ABUNDANCE AND AGE OF KVICHAK RIVER RED SALMON SMOLTS 



309 



0.06:1.00, and this ratio was applied to the 

 catches obtained at site A before June 7 to esti- 

 mate the index catches for this early period. 

 The estimated index catches for 1956 ai-e shown 

 in appendix table 2. 



During each year's smolt migration .some 

 hours and days of fishing were missed because 

 of ice in the river or failure of the fyke-net 

 anchors. Estimates of the number of fish pass- 

 ing the net site during hours not fished have 

 been calculated on the basis of the average 

 catch of the preceding and following hours. 

 Estimates for days missed have been made by 

 averaging the catches of the preceding and fol- 

 lowing days. Estimates for fishing hours 

 missed have never exceeded 8 percent of the 

 season's catch, and estimates for days missed 

 have never exceeded 2.5 percent of the season's 

 catch. The maximum estimate of hours missed 

 was for the 1955 run and the maximum esti- 

 mate of days missed was for the 1957 run. 



During the peak of the 1959 migration, the 

 cod end was placed on the net for 5 minutes of 

 each 15 minutes fished and occasionally 5 min- 

 utes of each 30 minutes fished. The catches 

 were multiplied by 3 or 6, as appropriate, for 

 estimates of the 15- and 30-minute periods. 

 This subsampling reduced handling of the fish 

 and consequent injury to them. The estimates 

 from the intermittent fishing in 1959 are con- 

 sidered reliable because of the homogeneity of 

 catches noted in 1958 during periods of con- 

 tinuous net checking at the peak of migration. 

 To test this homogeneity for 1958, all combi- 

 nations of every third 5-minute catch (for the 

 estimate of the 15-minute periods of fishing in 

 1958) were compared with the total of all con- 

 secutive 5-minute catches (total of 98), and the 

 maximum error was found to be about ±0.5 

 percent. When each combination of every sixth 

 5-minute catch (for the estimate of the 30 min- 

 ute periods) was related to the total catch, the 

 maximum error was less than ±4.0 percent. 



Timing of Migration 



The timing of the Kvichak River smolt mi- 

 grations is depicted by the annual cumulative 

 catch curves (fig. 9). Each spring the migra- 

 tion started near the final day of ice flow in the 

 Kvichak River (table 6) and continued through 



10 20 



JULY 



Figure 9. — Cumulative daily smolt catches at the index 

 site in the Kvichak River, 1955 through 1959. 



June or early July. The 1958 smolt migration 

 is the earliest on record and corresponds to the 

 early breakup of lake ice for that year. 



Table 6. — Final day of ice flow in the Kvichak River 



Each year the peak period of smolt abundance 

 has been less than 2 weeks in duration (table 7). 

 In the future this short period of major abun- 

 dance can probably be used to reduce the effort 

 previously necessary for smolt enumeration. 

 Fyke netting can be started near the time of 

 ice breakup and be continued only through the 

 peak period of migration. 



Table 7. — Periods of peak catches of smolts in the 

 Kvichak River 



Hourly Index Catches 



The hours of largest smolt catch for each 

 year are shown graphically in figure 10. In 

 1955 and 1956, the catches were largest from 

 2300 hours to midnight; during 1957 and 1958 



