mortality, whereas 

 the RSS contributed by fishing effort increased ( Table 

 2). Higher weighting factors resulted in higher total 

 RSS (Table 2). For the first fishing period, selectiv- 

 ity coefficients for the first length class were about 

 0.5, and no selectivity was detected for the other 

 length classes (Table 2). For the second fishing pe- 

 riod, catches concentrated on large-size crabs. 



The estimated legal male abundances of Kodiak 

 RKC peaked in 1965 and fell to a very low level in 

 the late 1960s (Fig. 6). The population moderately 

 recovered in the early and middle 1970s but declined 

 again in the late 1970s and completely collapsed af- 

 ter 1982. Model estimates of abundance with M = 

 0.3 and different weighting factors basically followed 

 the same trends as the fishery CPUE and the abun- 

 dances estimated by pot surveys (Fig. 6). Note that 

 no survey data were available until 1973 (Peterson 

 et al., 1986). Like the results for Bristol Bay RKC, 

 the abundance estimates with A = 1 followed the 

 trend of survey abundances best among all fits. The 

 scaled CPUE was very close to survey abundance 

 estimates from 1973 to 1977 but were much lower 

 from 1978 to 1981. 



Similar to Bristol Bay RKC, the estimated abun- 

 dances for Kodiak RKC with different natural mor- 



1 o 



09 



| 08 



w 



§. 07 



E 



8 06 



I 05 



04 



>. 



1 03 

 cd 

 A 



o 0.2 - 

 CL 



1 



00 



137 5 142 5 147 5 152 5 157 5 162 5 167 5 



Carapace length (mm) 



Figure 5 



Comparisons of length compositions of recruits (e: A=0 and 

 M=0.3, f: A = 10 and A/=0.3l, molting probabilities during 

 periods 1974-76, 1980, 1987-88, and 1992-93 (c: A=5 and 

 M=0.3, d: A = l and Af=0.3), and during periods 1977-79 

 and 1989-91 (a: A=l and M=0.5, b: A=0 and M=0.3) from 

 the two most disparate fits among all model fits for Bristol 

 Bay red king crab. 



talities had a similar trend over time, and a higher 

 natural mortality produced higher abundance esti- 

 mates (Fig. 7). The estimated abundances with natu- 

 ral mortality of 0.2 were lower than survey abundances 

 for most years, whereas the abundances estimated with 

 natural mortality of 0.5 were generally higher. With M 

 = 0.4, the closest abundance estimates to survey abun- 

 dances were produced among all fits ( Fig. 7 ). 



