357 



It is worth noting that many of the same aggregation issues surrounding the 

 transportation evaluations are inherent to the smolt travel time-flow regime-survival 

 studies, where aggregation is deemed appropriate by C8FWA member analysts. For 

 example, aggregateing daily release group data and using average or median values for 

 weekly (or even annual) comparisons. In effect by rejecting subgroup aggregation, the 

 C8FWA reviewers give the appearance of adopting a "double standard" for their analyses 

 and technical studies. 



Unknown Bias Levels in Upstream Sample Collection: 



The Review Group correctly identified that the studies have not separated hatchery fish 

 and wild fish. The techniques employed do not allow separation of these fish at the time 

 ttiey are marked and ti-ansported. The reviewers also correctiy identified that retums to 

 the dams do not separately measure the effect of ti-ansportation or sunwal to the 

 spawning grounds. The techniques applied within tine studies do no provide sufficient 

 numbers of retijrning adults to reliably assess retums to the spawning grounds (a factor 

 sfongly asserted by NMFS scientists). The research designs were never intended to 

 make such observations. 



But perhaps of greater importance for analysis purposes is the issue of representative 

 sampling between wild and hatchery fish and sampling procedures. The collection 

 faciiities-either hatcheries or upstream surveyArap sites— form the sample frame for the 

 total population of treatment and control fish. If there is not substantial assurance that 

 a representative sample of wild fish are being (or could be) counted at tiie upstream 

 survey/trap sites, then wild count comparisons to a hatchery collection site (where 

 hatchery numbers are directiy concentrated) is meaningless. It is not clear from the 

 CBFWA review whether the criterion for sample representativeness was adequately taken 

 into account 



Unsubstantiated Wild Fish Hypothesis: 



Given their analytical approach, tine CBFWA Review Group suggest tiiat wild salmon do 

 not survive as well as hatchery fish under transportation, or at least are more unlikely to 

 return to spawning sites. The review report fails to provide either the adequate empirical 

 documentation (given tiie comments above) or to offer reasonable support from the 

 technical literature to substantiate wfiy wild fish would be less responsive to barge 

 transportation than hatchery fish. Also, it is undear-or not fully unexplained-why 

 transportation effects would have a greater impact on spawning disorientation for wild 

 fish, as opposed to hatchery fish. 



Existing estimates of wild to hatchery fish ratios have varied greatly since the completion 

 of Lower Granite Dam. However, during ttie period of analysis by the Review Group- 

 1S86-1992-wild to hatchery fish ratios based on Lower Granite Dam counts and available 

 wild fish estimates have increased, not dedined. In 1985 and 1986 for spring and 



