139 



p. 30. Table 4 : 



Some values are incorrect. Ice Harbor (1968) was 14.5 percent not 32 percent. 32 

 percent was recorded for releases in the backroU of the turbine discharge. Bonneville first 

 powerhouse mortality estimate was 12-15 percent so 13.5 percent would be a more 

 appropriate value. The Lower Monumental data was for 1975, not 1972. This same 

 expenment indicated 27 percent mortality for steeihead passing over a standard spillway. It 

 is interesting that they chose not to use this value. The value quoted for Bonneville II of 18 

 percent is not appropriate. This was based on only one year of adult returns. The adult 

 returns were not sufficient from that one year of returns to conclude anything. The 

 appropriate data to use is the combined juvenile recoveries from all years. When this is done 

 the mortality ranges from 2-4 percent (Gilbreath et al., 1993 and Dawley et al., 1994). The 

 value of 18 percent for turbine mortality obtained in 1993 should not be used. The study 

 was designed mainly to determine if estimates of turbine, spill, and reservoir mortality could 

 be accurately estimated. Experimental difficulties in 1993 may have compromised this 

 estimate of turbine mortality. A revised, more accurate estimate will be available for the 

 1994 data. Not recorded is a turbine mortality 3.5 - 9.2 percent estimated by RMC 1994) at 

 Rocky Reach Dam and 5 percent estimated by RMC (1994) at Lower Granite Dam. 



P. 30. Par. 2. Line 1 : 



Smolt passage index. There are serious problems in using the smolt passage index to 

 develop reasonably accurate population estimates because the method does not account for 

 seasonal changes in FGE (fish guidance efficiency) which varies considerably through the 

 season and numbers collected also vary drastically depending on volume of spill. Assuming 

 a 1:1 ratio of spill volume to fish passage and a constant FGE for the season for each species 

 results in significant error. The smolt passage index is useful for comparison between or 

 among years, but could result in substantial errors, in estimating population at various 

 locations (dams) in the river. 



P. 43. Par. 3. Line 1 : 



In light of some of the errors (i.e. turbine mortality, FGE, and spillway mortality) 

 noted in values used in the risk analysis, I don't believe this is very accurate. 



P. 56. Par. 2. Depth Distnbution : 



Adults may remain at sufficient depth to compensate for fairiy high levels TDG, but 

 the fact remains that mortalities do occur when there are delays in migration and adults are 

 seeking fishway entrances. Even if there are minimal delays, adults must ascend t a 

 maximum of 6 feet of depth to enter and pass up the fish ladders. 



16 



