Anderson: Age. growth, and mortality of Pandalus borealis Kroyer 



547 



eating a higher growth rate for this year-class. For ex- 

 ample, the 1971 year-class attained average sizes of 

 16.47 and 21.75mm CL at ages 3.4 and 6.4, respective- 

 ly, whereas the 1975 year-class averaged 18.20 and 

 23.01mm CL at these ages. Skuladottir (1981) also 

 detected large differences in growth rates between 

 year-classes. In her study, one slow-growing year-class 

 had a K = 0.15 and L M = 28 which is similar to the 

 parameters calculated for the 1971 year-class in this 

 study. On the other hand, parameters calculated for 

 the average of five fast-growing year-classes in her 

 study indicated a K = 0.23 which is lower than the K 

 (0.29) calculated for the 1975 year-class in the present 

 study. 



While growth of P. borealis is probably not related 

 to overall population density, there is evidence of 

 an inverse relationship relative to within year-class 

 strength. Both the 1971 and 1975 year-classes hatched 

 during periods of high overall population levels (Table 

 1), but the faster early growth of the 1975 year-class 

 may be explained by its relatively lower abundance 

 (Table 2) and, presumably, reduced competition for 

 food during the juvenile phase. Pandalus borealis is an 

 aggregating species exhibiting differential distribution 

 by size, sex, age, and season (Shumway et al. 1985). 

 Although most larvae are captured between 20 and 

 30 m in the water column (Haynes 1983), Wolotira et 

 al. (1984) report a downward shift in vertical distribu- 

 tion with progressive stages of larval development. 

 They theorized distribution differences reflect either 

 a change in diet or distribution of food items. Berkeley 

 (1930) also describes the apparent segregation of juve- 

 niles from the adult population. The effect of ecological 



separation of life-history stages could, therefore, ex- 

 plain differing growth rates among year-classes even 

 though overall population density was high. 



Since 1979, the occurrence of small shrimp (< 12 mm 

 CL) in survey samples has been much less than in pre- 

 vious years (Figs. 2A, 3). Three possible explanations 

 for the virtual disappearance of this size-class are (1) 

 small shrimp may only be retained when overall catch 

 rates are as high as they were in 1972-77, (2) juvenile 

 shrimp may not normally be found within the same area 

 as larger adult shrimp except when high population 

 levels force them into the less-preferred adult habitat, 

 and (3) faster growth of juvenile shrimp may have led 

 to entering year-classes growing beyond 10 mm CL to 

 13-15mm CL since 1979. I believe the most plausible 

 explanation for the disappearance of small shrimp from 

 survey samples is faster growth of juveniles. Results 

 of this study suggest growth is inversely related to 

 year-class strength. The overall population decline of 

 P. borealis in Pavlof Bay (Table 1) is attributed to the 

 dying out of the relatively strong 1971 and 1975 year- 

 classes and a series of relatively weak entering year- 

 classes since 1979 (Figs. 2, 3). The 1.4 year-old group 

 is now between 13.2 and 14.4mm CL rather than the 

 10-11.8mm CL that was observed for 1972-79 survey 

 samples (Table 4). The possibility of missing size modes 

 in this recent data series is low because sample sizes 

 remained large (about 5000 shrimp per survey). Inde- 

 pendent sampling of shrimp length frequency from cod 

 stomachs captured in 1980 and 1981 trawl surveys 

 showed cod consume smaller shrimp, probably the 

 0.4 year-olds (6.5mm CL), than were found in trawl 

 samples (Albers and Anderson 1985). Beyond 10 mm 



