104 



NMFS concluded that it was appropriate to seek an operation that would 

 result in the EPA criteria of 110 percent being exceeded primarily because 

 of: (1) the abUity of fish in a river environment to compensate 

 hydrostatically for the effects of dissolved gas supersaturation, and (2) the 

 daily fluctuation in levels of dissolved gas througnout most of the river. In 

 a river environment, depth of migration reduces total dissolved gas effects 

 on migrants. Each meter of depth provides pressure compensation equal to 

 a 10 percent reduction in total dissolved gas. Shew et al. (Undated) and 

 Turner et al. (1984b) noted through tunnel studies that net entiy rates 

 through McNary and Bonneville Dam ladder entrance tunnels were highest 

 for the deepest (3.4m) tunnels. Other studies indicate that adult and luve- 

 nile salmon tend to spend most of their time at or below one meter of depth 

 (Smith 1974). Blahm (1975) concluded that shallow water tests were "not 

 representative of all river conditions that directly relate to mortality of ju- 

 venile salmon and trout in the Colvunbia River." In deep tank tests, 

 salmonids exposed to 115 percent TDG levels did not emerience significant 

 mortality until exposure time exceeded approximately 60 days (Dawley et 

 al. 1976). 



NMFS also concluded that it was not appropriate as an initial interim 

 level to seek an operation that would result in chronic exposure to TDG 

 level of 120 percent, as recommended by the states and tribes. In general, 

 chronic exposure to TDG levels of 120 percent with hydrostatic compensa- 

 tion does not cause significant mortality until exposure time exceeds 40 

 days (Dawley et al. 1976). This is generally more time than it takes Snake 

 River juvenile and adult migrants to travel between Lower Granite and 

 Bonneville Dam. Nevertheless, NMFS concluded that the more conservative 

 level of 115 percent is appropriate because of concerns about the potential 

 sublethal effects of gas bubble trauma. The state and tribal report on "Spill 

 and 1995 Risk Management" summarized the studies showing evidence 

 that swimming performance, growth and blood chemistry are affected by 

 high dissolved gas levels. The report correctly states that it is only inferen- 

 tial that these symptoms may result in susceptibility to predation, disease 

 and delay. In fact, studies conducted in 1993 and 1994 by the National Bio- 

 logical Service indicated that juvenile chinook salmon that have been ex- 

 posed for eight hours to high TDG (and exhibiting microscopic signs of gas 

 Dubble trauma) are no more vulnerable to northern squaw fish predation 

 than control fish that had been held in equilibrated water (Mesa and War- 

 ren, in review). Ultimately the analysis in the state and tribal report did 

 not assume any level of mortality as a result of these sublethal effects. 



NMFS concludes that the impairments to migrating fish as a result of the 

 sublethal effects of dissolved gas may be sufiiciently grave to warrant cau- 

 tion in setting long term exposure levels above 110 percent. In particular, 

 long term exposure to levels in excess of 110 percent decrease swimming 

 ability (Dawley and Ebel, 1975); fish stressed with high levels of dissolvea 

 gas have been reported to have less swimming stamina (Dawley et al., 

 1975); and gas bubbles in the lateral line can impair sensory ability. In ad- 

 dition, eilthough fish in deep tank studies are less affected by high levels 

 of TDG than fish in shallow tanks, some mortalities still occur despite.a 

 water depth that is apparently adequate for protection. There is no evi- 

 dence that fish can "sense" TDG supersaturated water and deliberately 

 sound to compensate. 



At specific projects where specific levels of spill, particularly daytime spill 

 have been shown to be detrimental to fish passage, timing and/or amounts 

 of spill may have to be adjusted (for specific detaus see NMFS 1994b). Spill 

 may also be limited at projects where it can be demonstrated that spill may 

 be detrimental to system spill allocation. One such project is John Day 

 Dam, where very low amounts of spill result in very high TDG levels. These 

 high TDG levels then limit the amount of spill possible at dams down- 

 stream. For instance, by reducing spUl by 10 to 20 kefs at John Day Dam, 

 it may be possible to increase spill at The Dalles or Bonneville Dams by 

 20 to 40 kefs. The exact relationship will need to be developed through in- 

 season spill/TDG testing. The limitation of spill may also apply at The 

 Dalles Dam to minimize the passage of spilled flow and fish over the high 

 predation risk area in the shoals below the dam (see specific details in 

 NMFS (1994b). The details regarding this limitation will be decided in-sea- 

 son through consultation with predation experts and will likely depend on 

 ambient flow and the spill levels obtainable under the TDG limitations. In 

 1995, spill at Ice Harbor, The Dalles, and John Day Dams may be modified 



