JUNE 



Figure 7. — Naknek River escapement and weighted 

 spawning ground tag recovery distributions by day of 

 tagging, 1962 (Grosvenor Lake, Brooks Lake, and 

 Naknek Lake). 



only in the Brooks River weir tagged fish count 

 but also in the tag recoveries from the streams 

 tributary to Brooks Lake (table 2). 



Grosvenor River had no tagged fish from the 

 early part of the run (fig. 6). Grosvenor River is 

 one of the areas where considerable difficulty was 

 experienced in recapturing tagged fish and in 

 identifying the color combination of those sighted. 

 It is possible, because of the generally smaller 

 numbers of fish tagged in the early part of the run 

 and the few tagged fish recaptured or identified 

 (table 2), that sampling was inadequate and early- 

 run fish were missed as a result of chance. On the 

 other hand, fish from late-run taggings (i.e., after 

 July 2) comparable in size to fish in some early-run 

 taggings were recaptured or sighted in Grosvenor 

 River. This plus the fact that tags were not 

 recovered from the only large tagging day, June 



30, in the early part of the run (table 1) strongly 

 suggests segregation of the Grosvenor River 

 population in the latter part of the run. Not 

 enough tags were recovered, however, to formulate 

 a definite conclusion concerning the segregation 

 of the Grosvenor River spawning populations. 



Results of Statistical Comparisons 



Chi-square tests of the hypothesis of like re- 

 covery distributions for comparisons between 

 similar and dissimilar spawning areas yielded 

 results that substantiate those presented above. 

 Probability values less than P=0.05 are con- 

 sidered to indicate unlike tag recovery distributions 

 and, therefore, segregation in time of occurrence in 

 the run. 



For the comparison between major valley or 

 terminal streams (American, Hardscrabble, Bay 

 of Islands, and Margot Creeks (table 3)), a value 

 of x 2 =5.38, d.f. = 6, P=0.50 was obtained. This 

 high value of P indicates no significant difference 

 between the tag recovery distributions for these 

 streams and, therefore, little segregation of their 

 spawning populations by time of occurrence in the 

 run. 



A value of x 2 =4.49, d.f. = 4, P=0.37 was 

 obtained from the comparison of recovery dis- 

 tributions for the grouped lateral streams of 

 Naknek, Brooks, and Grosvenor Lakes (table 4). 

 Coville Lake lateral streams could not be included 



Table 3. — Numbers of tags recovered by dates of tagging for 

 major valley streams, Naknek River system, 1962 



Table 4. — Numbers of tags recovered by dales of tagging for 

 grouped lateral streams, Naknek River system, 1962 



SALMON SPAWNING POPULATIONS IN NAKNEK RIVER 



471 



