SALMON RUNS OF THE COLUMBIA RIVER IN 1938 145 



may be greater after the closed period than before. This may be due in part to a 

 change in the gear used on the lower river after the height of the fall run of chinooks 

 has passed. The silvers, being smaller fish, may be more readily caught with gill 

 nets of smaller mesh than is most effective for the larger chinooks. However this 

 may be, it seems reasonably certain that in 1938 there was a small but fairly well 

 separated run of silver salmon that entered the river late in August. 



The main part of the run of this species comes from about the middle of Septem- 

 ber to about the middle of November. There is some evidence of separate modes in 

 the run during this time, but it is not conclusive or even very strongly marked. The 

 height of the entire run in the lower river comes close to the middle of October. 



Chum salmon do not begin to enter the river much before the first of October. 

 From that date on the run gradually increases to a peak that comes about the first 

 week in November. After this the run as gradually decreases to terminate late in 

 December. There is no evidence of significant minor modes. As in the case of the 

 silver salmon, comparatively few of these fish pass Bonneville Dam, although a small 

 catch was recorded from Zone G. It is clear that the majority of the fish of this 

 species spawns in the tributaries below Bonneville Dam. 



SUMMARY 



1. Exceptional data are available for the study of the salmon runs of the Co- 

 lumbia River for 1938. For the first time the catch data for Oregon and Washington 

 were given in similar form so that they could be combined. As a result, the daily 

 catch in pounds of each species in each of 6 zones (corresponding to the parts of the 

 river bounding the 6 contiguous counties of Washington) is available for study. Co- 

 incident with this the Bonneville Dam was closed and fish ladders were constructed, 

 by means of which the fish surmounted the dam. On their way through the ladders 

 the fish were conducted through narrow passages and over white surfaces, and the 

 number of each species was recorded. There have also been available for study the 

 counts of salmon passing through the fish ladders at the Rock Island Dam, on the 

 upper Columbia River near Wenatchee, Wash. 



2. By using appropriate conversion factors the catch in pounds has been con- 

 verted into numbers of fish, so as to make these data directly comparable with the 

 counts at Bonneville and Rock Island dams. Tables have been prepared in which 

 are given (1) the weekly catch for each of 3 major areas representing natural groups 

 of wnes, (2) the total catch, (3) the Bonneville count, and (4) the Rock Island count. 

 For each major area the data have been appropriately "lagged" so that, as nearly as 

 possible, those for the same part of the run will lie on the same line as the table is 

 read from left to right. This lag has assumed that fish entering the river and to be 

 found in Zones 1 and 2 one week will be found in Zones 3, 4, and 5 the second week, at 

 Bonneville and in Zone 6 the third week, and at Rock Island the fifth week. These 

 modified tables form the basis for study and analysis. 



3. The general course of the run of each species is shown so far as possible by the 

 available data. The chinook salmon enter the river throughout most of the year, but 

 two quite distinct peaks are shown: One near the end of April, the so-called "spring" 

 run, and the other the latter half of August. There is a period of marked scarcity 

 during June and July. The blueback run is of much shorter duration, the main por- 

 tion lasting only 6 or 8 weeks and showing a marked peak toward the end of June. 

 Steelhead trout enter the river throughout the year but the chief run is during the 



