FISHERY BULLETIN: VOL. 71. NO. 3 



2000 



2300 



I 



0200 

 I 



0500 



I 



0800 





v^AxmiKm'JKii 



III 



vimiaKi^nmimim 



t^y^y^y^y^AS>AM 



■y^y^^^y^^yf'y^y^y^^y^yi^ 



i^^^^^ 



Figure 2. — Fishing plan (nighttime) for study of dropout, 

 1965. Each unit of gear is identical; length of shaded area 

 refers to fishing time. 



units fished during the first and second 3-h 

 intervals. Loss of salmon from gill nets was 

 estimated by regression analysis based on data 

 from the catches in the various time periods (a 

 6-h catch vs. two 3-h catches). 



In the second experimental design, four units 

 of gill nets were set before daylight and fished 

 for various lengths of time until noon. Addi- 

 tional units were fished concurrently as shown 

 in Figure 3. In this design, we expected that in 

 the event fish did not enter gill nets fished dur- 

 ing daylight the dropout rate of fish caught 

 before daylight could be measured directly by 

 comparing catches of the gear units fished 

 before daybreak with those units fished for 

 various periods after daybreak. 



The indirect methods described above have 

 certain inherent limitations and potential 

 sources of error. These methods assume that: 

 a) each gill net is equally efficient; b) each net 

 fishes the same amount of time; and c) the 

 same number of fish are available to each unit 

 of net. 



0200 

 I L 



0500 



J I L 



0800 

 J I 



1200 



I 



I 



5 . 



Figure 3. — Fishing plan (night-day) for study of dropout, 

 1965. Each unit of gear is identical: length of shaded area 

 refers to fishing time. 



Additionally, it is not possible to differenti- 

 ate between losses due to dropout and losses 

 due to predators and carrion eaters.^ Thus, all 

 losses of salmon from gill nets must be attrib- 

 uted to dropout and predation. 



Direct Observations (1966-69) 



Loss of salmon in gill nets was estimated by 

 direct observation during experiments in 1966- 

 69. The experimental procedure was to patrol 

 the nets at night from small boats (when sea 

 conditions permitted), mark the locations of 

 gilled salmon, and determine their presence or 

 absence at later patrols and when the gear was 

 hauled in the morning. Portable spotlights 

 were used to observe the salmon at night; 

 reflected light made fish in the net visible. The 

 position of an enmeshed salmon was noted by 

 fastening a colored marker to the corkline 

 directly over the fish. 



The time and procedure are described 

 below; some variation in timing occured dur- 

 ing various sets. 



2000 Put out 24-36 shackles of gill net. 



2100 First patrol: Examine the nearest 7-14 

 shackles; mark positions of gilled 

 salmon with colored pins. 



2130 Second patrol: Repeat the patrol and 

 record the presence or absence of 

 salmon at locations marked at the 

 first patrol; mark locations of newly 

 observed salmon with pins of a 

 second color. 



2330 Third patrol: Repeat the above pro- 

 cedure, again marking locations of 

 newly observed salmon with pins of 

 a third color. 



0800 Haul the gill net string; record pres- 

 ence or absence of salmon for all 

 marked locations. 



Variations in timing and frequency of 

 observations occurred among years. For exam- 

 ple, in 1967, an additional observation at 0230 



■* Total loss due to predators and carrion eaters is 

 designated as loss due to predation in the remainmg part 

 of our report. 



848 



