FISHERY BULLETIN: VOL. 73, NO. 2 



60 

 50 

 40 

 30 

 20 

 I 

 

 70 

 60 



Ld 



^ 50 



I- 



z 40 



UJ 



o 30 



LiJ 



a. 20 



I 







70 



60 



50 



40 



30 



20 



10 







1965 



m^ 



1966 



1967 



1968 



■4=1=1 



969 



n 



■3' '5 7' 9 II 

 AGE (years) 



1 — r- 



FiGURE 5.-Age composition of Pacific hake taken off Washing- 

 ton during May-August, 1965-69. (Shaded bar denotes the 1961 

 year class.) 



collected in a stratified manner. Before the com- 

 mercial samples could render representative age 

 compositions, they had to be randomized. This was 

 accomplished by constructing an age-length key 

 for each sample. The percentage age frequency 

 per length interval observed in stratified sub- 

 samples was applied to the length frequency dis- 

 tribution of the entire sample. 



VARIABILITY IN 

 AGE COMPOSITION 



Several potential sources of variation in age 

 composition were analyzed using research and 

 commercial samples. Age composition is used for 



evaluation of the effects of fishing, estimation of 

 recruitment, growth, and mortality. Therefore the 

 major sources of sampling variation should be 

 identified. The effects of annual, seasonal, lati- 

 tudinal, and sexual variation are considered in this 

 section. 



Annual Variation 



Annual variation in age composition was 

 studied using samples collected off the Washing- 

 ton coast during May-September 1965-69. It has 

 been assumed (Tillman 1968; Nelson and Larkins 

 1970) that Pacific hake are fully recruited to the 

 fishery at age 5. If this assumption is valid and if 

 recruitment and mortality rates are constant then 

 one would expect, from a relatively unexploited 

 population, a typical catch curve with the 5-yr-olds 

 most numerous and the succeeding ages decreas- 

 ing at a rate equal to the rate of natural mortality. 

 This pattern was apparent in 1965 (Figure 5). The 

 partially recruited 4-yr-olds were not as numerous 

 as the 5-yr-olds which predominated. From age 5 

 there was a progressive decrease in the relative 

 abundance of succeeding age groups until only a 

 few 13-yr-olds remained. By examining the age 

 compositions in subsequent years, it became ob- 

 vious that the 4-yr-old age group in 1965 was con- 

 siderably larger than usual. This was the first in- 

 dication that the 1961 year class was unusually 

 large. In 1966 the 1961 year class (5-yr-olds) was 

 probably fully recruited and strongly dominated 

 the age structure. The relative abundance of the 

 incoming 1962 year class (4-yr-olds) was much 

 smaller than the 1961 year class in the 1965 

 samples. The 1961 year class can be followed 

 through the age composition in 1967-69. In 1969 

 the 1961 year class lost its dominance to the 1962 

 year class, but still produced extraordinarily large 

 numbers of 8-yr-olds. Apparently the 1965-66 year 

 classes were smaller than those observed 

 previously since in the 1969 sample no 4-yr-olds 

 were observed. 



Obviously annual variation in age composition 

 does occur in Pacific hake and is at least partly due 

 to varying levels of recruitment of incoming year 

 classes. 



Seasonal Variation 



In 1966-67 regular sampling of Pacific hake from 

 off the southern coast (lat. 46°00'-46°59'N) of 



342 



