DATE OF TAGGING 



by the counters (table 10). Many chinooks 

 (obviously jacks) were called sockeye. Jacks are 

 precocious male chinooks, and are similar in size 

 to sockeye. The error in mistaking sockeye for 

 chinooks was relatively small. No tagged steel- 

 head were reported by the counters at Rock Island 

 Dam; one tagged steelhead, however, was observed 

 upstream at Tumwater Dam. Similar errors were 

 made in 1955 and 1956, although the fish counters 

 were asked to keep the counting boards at the 

 minimum depth consistent with efficient fish 

 passage. The magnitude of error was not deter- 

 mined for these years. 



2. The disproportionate returns of the different 

 types of tags used gave evidence of errors in tag 

 identity at Rock Island Dam. A far greater 

 percentage of bar tags was observed at Rock Is- 

 land in 1954 and 1955 than of spaghetti tags (table 

 11). Upstream, however, the percentage re- 

 turns of the two tags were approximately the same 

 for both years. The streamer tag was substituted 



Figure 6. — Numbers of sockeye tagged below Rock 

 Island Dam, and recoveries of tagged fish upstream, 

 expressed as percentages of the number tagged each day, 

 1954. 



identifying accurately the various tags and species 

 of fish as the fish crossed counting boards. The 

 tags were more difficult to identify at Rock Island 

 than at upstream points because of the need for 

 counting large numbers of fish while simultane- 

 ously identifying tags, and because of the speed the 

 fish moved across the counting boards. The 

 misidentification of both tags and species was a 

 source of error in the Rock Island tag return data 

 as exemplified in the following illustrations: 



1. In some of the 1954 experiments, fish counters 

 recorded as tagged species which had not been 

 tagged. In other experiments, they recorded 

 greater numbers of tag observations of species than 

 fish tagged, and on several occasions they counted 

 fewer jack chinook in the traps than were tallied 

 during subsequent tagging. We revised our 

 tagging procedure for some experiments in 1954 

 to determine the magnitude of error in species 

 identification. Instead of tagging all species with 

 one color combination as was usually done, we 

 tagged the various species with different color 

 combinations and noted the returns as identified 



10 20 30 

 JULY 



DATE OF TAGGING 



FuiURE 7. — Numbers of sockeye tagged below Rock 

 Island Dam, and recoveries of tagged fish upstream, 

 expressed as percentages of the number tagged each 

 day, 1955. 



354 



U.S. FISH AND WILDLIFE SERVICE 



