reduced in size for a period of days, but 

 the wandering fish would eventually return to 

 the weir area at the lake outlet. Although this 

 upstream school was usually quiet and seden- 

 tary, sometimes it would rove actively back 

 and forth across the face of the weir. At these 

 times major downstream migrations through 

 the counting openings in the weir were made. 

 These mass behavior displays could not be 

 related to any obvious physical feature of the 

 environment, and further study is needed to 

 explain them. 



Eventually 86 percent (27,183 fish) of the 

 total upstream count were counted back down- 

 stream into Brooks River, where they ulti- 

 mately spawned (figs. 3 and 4), leaving only 

 4,414 spawners in the lake system. Because 

 no provision had been made for estimating 

 downstream migration of spawners from 

 Brooks Lake into Brooks River between 1940 

 and 1955, the annual counts for these years 

 are not indicative of the true numbers of 

 spawners in Brooks Lake and its tributaries. 

 Therefore, caution must be used in comparing 

 the annual counts shown in figure 4. 



Counts of sockeye salmon migrating up- 

 stream through the weir should have provided 

 a reliable measure of the number entering the 

 lake, but during the tagging program, it was 

 discovered that many sockeye salmon, as well 

 as other fish, were not counted because of the 

 following reasons: (1) Small sockeye salmon, 

 trout, and pink salmon would sometimes 

 squeeze between the pickets; (2) schools of 



NET COUNT 



DOWNSTREAM 



n ADDITIONAL EST 



46 48 50 

 YEAR 



Figure 4.— Annual counts of sockeye salmon at Brooks 

 Lake weir, 1940-57. 



sockeye salmon would occasionally be dis- 

 turbed below the weir and would hit the pickets 

 with great force, pushing themselves through 

 by spreading the pickets; (3) late in the season 

 bears would occasionally molest the weir and 

 dislodge pickets; and (4) the weir had a struc- 

 tural defect. Methods of estimating the un- 

 counted salmon are discussed in the section on 

 the distribution of spawners, p. 9. Since the 

 weir was constructed the same in 1957 as in 

 previous years, unknown and uncounted num- 

 bers of salmon probably entered the lake 

 before 1957. 



The downstream migration through the weir 

 in 1957 (fig. 3) is assumed to be accurate 

 because fish moving downstream passed only 

 through the openings that were purposely pro- 

 vided for counting and did not attempt to force 

 their way through the pickets. 



In addition to the sockeye, other salmon 

 passing through the weir in 1957 included 18 

 king, 7 chum, 461 coho, and 1 pink. 



AGE AND LENGTH COMPOSITION, 

 SEX RATIO, AND FECUNDITY 



During the main migration period (June 28 to 

 September 10) 704 adult sockeye salmon were 

 captured in a 6- by 8-foot trap located on the 

 upstream face of the weir 40 feet from the 

 west end (fig. 2), These fish made up a sample 

 to determine age and length composition, sex 

 ratio, and fecundity of the total run. Fish were 

 dipped out of the trap individually with a long- 

 handled net. After they were tagged and the 

 necessary data were recorded, they were 

 immediately released above the weir, excepting 

 those killed to determine fecundity. The trap 

 entrance was usually opened early in the 

 morning and was left open until 2 percent 

 of the previous day's count of adult migrants 

 had entered. It was then closed, the fish tagged 

 and processed, and counting through the weir 

 begun. The trap was checked frequently and the 

 entrance closed before overcrowding occurred. 

 Samples were taken every day there was an 

 upstream migration. 



To determine age a scale was taken from 

 the left side of each sample fish about halfway 



