103 



combined- ef f ects objective is met for Snake River Chinook salmon 

 populations. The goal is that the 5 -year average population in 

 the years 2004-2008 is greater than that observed in 1986-1590. 

 According to the Dennis model, and assuming that 1! abundance 

 just meets the goal in 2008, and 2) tne overall trend in 

 abundance is neither decreasing nor increasing, the probability 

 of extinction in the next 100 years would be minimal for both 

 fail and spring/ summer chinook salmon. 



3ne feature of the proposed goal is that, because the populations 

 can be expected to continue to decline for a period of time 

 before implemented changes fully take effect, population 

 abundance will decline below target levels before rising to meet 

 the target at the end of the time frame. This means that the 

 analysis above is in one sense conservative, because it assumes 

 that the mean rate of population change in 2008 will be zero, 

 rather than the positive rate expected if the population follows 

 the expected trajectory. 



Dn the other hand, the temporary decline in abundance that would 

 be expected even if actions that meet the proposed goal were 

 implemented might increase the level of risk faced by the 

 population during this transition period. This risk is difficult 

 to evaluate using the Dennis model because that model assumes no 

 change in conditions that were obtained during the base period; 

 whereas, changing these conditions is the explicit purpose of 

 actions taken to avoid jeopardy to the species. As an 

 alternative, it is possible to calculate confidence intervals 

 around the population abundance trajectories shown in Table 2. 

 The lower 95% confidence bounds for abundance for spring/summer 

 and fall chinook salmon calculated around the population 

 trajectories in Table 1 are shown in Figure 1. This figure 

 provides some insight into the short-term demographic risk cf 

 extinction posed by the temporary population decline that is 

 expected even if actions are taken to meet the goal. As is 

 apparent from Figure 1, the lower bounds for both species are 

 substantially higher than one indicating little risk of 

 extinction during the period 1992-2008. 



The Dennis model should be used with caution, as it 1) does not 

 include any density dependence, 2) does not incorporate age- 

 structure, 3) does not compensate for effects of past management 

 (harvest rates, hydropower system operations, etc.! , and 4; may 

 not be applicable to very small populations (<100) . Conclusions 

 regarding extinction are thus far from exact, and must not be 

 construed to suggest that the populations will be out of danger 

 if mean change is zero over sixteen years. However, the general 

 indication is that the proposed standard, if achieved, would put 

 both populations in a much better position than they are now. 



As a result of these considerations" of genetic and demographic 

 concerns, NMFS has concluded that stabilizing the population at 



14 



