for the spaghetti tag in 1956, and, as observed, 

 returns were similar at Rock Island Dam and 

 upstream for the two tags. It is evident, there- 

 fore, that the counters at Rock Island Dam did 



Table 10.- 



- Errors in species identification at Rock Island 

 Dam in 1954 



1 The color and tag combination used here was also used 15 days previous. 

 Some returns may have been attributed to the previous experiment. 



2 The total of 79 fish observed is an obvious error in tag identity. 



~! 1 I 



10 2 



JULY 



DATE OF TAGGING 



Figure 8. — Numbers of sockeye tagged below Rock Island 

 Dam, and recoveries of tagged fish upstream, expressed 

 as percentages of the number tagged each day, 1956. 



~ 450 



r- 



UJ 



Hi 



u.400 



o 



O350 

 o 



LU 



in 

 ^300 



o 



to 



§2 50 

 < 



O200 



x 



S 150 



o 



_l 



LL. 



100 



LEFT BANK 



Jl 



RIGHT BANK 



10 20 30 10 20 30 



JULY AUGUST 



DATE OF TAGGING 



Figure 9. — Recoveries of tagged sockeye upstream ex- 

 pressed as percentages of the number tagged at each 

 bank below, compared with water flow, 1954. 



not discriminate between the tags in 1954 and 

 1955. 



These errors in species and tag identification 

 thus affected the reliability of the tag return data 

 at Rock Island Dam, and the data must be used 

 with caution. 



Sockeye and Chinook 



The results of the fish counters' observations of 

 tagged fish (including the few steelhead tagged) 

 crossing the counting boards for the four seasons 

 of tagging are shown in table 12. Although the 

 salmon were released below the dam at various 

 places on both sides of the river, most of the tags 

 were observed at the left ladder. Comparatively 

 few were recorded at the right ladder, even though 

 roughly half of the salmon were released a short 

 distance away. The counters observed 85 percent 



LOSS AND DELAY OF SALMON PASSING ROCK ISLAND DAM 



355 



