FISHERY BULLETIN: VOL. 70, NO. 1 



be taken in a larger floating trap positioned at 

 the upstream end of the ladder. 



A conventional 1,000-gal (3.79-m3) tank truck 

 transported sockeye salmon in 1964 and 1965 

 and Chinook salmon in 1965 but was not available 

 for the 1964 chinook salmon experiments. In- 

 stead, we used 1.2 m by 1.2 m by 1.1 m plywood 

 boxes, equipped with aeration systems and 

 mounted on -Vfton (680-kg) trucks. These units 

 were suitable for transporting fish to the adja- 

 cent left-bank release site but inadequate for 

 moving more than seven or eight fish per unit 

 to the opposite bank via Wenatchee, Wash., — a 

 48-km trip that took about 1/2 hr. Because of 

 this limitation, only one-half as many chinook 

 salmon were released on the right bank as on 

 the left bank in 1964. 



Each batch of fish liberated was distinctively 

 marked. Several types and colors of tags were 

 used. Petersen plastic disks were used either 

 alone or in combination with plastic bars and 

 vinyl streamers. Nickel pins, inserted through 

 the body just below the dorsal fin, provided the 

 attachment. Tags were applied in pairs, so that 

 the same color and type of tag showed on both 

 sides of the fish. Tagging time seldom exceeded 

 30 sec per fish. 



ARTIFICIAL MANIPULATION OF SPILL 

 PATTERN, AUGUST 3-5, 1964 



The spill pattern throughout most of the 1964 

 and 1965 tagging was developed from results of 

 experiments at Rock Island Dam in 1958 (Leman 

 and Paulik, 1966) with subsequent refinements 

 from a model study in 1964 and 1965 (Ward, 

 1965, see footnote 5). On August 3-5, 1964, 

 however, gates 16 to 18 (adjacent to the center 

 ladder) were closed and gates 34, 36, and 37 

 (adjacent to the right-bank ladder) were opened. 

 This departure from the recommended spill pat- 

 tern was undertaken to measure its eflfect on the 

 passage of tagged fish over the dam. 



TAG OBSERVATION AND 

 DATA RECORDING 



Four steps were taken to insure the accuracy 

 of the tag observations: First, hydroscopes 



(floating "windows") were installed over the 

 counting boards on the right- and left-bank lad- 

 ders to suppress glare and surface disturbance. 

 Second, all fish counters were tested for color 

 blindness. Third, samples of tags were mounted 

 on the tally boards to facilitate instant recogni- 

 tion and recording of the tags. Fourth, fish 

 counters were systematically rotated between 

 ladders to distribute any bias by the counters be- 

 tween the ladders. 



The gates in the fish ladders were open and 

 the passing fish were counted 16 hr a day — 

 5:00 AM to 9:00 PM — during these experiments. 

 The half-day units used to measure travel times 

 were adapted to the counting schedule. Fish 

 observed during the same 8-hr period in which 

 they were released were assigned a travel time 

 of one-quarter day or 0.5 half-day. Fish re- 

 leased just before noon (as most were) were 

 given a travel time of 0.5 half-day, if observed 

 the same day, or a time of 1.0 half-day if ob- 

 served the next morning. Tag observations 

 were grouped by tag combination, ladder, and 

 travel time in half days. Data were punched 

 on IBM cards — one card containing the release 

 and recovery data for each fish. The numbers 

 of tagged salmon released below Rock Island 

 Dam and later observed passing the dam in 

 1954, 1955, 1964, and 1965 are summarized in 

 Table 1. 



Table 1. — Numbers of salmon that were tagged and re- 

 leased below Rock Island Dam in 1954, 1955, 1964, and 

 1965, and the numbers and percentages of tagged fish 

 that were later observed passing over the dam's fish 

 ladders.^ 



Species of 

 salmon 



Year and ^released l^'elow Tagged fish observed 

 dam passing dam 



' The numbers of spring chinook salmon released in 1954 and 1955 

 differ from those reported by French and Wahle (1966). They used the 

 July 13 date suggested by Fish and Hanavan (1948) as the termination 

 of the spring run and the beginning of the summer run. We used the 

 scale method described by Koo and Isarankura (1967) to determine that 

 the dotes of least overlap between the two races were July 30, 1954, and 

 July 8, 1955. We believe that our separations, based on the more recent 

 study, are the more accurate. 



130 



