escapement were determined from scale readings. 

 Percentages for the four dominant age groups 

 (4 2 , 5 2 , 5 3 , and 6 3 ), which made up 98 percent of the 

 1962 escapement, were plotted graphically for 

 each day of sampling. This was done for males 

 and females, both separated and combined. The 

 graphs were examined to determine if particular 

 age groups tended to be segregated in certain 

 portions of the run. Similar graphs were prepared 

 for the percentage of salmon that had spent 2 and 

 3 years in fresh water and 2 and 3 years in the ocean. 

 The 1962 escapement was composed almost en- 

 tirely (99.8 percent) of fish of 2- and 3-fresh- 

 water age, and 98 percent of 2- and 3-ocean age. 

 These graphs were examined for the extent of 

 segregation by both fresh-water and ocean ages. 



Age segregation on the spawning grounds was 

 studied by comparing the age composition of 

 individual spawning ground samples with each 

 other and with age composition of the total es- 

 capement. The age composition of the total 

 escapement was derived from the daily percentage 

 age composition, weighted according to the size 

 of the daily escapement. The escapement and 

 spawning ground age compositions were compared 

 on the basis of age group and of fresh-water and 

 ocean ages. Theoretically, with representative 

 sampling of all spawning areas, a lack of segrega- 

 tion by age categories would be shown by close 

 agreement between the total escapement and the 

 individual spawning ground age compositions. 

 Substantial disagreement would indicate segrega- 

 tion. 



The frequencies of occurrence of the four age 

 groups (i.e., 4 2 , 5 2 , 5 3 , 6 3 ) in most of the run were 

 arranged in a contingency table to compare the 

 age compositions of the individual major spawning 

 grounds of the Naknek system. The proportion 

 of fish in each age group was then tested for in- 

 dependence by chi-square. A probability value 

 of less than P=0.01 was considered to indicate 

 unlike age compositions between the areas tested 

 and, therefore, segregation by age on the spawning 

 grounds. 



RESULTS AND DISCUSSION 



Extent of Segregation by Time of Occurrence in the 

 Run 



Graphical comparisons of the pattern of weighted 

 recoveries by day of tagging with the pattern of 

 the daily escapement are a somewhat subjective 



approach to studying the extent of segregation. 

 On the other hand, statistical tests of like tag 

 recovery distributions between spawning areas are 

 objective in nature. Results obtained by both 

 methods were similar. 



Results of Graphical Comparisons 



Graphical comparisons of the pattern of weighted 

 tag recoveries for individual spawning areas by 

 days of tagging with the pattern of daily escape- 

 ment to the entire system are presented in figures 

 4-7. 



It is obvious from these comparisons that most 

 of the spawning areas derive their escapement 

 from all portions of the run. In most areas the 

 proportion of tag recoveries appears, in general, 

 to be related to the size of the escapement on the 

 date of tagging. Greater variability between the 

 tag recovery and escapement pattern is apparent 

 for such areas as Bay of Islands and Margot 

 Creeks, Grosvenor River, and Naknek River at 



o 

 o 



< 



cr < 



uj h- 



> o 



o i- 



U u- 



ui o 



o 



z 

 < 



a 

 < 



70 



60 



50 



40 



30 



20 



10 







60 



50 



40 



30 



20 



10 







40 



30 



20 



10 







25 30 



JUNE 



BAY OF ISLANDS 

 CREEK, NAKNEK 

 LAKE-RECOVERIES 

 n=40 



MARGOT CREEK 

 NAKNEK LAKE- 

 RECOVERIES 

 n=6l 



NAKNEK RIVER 

 ESCAPEMENT 



Figure 4. — Naknek River escapement and weighted 

 spawning ground tag recovery distributions by day of 

 tagging, 1962 (Margot and Bay of Islands Creeks). 



SALMON SPAWNING POPULATIONS IN NAKNEK RIVER 



469 



