548 



Fishery Bulletin 89(4). 1991 



CL, length frequencies were similar for 

 both cod and trawl data, suggesting that 

 the age designation of this larger 1.4 

 mode is correct. 



Variability in growth rates of year- 

 classes also causes variability in the age 

 of full recruitment in survey sampling, 

 owing to the size selectivity of our 32 mm 

 mesh trawl. Recruitment was complete 

 at ages 5.4 and 3.4 (18.96 and 18.20mm 

 CL) for the 1971 and 1975 year-classes, 

 respectively. Likewise, selectivity ex- 

 periments by Blott et al. (1983) showed 

 complete vulnerability of 19 mm CL P. 

 borealis in a 32mm stretch mesh trawl. 

 Fox (1972) reported full recruitment at 

 age 3 (=18. 0mm CL) in stocks of P. 

 borealis from Kodiak Island, Alaska, in 

 which sampling was accomplished with trawls having 

 a mesh size similar to those in the present study. 



Mortality estimates 



Total mortality estimates were based on year-class 

 abundance after full recruitment to survey sampling 

 (eq. 5; Table 5). The age of full recruitment to sur- 

 vey sampling was identified as 5.4 for the 1971 year- 

 class and 3.4 for the 1975 year-class based on the 

 visual inspection of the catch curve (Fig. 5). Since 

 the commercial fishery was closed before the 1979 

 survey, total mortality rates estimated for the 1975 

 year-class beyond age 4.4 are equivalent to natural 

 mortality. Table 5 also presents annual exploitation and 

 fishing mortality rates estimated for both year-classes 

 (from eq. 6). The 1971 year-class showed increasing 

 natural mortality between ages 5.4 and 8.4, while 

 fishing mortality remained relatively stable during this 

 period. 



Increasing total mortality with age for the 1971 year- 

 class is attributed to increasing natural mortality since 

 fishing mortality was constant. Apparent increases in 

 natural mortality could also be caused by emigration 

 of older individuals from the Pavlof Bay population, but 

 this seems unlikely since it has been demonstrated that 

 larger (older) shrimp are less active in diel vertical 

 migration (Barr 1970) and would tend to be retained 

 by the shallow entrance of Pavlof Bay. The observed 

 higher natural mortality in this study may be the result 

 of spawning stress or senescence since the 1971 year- 

 class may have approached their maximum longevity. 

 Predation by Pacific cod Gadus macrocephalus may 

 have also contributed to the high mortality observed 

 from ages 6.4 to 8.4 when cod catches increased from 

 10 to about 500 kg per nautical mile towed (Albers and 

 Anderson 1985). 



10000 





a. 



<f) 



CO 



c 

 o 





10 







-J h 



1 



Year 



Figure 5 



Catch curves (fitted by eye) of the 1971 and 1975 year-classes 

 from Pavlof Bay survey data. 



The 1975 year-class showed its highest calculated 

 natural mortality between ages 3.4 and 4.4 (Table 5); 

 however, these rates were subsequently lower. After 

 closure of the fishery, the total mortality rates for the 

 1975 year-class beyond age 4.4 are equivalent to 

 natural mortality. Albers and Anderson (1985) reported 

 the significance of Pacific cod predation on P. borealis 

 in Pavlof Bay. Pacific cod abundance peaked in 1979 



