FISHERY BULLETIN: VOL, 77, NO 1 



Table l. — Summary of exploitable walleye pollock biomass and density by depth strata and 

 survey regions in the western Gulf of Alaska. Catchability coefficients 1.0 and 0,5 are used to 

 produce a range of pollock biomass and density. Biomass estimates were calculated from the 

 summer survey period due to limited spring survey coverage. 



added on the assumption that at age length is 

 near (Alverson and Carney 1975), and 3) with 

 nominal ages or ring counts incremented by the 

 fraction of a year between middates of spawning 

 and sampling (A? in Table 2). 



Because each growth pattern in Figure 6 repre- 

 sents a synthetic cohort, i.e., a composite of year 

 classes, the departures from the pattern, gener- 

 ated by members of the 1967 and 1970 year class- 

 es, were examined in detail. According to the age 

 composition discussed earlier, both year classes 

 were encountered in relatively high abundance at 

 one extreme of the survey range (Figure 3) and in 

 low abundance at the other (the 1967 year class 

 was prominent at Sanak-Unalaska, the 1970 year 

 class at Kenai). To examine the evidence for a 

 growth-density relation, the size differences be- 

 tween the synthetic growth curves and observed 

 mean lengths of the sampled age-groups of these 



year classes were calculated (lower part of Table 

 2). 



The results are shown along a schematic east- 

 west axis in Figure 7. In the easternmost region 

 ( Kenai), the strong 1970 year class indicates nega- 

 tive departures from expectation (at age 5), 

 whereas corresponding departures are positive for 

 the relatively weak year class of 1967 (at age 8). In 

 the westernmost region, the relative strengths of 

 these year classes seemed to be reversed, and the 

 direction of departures of their mean lengths at 

 ages 4 and 7 also reversed. By this criterion, the 

 segregation of the two components of each year 

 class was most pronounced at the extremes and 

 least SO in the intermediate Kodiak-Shelikof re- 

 gion where the lines cross. 



Age-specific observed lengths of the 1970 and 

 1967 year classes were also compared directly 

 with those of others, apparently weaker year class- 



270 



