Kope and Botsford: Recruitment of Oncorhynchus tshawytscha in central California 



263 



at the 0. 1 level is half the number expected if all series 

 were independent (i.e., 4 versus 8 out of 80), and the 

 only correlations significant at the 0.05 level using the 

 conservative test— a positive effect of diversions on 

 Feather River fish— has no apparent physical or bio- 

 logical explanation, because diversions during down- 

 stream migration would be expected to have a negative 

 effect on survival. We consider this one a spurious cor- 

 relation. Lack of correlation with delta diversions and 

 total delta outflow may be due in part to the fact that 

 hatcheries on the Feather and American Rivers have 

 released most of their smolts below the delta since the 

 mid-1970s. 



Marine influences on spawners were assessed by 

 computing correlations between the same population 

 variables and upwelling index, sea level height, and sea 

 surface temperature. The resulting correlations (Table 

 2) are significant at the 0.05 level only for adult spawn- 

 ers and reci'uitment to the adult spawner population. 

 They show a negative influence of conditions associated 

 with ENSO events (i.e., higher ocean temperatures, 

 higher sea levels) in the third spring and summer, im- 

 mediately before the adults enter the rivers to spawn. 



There is also a positive correlation with upwelling in- 

 dex in the first spring and summer, during which 

 smolts migrate downriver, through the delta, and enter 

 the ocean. 



We also tested for marine and freshwater influences 

 on catch data from both the commercial and the recrea- 

 tional fishery. We attempted to remove the effects of 

 fluctuations in effort from, the catch by computing com- 

 mercial catch-per-unit-effort and recreational catch- 

 per-unit-effort using the marginal effort data. We com- 

 puted recruitment estimates from deconvolution of 

 each of these. There is a substantial number of correla- 

 tions significant at the 0.1 level using the standard test, 

 but none reflect freshwater influences (Table 3). Of the 

 correlations that are significant at the 0.05 level, two 

 are with temperature in the third summer, and the re- 

 maining four are in the first marine year, three of them 

 being with upwelling index and the other with sea level 

 height. None of the catch-per-unit-effort variables were 

 significantly correlated with any of the environmental 

 variables using the conservative test. 



These results indicate that oceanographic conditions 

 during the first and third years of life may influence 



