SALMON RUNS OF THE COLUMBIA RIVER IN 1938 125 



with the original assumption. Whether, without the closed period, the peak in the 

 Bonneville count would have come in the week ending September 10 is perhaps some- 

 what doubtful, and no method has occurred to us whereby that can be independently 

 determined. From the total run (table 12) this would seem to be a reasonable infer- 

 ence, but it has been based on the assumption that 2 weeks are required for the journey 

 from the mouth of the river to Bonneville. 



In passing, it should be emphasized for future use in similar situations that the 

 effect of the closed period has been to so increase the Bonneville count immediately 

 following the beginning of the closed period that it has the effect of shifting the peak 

 of the count upward. This would be true even if the final week of the open period 

 had consisted of 6 days instead of 4 days of fishing. In general, the incidence of a 

 closed period will increase the escapement in the following weeks, but in this case 

 the peak of the run happens to coincide so closely with the beginning of the closed 

 period (probably actually preceding it on the lower river) that the effect is to shift 

 the peak of the escapement upward. Also, in this particular case, the fact that the 

 last week of the open season contained only 4 fishing days had the effect of appar- 

 ently shifting the peak of the catch downward. The combined result was an apparent 

 lag of 3 instead of 2 weeks between the peak of the catch in Zones 1 and 2 and the 

 peak of the count at Bonneville. Similarly, at the beginning of an open period there 

 will be the reverse tendency for the peak of the escapement to be shifted downward 

 and the peak of the catch to be shifted upward. Doubtless the peak of the Bonne- 

 ville count that occurs during the week corresponding to that of April 30 has been 

 so modified. Actually this count was made during the week ending May 14, and the 

 fish passing Bonneville during that week were doubtless partly through Zones 1 and 

 2 before the fishing season opened on May 1. These rather confusing effects are, of 

 course, due to the complementary relationship existing between the catch and the 

 count at Bonneville. 



Related to these phenomena is the fact that there uppears to have been some 

 delay in the passage of fish through Zone 6 following the peak of the run and the 

 closed season. This is shown particularly by the fact that during the weeks ending 

 September 10 to October 15 (almost the entire effective fall season) the catch above 

 Bonneville exceeded the Bonneville count. However, we believe that this does not 

 indicate a general lower average rate of travel, but is due, rather, to the combined 

 influence of individual variation in the rate of travel and a constant reduction in the 

 number of fish passing Bonneville. The anomaly, then, of the existence over a number 

 of weeks of a greater catch above Bonneville than count over the dam is closely 

 related to the fact that the peak of the escapement curve is shifted to an earlier date 

 by the incidence of an open season. 



THE JUNE-JULY RUN 



As previously mentioned, the June-July run of chinooks is poor compared witli 

 that in May or August, and it is rather generally thought that the populations form- 

 ing this part of the run are the most seriously depleted of any. Some evidence of 

 this was developed at the time the original study was made, but was not included 

 in the original report. It has seemed worth while to pursue the investigation further. 



As bearing on the extent to which the June-July run has been depleted, we have 

 examined data secured through the cooperation of the Columbia River Packers Asso- 

 ciation. These data are in the form of reports of daily deliveries to this company 



449668 — 42 * 



