added to our understanding of the causes of trends in returns, and little contribution would be made 

 toward the refinement of the program. 



The program defines the doubling goal in terms of "the number of adults returning to the mouth 

 of the Columbia River plus the number of adults caught in the ocean." Under this approach, adult 

 returns to the river mouth would be estimated as described above, plus prior ocean catch of each 

 stock. By adding the estimated ocean catch, the effect of fluctuations in harvest rate on returns would 

 be taken into account, and would provide a truer measure of program progress. 



However, it would still not be possible to totally isolate the contribution of the program from other 

 factors, notably variation in the natural survival rate in the ocean. Even if the ocean survival rate could 

 be assumed to be more or less constant (excepting dramatic and obvious changes such as an El Nino 

 event), and progress toward the doubling goal could be attributed to the program, we would still not 

 know how the program affected returns. No information would be provided as to the efficacy of 

 different types of actions, our understanding of the system would not be increased, and no contribution 

 would be made to the refinement of the program. In addition, adding the unadjusted catch to the 

 estimate of return would consistently overestimate program production since some of the fish caught 

 would have died in any event from natural causes prior to reaching the river mouth. 



The measure could be improved by adjusting the estimates of ocean catch by the number of fish 

 that would have died in the ocean from natural causes. This measure is termed the "adult equivalent 

 run size" and is a true measure of the adult production from the basin or from a particular production 

 scenario. Adult equivalent run size is used in the Pacific Salmon Treaty process, and by the federal 

 courts in determining allocations between Indian and non-Indian fisheries. This measure would thus 

 more accurately reflect actual adult production and would be consistent with other coastwide 

 approaches. 



However, the other shortcomings of the previous method would apply equally to adult equivalent 

 run size as well. It is a fairly complex process that would not properly credit the program, and would 

 contribute little to our knowledge or ability to refine the program. 



2. Analytical Methods 



Observational methods, while attractive because of their intuitive appeal, suffer as measures of 

 progress because of their limited ability to increase our knowledge of the salmon and steelhead 

 resource. Observational indices only address stock abundance and provide only limited amounts of 

 information to isolate program effects or to refine the program. Analytical methods build on 

 observational methods. They attempt to increase the information content by integrating environmental 

 indices, research results, or monitoring data into mathematical expressions that are hypotheses 

 explaining the trends evidenced by observational data. These expressions are refined by testing 

 through research or monitoring programs. 



Analytical methods can be divided into at least two general categories that will be termed here 

 statistical and life-cycle approaches. Statistical methods can be used to discern relationships between 

 variables such as run size and flow during the outmiqration, number of spawners, or the relationship 

 between year-classes. A life cycle approach, on the other hand, uses a computer model as a 

 conceptual basis for explaining trends displayed by observational indices. The degree of relationship 

 between the model and the observed trends is the basis for refinement of knowledge using information 

 gleaned from directed research and monitoring programs. 



Statistical methods . A variety of statistical techniques are available to examine the relationship 

 between variables or to partition observed variation into component parts. For instance, the Oregon 

 Production Index (OP!) uses statistical correlation to provide harvest managers with an indication of the 



