could be distinguished by length-frequency 

 distributions; therefore, gill net-caught fry 

 and yearling fish were distinguished on the 

 basis of size alone without recourse to scale 

 reading (fig. 19). 



The decrease in mean length in June and 

 July may be explained by the inclusion in June 

 samples of larger 2-year-olds which sub- 

 sequently migrated from the lake and were 

 no longer available in August. Yearlings in- 

 creased in mean length from 95 mm, in August 

 to 101 mm. in September. Two-year-olds of 

 the same age class in the 1957 spring out- 

 migration (fig. 12) were only 109 mm. long, 

 or only 8 mm. longer than these fish in the 

 fall. If it is assumed that yearlings of each 

 of these year classes grew at the same rate, 

 the fish we sampled in the fall of 1957 would 

 gain only an additional 8 mm. during the 

 ensuing 7 winter months. This projected slow 

 winter growth contrasts with the 6 mm. of 

 rapid growth in only 1 month during the 

 latter part of the summer of 1957. 



The average daily weight of combined sam- 

 ples of gill net-caught fingerlings is shown 

 in figure 23. Points are scattered, indicating 



14-1 



12- 



X 



= 6 



2- 



YEARLINGS 



FRY 



I 



JUNE 



10 

 JULY 



20 



I 



10 



AUG. 



20 I 



10 20 

 SEPT. 



I 



10 



OCT. 



Figure 23.— Average daily weights of sockeye salmon 

 caught in gill nets in Brooks Lake, from June to 

 October 1957. (Curves fitted by inspection.) 



variable growth, possibly because of differ- 

 ences in diet. This contrasts with more 

 uniform weights of the 1957 outmigration 

 (fig. 13). Any final conclusions based on these 

 limited 1957 weight data would be premature, 

 but one may speculate that fish far below 

 average weight either perish during the winter, 

 or compensatory growth allows them to catch 

 up with the rest of the fish by the time of 

 spring migration. Apparently growth condi- 

 tions in Brooks Lake were about the same in 

 1957 as in 1956 since the 2-year-olds of the 

 1954 brood were the same weight at the start 

 of the 1957 outmigration as were the 1955 

 brood fish at the end of the 1957 summer 

 season. This conclusion assumes that little 

 weight gain occurred during the winter of 

 1957-58, as indicated by the flattening out of 

 the weight curve in figure 23. Too few fish 

 were caught from the 1956 brood to draw any 

 conclusions on their growth. 



FOOD AND ITS AVAILABILITY 



THROUGH SUMMER GROWING 



SEASON 



Stomach contents of all fish from gill nets 

 were examined. Each fish was weighed and 

 measured, and a scale sample was taken for 

 mounting in the same manner as previously 

 described for fyke net-caught fish. Stomachs 

 were then excised with a small scissors, 

 and the estimated percentage of fullness re- 

 corded (empty = percent, distended = 100 

 percent). The contents were placed on paper 

 towels and allowed to drain, after which they 

 were placed in graduated centrifuge tubes 

 where total volumes were measured by water 

 displacement to the nearest 0.05 ml. At the 

 beginning of the season an attempt was made 

 to measure volumes for each food organism, 

 but the volumes were so small that they 

 could not be accurately measured with avail- 

 able equipment. 



After total volume was measured, counts of 

 each organism were made. Zooplankters were 

 classified by genus and insects by common 

 names (or orders, if specific identification 

 was in doubt). Only whole or major portions 

 of insects were counted as individuals. When 

 insect fragments occurred, no attempt was 



35 



