77 



standard that has protected aquatic life in the Columbia for many years does not promote 

 well-mformed decisionmaking. Our ability to respond to NMFS' proposal has been seriously 

 hampered by the shonness of the time penod allowed for public comment. 



2. The spill program will decrease salmoD survival, not increase it. NMFS 

 recognizes that the most accurate computer model of juvenile salmonid survival is the CRiSP 

 model, developed by University of Washington researcher Dr. James Anderson. Applying 

 the model to NMFS' spill program, Dr. Anderson reports that it will decrease the survival of 

 migrating spnng chinook salmon by at least 2.5% (and probably more) (Tab 6, at 3-4), 

 pnmanly because the spill program will decrease the percentage of salmon transpcrtsd 

 around the dams. A staff report by NMFS' own scientists offers an even higher estimate of 

 the decrease in survival than Dr. Anderson: they report a 4% decrease in survival. (Tab 1, 

 at 5 & Table 1 (decrease from 79% to 74% survival).) Dr. Don Chapman confirms that 

 "relaxed restnctions on gas supersaturation, to the extent that they lead to increased 

 voluntary spill, will do smolts no favor. The available data indicate that juveniles are better 

 off if left in barges." (Tab 7, at 4.) 



3. The state and tribal "1995 Spill and Risk Assessment" document cannot be 

 relied upon to support any increase in allowable TDG. Two NMF,S scientists who 

 reviewed the Risk Assessment document identified "major deficiencies" in it. (Tab 5, at 1.) 

 This conclusion is shared by former NMFS scientist Dr. Wes Ebel (Tab 8, at 2 & 



Exhibit A), Dr. Larry Fidler (Tab 9), Dr. Don Chapman (Tab 12), and Dr. James Anderson 

 (Tab 22), all of whom have advised the Oregon DEQ that the Risk Assessment should not be 

 relied upon. DEQ staff appears to share this assessment. (Tabs 18-21. ) Frankly, the state 



