162 



cages offers an excellent surface for bubble formation, so gas levels may be 

 lower inside the cages than outside. The optimum procedure would be to 

 measure the gas level in each cage. 



The calibration of each gas instrument should be checked weekly. The 

 discrepancy between gas levels measured by your crew and those measured by 

 the USAGE indicate the need for this calibration. Gas levels in your cages 

 below Ice Harbor Dam on May 1 1 were measured at over 128% when Ken Witty 

 was present, while the 24 h high reported by the Fish Passage Center was only 

 122% for downstream of Ice Harbor. Such large discrepancies, In measurement 

 are critically important and should be immediately accounted for. 



We suggest that levels of dissolved oxygen, as well as total dissolved gasses, 

 should be monitored. The protocols listed in "Standard Methods" 1 8th Edition, 

 should be followed. 



STUDY DESIGN 



A statistical design, including specification of desired confidence intervals, 

 should be developed This design should specify the number of fish held in 

 each live cage and the number of replicate live cages Our experience with the 

 variability in expression of GBD signs indicates that replication is essential to 

 sound statistical inference. 



The net pens in use, particularly the volitional depth pen, are too small. Their 

 size is adequate for holding the fish, but is confining enough that it probably 

 affects fish behavior. Fish in the volitional depth pen could feel quite confined, 

 and may not freely range in depth as an unc/a^ged fish would. This problem 

 was aggravated by the tendency of the nver currents to partially collapse the 

 sides of the volitional depth pen. An internal frame, combined with greater width 

 and length of the pen, are needed. 



Resident fishes should be thoroughly examined at the time they are captured in 

 their natural environment, and a representative sample should be sacrificed for 

 internal examination. The live pens in v\^ich resident fish are held should be 

 substantially enlarged. The present small size of the live pens and the 

 confinement they impose probably causes resident fish (some quite large) to 

 hide at the bottom of the cage. Further, the trauma associated with their capture 

 and processing, followed by lack of acclimation to the live cage, is likely to lead 

 to abnormal behavior of the fish placed in live pens. 



Sampling of zooplankton and insect larvae should be added. This sampling is 

 easy to do and would serve as an indicator of effects on the food chain. These 

 organisms should be examined immediately after capture, and need not be held 



