88 



In if}iU and J99S Risk Managemeiii, the agenciss present data for the r«c«pf»ire of juvenile talmonidt 

 at McNary dam which were markeri and rdwEod at Rock Island Dam on the mid-ColumWa and ai 

 IJflle Gooie Dam on ihe Snake River. Tobies ] and 2 arc takeo directly frou) SpUl uikI 1993 Risk 

 Managcmau I have added one addliiuiinl tulumn orinAjmiailon to Uie lablej which jhowj lh« 

 lurbine flows (/.«., Average Flow Ie3$ Daily Averajje Spill) for the years examined in the a«enaei' 

 report The intent of the agencies in presenting those tables waj to demonatrare (hat there wu not a 

 majsive mortality resuiting ftom Hi«olved gm wpertaturation (DOS) during this period. Howcv«r, at 

 indicated in TabU J, most of the DOS levels nro below the US. nnvironmemal Piolwlion Ajjency ' 

 guideline of 1 10%. In ftict, the hlgliesi lolid tjas pressure (TCP) Is 1 13.7 %, which is sianlflcaiUly 

 below the ma.\lmum valuei of 120% . 125% recommended in Spilt aid J99S RiskManagemeni. 

 Similarly, in Table 2, the TOP leveb are also well below the aJiowabie levels recommended in S^HI 

 and 1995 Rixk Management. Conawjiif miy, the dissolved gat ralalioniWpa ihown in Tablea 1 and 2 

 have no relevance to the dissolved gai levels proposed for the Columbia and Snake RJverj by lliB alate 

 end tribal agenciea and which are preseni in UH»e rivers in 1993. 



The most Imponatit aspects of these daU are the relationships between spill, turbine flow, and the 

 proportion of recaptured fish The argiimenf advanced by Spill and 1995 Risk Managi/ituinf is that 

 high recapture proportions reflect higher smolt uutvivol In the sections of river examined (I.e., Rocli 

 Island Dain to McNary Dam &11J LiiiJo Oouse Dam to McNiry Dam). In Table 1 , It li clear thtt the 

 highest recapture proportion occurs in 19y3 when the spill is highest {I.e., 417 kefs). However, this is 

 also the year with the hiRJicst flow through the tuibines (/.? , 88.1 kcft). In fkct. the second highest 

 recapture proportion (0 4fifi) of<^irrpd in 1989 when spill (Tow was well boJow the 1993 spill level, but 

 turtjne flow waa the second hiffhen for the period {I.e., 70.0 koft) Clearly, because ofthe contrasting 

 results, ilieae data Oiil lo dciiioiiiiiiiatt thm npill produces the highest sinoh survival in these river 

 reaches. 



Table 1 also yields Important infcrmstion about siiidt survival, spill, and turbine flows. In the table, 

 1991 corresponds to the year of highest dnily avcrngc spill (56.6 kefi) with a corresponding recovery 

 piopuiliuii uro.32 fui thinouk aalmoji and 3'4J fbr steelhcad trout. Yet, In 1989, when splU levels 

 are about i/2 those of IVVl (/.<?, J) 1 kefs), the recover proportions are higher (',c., 0.342 for 

 chinoolc salmon and 0.367 for steelhead trout) Similarly, in 1992. when spill is less than 1/2 thai of 

 1991 (;> . 54 4 kefs), thf recovery proportion k ne"'!' src'ter lh«n that of 1991 (/.c, 0.34 for chinook 

 salmon ond 0.381 for Jteolhcod trout) nnd is the highest for stcclhcad trout shovrTi in the tabic. When 

 examined iij leiiiis uf turbine flows, seme yf the hijjhest recovery proportions shown In Table 2 

 correspond to the nighest turbine flows (120.6 kcli ill 1989 and 130.9kclSin 1991). Again, because 

 ofthe highly variable results, these data fail to dcmoiisirai? that spill is the optimal means offish 

 passage at dnms 



In the case yf adult sun'ival, Figure 1 shows piois of migrant sup-ival indices fbr two populations of 

 Snake River Chinook salmon tor the period 1980 through 1990. These data are from Table 3 of Spill 

 (uid 1995 Risk Mana^mcnl. which l:een included for reference. In the figure, it is seen that for Mar»h 

 Creek wild spring chinook, jurvivil was high when spill leva's ^vers 54 8 kefs. However, survival was 

 also high when spill levtls were only 8 2 kcfi When spill \c\zU were bctvfoon 27,3 aud 48.1 kcfa, 

 survival was less than that fbr either 54.8 or 8.2 kefs. 



