136 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



The general features of the run are much simpler than in the case of the chinook 

 salmon just considered. There is a single, well defined peak formed by fish that enter 

 the lower part of the river late in June and early hi July. The first fish of this species 

 to appear in the records were in the Bonneville count for the second week hi May, 

 and a few were counted past the dam during the next 4 weeks. It was not until the 

 week ending June 18, however, that bluebacks began to show up in large numbers at 

 Bonneville. From that time on for the next 6 or 7 weeks bluebacks were a very important 

 element in the Bonneville count, but after the first of August their numbers dwindled 

 rapidly although a few were recorded as late as the second week in November 

 (table 7). It is to be noted that bluebacks did not appear in the catch of corresponding 

 weeks as early as they were recorded in the Bonneville count, nor as late (tables 

 20 and 21). This is probably due in part to the use of small-meshed nets especially 

 adapted for catching bluebacks while this species is most abundant; and also in part to 

 inaccurate identification in the Bonneville count. Evidence has been given above to 

 show that during the height of the blueback run there is a tendency to mistake the 

 smaller chinooks (grilse) for bluebacks. It seems not unreasonable to suppose that the 

 same error may also be made while bluebacks are scarce (or even entirely absent) 

 which would account for the very long "tails" to the time-frequency curve given in 

 figure 6, since these tails are formed almost entirely from the fish recorded in the 

 Bonneville count (tables 20 and 21). 



There are no complications due to spawning below Bonneville because in all proba- 

 bility all of the fish of this species spawn in streams tributary to lakes far above the upper 

 limits of commercial fishing. The bluebacks of the Columbia undoubtedly represent 

 a number of races, populations, or stocks, each breeding in its "home" lake basin; 

 but so far as the immediate problems are concerned they act from the mouth of the 

 Columbia to Cclilo as a unit run. Above Celilo not much is known of the repre- 

 sentative races; the available data consist chiefly of the counts made at the Rock 

 Island Dam. 



In preparing the modified tables for the blueback run the same rate of migration 

 up the river has been assumed as for chinooks; i. e., that fish that were hi Zones 1 

 and 2 in the first week would be found in Zones 3 to 5 the second week, at Bonneville 

 and between Bonneville and Celilo during the thud week, and at Rock Island the 

 fifth week. The data in tables 20 and 21 and figure 6 show that this assumption is 

 well justified, since the conspicuous peaks are made to coincide almost perfectly. 



These data have been analyzed by applying methods similar to those used in 

 the study of the chinook salmon. From the totals given hi table 20 it may be seen 

 that, for the entire season, the ratio of the estimated number of fish taken hi the 

 commercial fishery to the estimated escapement is approximately 3.32:1. In other 

 words, as shown by this estimate, over 3 fish are caught to 1 that escapes, passes 

 through the commercial fishing area, and becomes available on the spawning grounds 

 for perpetuating the run. The Rock Island count was 17,123. Multiplying this by 

 the ratio of catch to escapement gives 56,800 as an estimate of the number of fish 

 caught out of the populations normally spawning in the Columbia River above 

 Rock Island. Reconverting this to pounds by multiplying by 3 gives a total of 

 170,000 pounds caught that may be attributed to the runs spawning above Rock 

 Island. 



These figures are based upon a consideration of the catch and escapement for the 

 entire spring season up to and including August 25, and the total Rock Island count. 



