0.15 0.25 0.35 0.45 0.55 



CV of Egg-to-Adult Survival 



Figure 11 — Effect of the coefficient of variation (CV) of egg-to- 

 adult survival on the mean number of spawners in year 30 and 

 the CV of spawners. 



MODEL CALIBRATION 



Proper calibration of a model such as the SLCM is always problematic. 

 Rarely, if ever, is there sufficient information to provide independent esti- 

 mates of all model parameters. For many stocks, the most readily avail- 

 able information likely will be that collected from hatchery fish, such as 

 sex ratios and fecundity information. Given the large number of hatcher- 

 ies within the Pacific Northwest, reliable information on most stocks or on 

 stocks similar to those in question should be available. Users should be 

 able to obtain fairly accurate estimates of the average potential egg produc- 

 tion from a given number of spawners and estimate a variance. Howell 

 and others (1985) and the recently completed subbasin production plans 

 (Columbia Basin Fish and Wildlife Authority 1990) are useful sources of 

 this type of information. 



The second most readily available source of information is CWT data. 

 The Pacific Marine Fisheries Commission maintains a comprehensive data 

 base on tagged fish releases and recoveries for Columbia River and other 

 Northwestern stocks. Further information on using CWT data to estimate 

 harvest and escapement parameters is provided in appendix B. Two as- 

 pects of CWT data should be remembered: (1) most CWT data are from 

 hatchery fish, and (2) the ratio of tags returned to the total number of 

 tagged fish released generally is extremely small. Because most data are 

 from hatchery fish, caution is called for when extrapolating data to natu- 

 rally produced fish. Because the number of returns is so small, acute bi- 

 ases in parameter estimates might arise because of uneven reporting rates 



20 



