RED SALMON COUNTS FROM OBSERVATION TOWERS 



361 



Stat ; however, wlien the lights were sufficiently 

 dimmed so that the salmon passed without hesita- 

 tion, the counts were not so accurate as desired. 

 Consequently, night counts were taken from only 

 one tower in 195" and not at ail in 1958. In 1959, 

 night counts were successfully obtained through- 

 out the season by using a single, bright spotlight • 

 focused outside of tlie panels from the top of the 

 towers. The fish passed in the dim area inshore 

 from the spotlight and could be seen satisfactorily 

 when crossing the background panels. Amber 

 and red lenses were used experimentally in 1957 

 to determine whether the salmon would tolerate 

 high illumination intensities wiien tlie lights were 

 colored. The colored lights caused sti-onger avoid- 

 ance reactions than dimmed white lights, and also 

 were less penetrating in waters that were moder- 

 ately turbid. 



The tallying procedure varied with the intensity 

 of the migration. Fish can be tallied separately 

 when the number passing is fewer than 250 per 

 minute. After the run exceeds this quantity, fish 

 must I)e tallied by estimating groups of 10. Heavy 

 runs requiring such grouped counts occurred only 

 near the peaks of the larger escapements, such as 

 the ones that occurred in 1956 and 1957. 



Method of Calculation 



The calculating procedure for estimating num- 

 bers of red salmon utilized the sample count as 

 an estimator of the number of fish passing in a 

 given unit of time. Therefore, with 10-minute 

 counts taken hourly, the estimate of numbers of 

 fish passing a tower is the product of multiplying 

 the count by six. Daily estimates are the sum of 

 calculated hourly estimates for all towers, and the 

 seasonal estimate is the sum of all daily estimates. 



Whenever systematic counts ceased at night or 

 wjien the waters were turbid with sediment, the 

 number of fish passing has been determined by 

 averaging tlie counts preceding and following the 

 periods with no counts. The estimate has been 

 calculated by multiplying this average by the ap- 

 propriate time factor. 



Results 



Estimates for Kvichak Kiver escapements have 

 varied widely from year to year, as would be ex- 

 pected from the cyclic nature of the runs and the 

 changing intensities of the commercial fishery. 



Tahle 1. — Kvichak daily fstimates of red salmon escape- 

 ments, ]955 through 1959 



' Estimate for July 26, 27, and 28 based on Irregular daytime samples. 

 ! Estimated late season migration. 



These totals for the past 5 years were: '250,546 

 in 1955; 9,-143,318 in 1956 ;' 2,842,810 in 1957: 

 534,785 in 1958 ; and 680,000 in 1959. 



Daily estimates for these years are listed in 

 table 1, and a graphic comparison of the daily 

 escapements thus obtained is presented in figure 

 7. The escapement of 1956 was the largest red 

 salmon run ever counted moving upriver in Alas- 

 ka. The escapement of 1955 undoubtedly was one 

 of the smallest noted in the history of the Kvichak 

 River system. 



The significant part of the Kvichak River 

 escapement has always passed the Igiugig towers 

 within approximately 3 weeks (table 2) . The peak 

 varied from July 8-9 in 1958 to as late as July 17 

 in 1956. The end of the runs, defined as the points 

 where the daily escapements are less than 1 percent 

 of the final total, are relatively unimportant to 

 the total estimates. Small numbers of fish usually 

 continue upriver even after the counts are termi- 

 nated. Earlier, before the counts commence in the 



