SALMON RUNS OF THE COLUMBIA RIVER IN 1938 123 



is only 0.72 percent of the escapement. The catch that may be attributed to Rock 

 Island is, therefore, 1,500 fish with an aggregate weight of 40,600 pounds. 



The data last presented (for the period July 31 to August 27) are similar to those 

 given in the original report and are presented here chiefly for comparison with those 

 that follow. We have already stated that this is not a natural subdivision of the run 

 and that, properly, the period from July 31 to the end of the year should be treated 

 as a unit. This larger period takes in the major peak of abundance that occurs in 

 late August and early September and includes completely the closed period, August 25 

 to September 10, and all catches that may be referred to the stocks of fish affected by 

 the closed season. The total catch was recorded as 8,326,000 pounds, which we esti- 

 mate included some 318,000 fish. In contrast to the other selected periods, the catch 

 above Bonneville Dam forms a large part of the total and it is of interest to note 

 (tables 1, 12, and 13) that the major part of this catch above Bonneville takes place 

 after the closed period. The total catch during the fall season alone was 2,685,000 

 pounds (109,000 fish), of which over half, 1,395,000 pounds (55,000 fish) were taken 

 above Bonneville. It is obvious that one important result of the closed period is to 

 permit enough fish to escape the fishery on the lower river so that upwards of a million 

 pounds may be taken above Bonneville Dam. 



The Bonneville count during the period July 31 to the end of the year was 229,000 

 fish. The net escapement (Bonneville count less the catch above Bonneville) was, 

 therefore, approximately 175,000 fish. The ratio of catch to escapement is 1.9:1, 

 which, while still high, is much less than that duiing the earlier periods. It is to be 

 noted, however, that this ratio is considerably higher than that for the month of 

 August, when the ratio is 1.4:1. This was one of the results of treating the period 

 from July 31 to August 27 as a unit. The facts that thi> period is not a natural 

 subdivision of the run and that the count at Bonneville for the period is undoubtedly 

 influenced by the incidence of the dosed season on August 25 have resulted in this and 

 other differences between the data for the month of August and those for the entire 

 period of the fall run. 



The Rock Island count for the period corresponding to that from July 31 to 

 the end of the year was 1,879, or 1.1 percent of the estimated net escapement. Taking 

 this as the percentage of Rock Island fish in the run as a Whole, the da'tcn that may 

 be attributed to the Rock Island runs is estimated at 3,500 fish, or 91,500 pounds. 

 This is to be compared with an estimate of 1,500 fish of an aggregate weight of lo.iino 

 pounds for the month of August. 



Table 15 presents the more significant figures bearing On fchfe absolute and relative 

 importance of the Rock Island runs of Chinook salmon. There are given not only the 

 figures obtained through the basis of estimate adopted in this report, but also, for com- 

 parison, those obtained through the two bases used in the original report by Calkins, 

 Durand, and Rich. (The estimates given here for the full season on the bases used 

 in the original report were not, of course, given in that report, which treated the catch 

 only up to August 25.) It is apparent that, in general, the results of all three pro- 

 cedures are of the same order of magnitude so that one may assume with some confi- 

 dence that no gross errors have been introduced. Although we believe that the 

 estimates based on the average weights obtained in 1939 are the most accurate, and 

 should certainly be used for detailed study of parts of the run, it is clear that simpler 

 methods will give approximate results of real value. 



