12. The early run of sockeye salmon con- 

 sisted mostly of lake tributary spawners; the 

 middle run was a mixture of tributary and 

 Brooks River spawners; and the late run was 

 primarily Brooks River spawners. This dif- 

 ferential distribution by time period suggests 

 that the Brooks run is composed of several 

 races, each with different migratory habits. 



13. Egg retention by Brooks spawners is 

 low, suggesting successful egg deposition. 



some migration continued all summer and 

 fall. 



21. Throughout the season, 1,293 outmigrant 

 sockeye salmon were caught with fyke nets. 



22. Outmigrants during the main migration 

 were composed of two distinct age and size 

 groups: 2-year-olds, averaging 109 mm. mid- 

 eye-fork length, and yearlings, averaging 83 

 mm. 



14. By marking dead salmon with distinctive 

 fin clips on successive stream surveys, we de- 

 termined that within 7 days the dead fish dis- 

 appeared almost completely. Thus, counts of 

 dead fish on successive weekly surveys are 

 additive for minimum total counts. 



15. Bears eat many sockeye salmon spawn- 

 ers but take mostly spent fish. 



16. Because of removal of salmon from the 

 streams by bears, tag ratio samples from dead 

 fish in Brooks tributaries are impractical. 



17. On the basis of 2,333 observations of 

 spawning activity on Brooks River, the follow- 

 ing conclusions were made: (a) A 1:1 sex 

 ratio was most common; (b) males were twice 

 as aggressive as females in defending a redd 

 territory; (c) three separate and distinct 

 spawning waves occurred on Brooks River on 

 the same spawning area; (d) there was a rela- 

 tion between downstream weir counts and 

 spawning waves in Brooks River; (e) there 

 was no relation between measured environ- 

 mental factors and spawning behavior; (f) fe- 

 males lived considerably longer than males 

 and defended a redd until death; and (g) females 

 averaged 7.75 days on a redd from time of 

 selection through spawning to death. 



23. The total season's outmigrant age com- 

 position was 2 percent fry, 42 percent year- 

 lings, and 56 percent 2-year-olds. 



24. Representative daily samples were ob- 

 tained from samples with as few as five fish, 

 taken at any location at Brooks Lake outlet. 



25. Sockeye salmon fry and finge r lings 

 emerging from Brooks River gravels migrate 

 downstream into Naknek Lake and do not mi- 

 grate upstream into Brooks Lake in significant 

 numbers. 



26. Sockeye salmon migrants feed actively 

 and grow rapidly up to the time of seaward 

 migration. 



27. Sockeye salmon fry in Brooks Lake trib- 

 utaries migrated completely out of small trib- 

 utaries by June 1, 



28. Canadian tow nets and beach seines were 

 unsatisfactory for representative sampling of 

 sockeye salmon in Brooks Lake. 



29. Small-mesh gill nets, although subject 

 to limitations, are the most practical gear for 

 sampling lake-dwelling sockeye salmon in 

 Brooks Lake. 



18. Diurnal mass movements of sockeye 

 salmon spawners in and out of One Shot and 

 Hidden Creeks did not occur in 1957. 



19. Fewer than 400 sockeye salmon spawned 

 on Brooks Lake beaches in 1957, indicating that 

 beach spawning was of negligible importance. 



20. Peak migration of juveniles out of 

 Brooks Lake occurred on May 29, but 



30. In nylon gill nets, 3/4-inch stretch mesh 

 was the most efficient size for capturing 

 yearling sockeye salmon. 



31. Five hundred and five juvenile sockeye 

 salmon were caught in gill nets during the sum- 

 mer. Other catches included: 138 adult sockeye 

 salmon, 26 coho salmon, 62 rainbow trout, 

 4 Dolly Varden, 23 lake trout, 120 round white- 

 fish, 26 pygmy whitefish, 78 sculplns. 



62 



