SALMON RUNS OF THE COLUMBIA RIVER IN 1938 143 



might have been the result of misidentificatiou of this species in the Bonneville count — 

 steelheads being mistaken for the much more numerous chinooks. In order to test 

 the possibility of such misidentification on a large scale in the Bonneville count, a 

 study was made of the ratios of the number of steelheads to chinooks in (1) the catch 

 below Bonneville, (2) the Bonneville count, and (3) the catch above Bonneville for 

 each week over the period beginning June 5 and ending October 29. It is to be 

 expected that such series of ratios would vary over the entire period with the rela- 

 tive numbers of fish of the 2 species, but the general trends of the ratios should be 

 similar in the 3 localities in the absence of disturbing factors — such as misidentifica- 

 tion in the Bonneville count. Figure 8 is a graph of these ratios wherein ordinary 

 arithmetic coordinates are used, since the absolute values are the significant ones. 

 It is apparent from this that the trends are very similar in the 3 localities; which is 

 evidence that the identification at Bonneville was sufficiently accurate and probably 

 was not responsible for the anomalous fact that more fish were recorded in tlie com- 

 mercial catch above Bonneville than were counted over the dam. 



The data thus graphed are interesting in themselves in addition to their bearing 

 on this particular problem. It is quite obvious that, in numbers of fish, the steel- 

 heads approach the chinooks and, during the June-July period when chinooks are 

 few, greatly exceed them. It is chiefly during the peak of the chinook run in August 

 and September that the ratio is down to about 1 :5 in the catch below Bonneville 

 and the Bonneville count. The parallelism in the 3 trends up to about the middle 

 of September is quite striking and is supporting evidence that, for this part of the 

 run, the assumed rate of travel is satisfactory. 



SILVER AND CHUM SALMON 



As mentioned in the introduction, the purposes of the original report by Calkins, 

 Durand, and Rich were such that consideration of the catches of silver and chum 

 salmon was not important. In this revision, however, it is pertinent to include the 

 data available on these 2 species, and to examine these for whatever light they may 

 throw upon the characteristics of the runs. The general features of the runs of silvers 

 and chums are so similar that it is convenient to treat them together. 



The data for these species are given in modified form in tables 25 and 26. In 

 converting poundage to numbers of fish an average weight of 10 pounds per fish has 

 been used for both species. The same rate of migration up the river has been used as 

 with the other species, although the rate of migration of both Isilvers ,and chums is 

 more doubtful and of far loss significance than in the case of the other species. There 

 is, however, no good evidence that the rate of travel is any different in the case of 

 these 2 species than in the others, although the obvious irregularities in the time at 

 which the main portion of the catches is made in the different zones (tables 4 and 5) 

 lead one to suspect that the rates of travel of these species may be somewhat different. 

 This is a matter that should be investigated, but it is necessary for the present to 

 assume the same rate of travel — which has been done in preparing the modified tables. 



