migrants followed the lake shores on each side 

 of the outlet stream, thereby being concentrated 

 along the stream-banks where the smaller 

 shore nets were located. 



From June 1 to 11, to determine if small 

 samples from each net's total nightly catch 

 would yield representative age data, ages of 

 approximately five fish from each net's catch 

 per night were compared with ages of the total 

 catch of each net (table 7). The random 

 method of sample selection described pre- 

 viously was used. 



Differences in age between small samples 

 and total catches on individual days were 

 quite large; however, age composition from 

 the small samples was identical with the over- 

 all age composition from the entire catch 

 for the period from May 27 to June 11 

 (table 6), and was only 3 percent different 

 from the total sample of the period June 1-11 

 (table 7). Small samples were therefore ade- 

 quate for determining age composition for the 

 season but not for individual days. 



apparent leveling off of weight after June is 

 not proved by these data because samples 

 taken after June represent small numbers of 

 fish. A rapid weight increase was noted through 

 June for yearlings of the 1955 brood, indicating 

 an adequate supply of food. Average weights for 

 the season were 11.46 g. for 2-year-olds, 

 8.80 g. for 1-year-olds, and 4.14 g. for fry. 

 In comparison with migrants from Canadian 

 and Russian lakes, the yearlings were some- 

 what heavier than those of Cultus Lake in 

 1927 and 1928, but were lighter than the aver- 

 age for yearlings from Kamchatka lakes 

 (Krogius and Krokhin, 1948). 



Almost all 2-year-olds of the 1954 brood 

 migrated from the lake by the middle of June 

 and during the short spring period gained 

 weight rapidly. Active feeding evidently con- 

 tinues at least until the time of migration 

 from the lake. 



DAILY AND SEASONAL MIGRATION 

 PATTERNS 



LENGTH-WEIGHT RELATIONSHIP 



All migrants processed from the fyke net 

 catches were weighed to the nearest 0.5 g. on 

 a triple-beam balance. The average weight 

 by age was computed for each day's sample, 

 and weight-gain curves were plotted by in- 

 spection (fig. 13). Each age class was treated 

 separately since there were three separate 

 size groups with little overlap, and combined 

 mean weights would be meaningless. The 



16 



I 10 20 



AueusT 



10 20 I 

 SEPT. OCT 



Figure 13.— Mean daily weights In 1957 of Brooks Lake 

 migrants. 



Although the nets were fished for five 24- 

 hour and five all-night periods to determine 

 the diurnal characteristics of the migration, 

 only the catches from May 30 to June 1 were 

 large enough to warrant study. 



Only newly hatched fry made any appre- 

 ciable daytime migration, and these could not 

 be effectively sampled with our coarse-mesh 

 nets. Personnel working near the outlet daily 

 observed only insignificant numbers of year- 

 lings or 2-year-olds during the day. A few 

 yearlings and 2-year-olds were captured dur- 

 ing the day in the five 24-hour sets. 



Catches on the nights of May 30-31 and 

 May 31 to June 1 (at the peak of the seasonal 

 migration) indicated that most fish left the 

 lake in the darkest hours of the night (fig. 14). 

 All three nets showed similar catch patterns 

 except that catches of net 2 in midstream 

 peaked slightly later than those from nets 1 

 and 3. 



21 



