of Commercial Fisheries on the Igushik River 

 and Nuyakuk River in 1958 and 1959, respec- 

 tively. Within the Wood River lakes system, 

 escapement estimates of individual areas are 

 being continued to assess the percentages and 

 numbers of fish spawning in each lake and 

 major river. Escapement estimates for the 

 period 1946-58 were evaluated by J, R. Gilbert 

 and for 1959 by M. O. Nelson, 



Escapement sampling . — Since 1946, exten- 

 sive annual sampling of dead fish on the 

 spawning areas has been conducted for length 

 frequency, scale samples, and sex ratio de- 

 termination. Dead salmon were measured 

 from mideye to hypural plate and measure- 

 ments converted to mideye-tail fork lengths 

 based on a conversion table worked out from 

 commercial fishery samples by R, E. Duncan 

 (1956). No correction has been made for pos- 

 sible differences in length between fish as 

 measured dead in commercial fishery samples 

 and as measured dead on the spawning grounds. 

 In general, larger samples were taken from 

 the more heavily utilized spawning areas. 

 Sampling was more comprehensive in the 

 Wood River lakes. Total numbers of fish 

 measured annually by sex are given in table 1. 



Accurate determination of sex ratio of fish 

 on the spawning grounds is difficult because 

 of observed differences between sexes in 

 behavior during and after spawning, in time 

 of dying after spawning, in drifting after death, 

 and in selective feeding by predators. The sex 

 ratios from live and dead counts that were 

 considered to be most representative have 

 been used to arrive at escapement sex ratios 

 in the Wood River lakes for the period 1946- 

 58. In 1959, sex ratios for the Wood River 

 lakes escapement were obtained from daily 

 samples collected throughout the season by 

 beach seine as the salmon entered Lake 

 Aleknagik from Wood River, Daily sex ratios 

 in 1959 were weighted by daily escapements 

 past the Wood River tower enumeration site. 



DETERMINATION OF AGE AND DIS- 

 TRIBUTION OF AGE GROUPS 



The red salmon of Bristol Bay mostly spend 

 one or two winters in fresh water and two or 



three winters in the ocean before maturing. 

 This gives rise to four major age components 

 in addition to several minor age categories. 

 The IBM program discussed earlier will 

 calculate the different age components of 

 both catch and escapement without the use of 

 length frequency measurements if represen- 

 tative age data samples are available. These 

 components can then be combined to examine 

 any relationships between year classes or 

 between escapement and return. 



For the period covered by this report a 

 somewhat different approach was followed 

 because escapement sampling for age was not 

 conducted in the trunk streams as it has been 

 since 1960. ' For the years 1946-59, scale 

 samples from the escapement were taken on 

 the individual spawning beds, where resorp- 

 tion of the marginal areas of scales prevents 

 an assessment of ocean age. Therefore, it 

 was necessary to ascertain ocean ige from 

 length frequency distributions. 



The size of adult salmon is largely deter- 

 mined by the number of years the fish spend 

 in the ocean. Those that have spent two winters 

 in the ocean form the small size group, and 

 those that have spent three winters in the 

 ocean, the large size group. These two size 

 groups usually differ in mean length by about 

 5 cm. 



In addition to the above two size groups, 

 there are a small number of jack salmon 

 (males that return after one winter in the 

 ocean) in some years and a variable but in- 

 significant percentage of fish that spend four 

 winters in the ocean. In calculating age dis- 

 tribution, the jack salmon are included with 

 the small size group, and the 4-year-ocean 

 fish, with the large size group. 



Further, since age composition in the catch 

 within individual fishing periods was generally 

 based on a scale sample one-fifth the size of 

 the sample used in determining length fre- 

 quency distribution, we felt that a better es- 

 timate of the distribution of ocean age would 

 be obtained by also using the available length 



' Trunk stream sampling for age composition was 

 initiated at Wood River in 1959, but data were not 

 available for the other trunk rivers of the district. 



