In correlating the above index figures of the time of run in the 

 commercial fishery idth the time of run at Bonneville it is necessary to 

 note that the earlier the run in the commercial catch for any year the 

 larger the fit'ure; conversely the earlier the count at Bonneville, the 

 smaller the figure. Thus if the two variables are positively correlated 

 the correlation coefficient using the designated indices will be negative. 

 Using the figures from Table 2 and the list above, a correlation of r ::: 

 -,58ii was found, which is significant at the 5-percent level. 



The correlation between the size of run and the time of run each 

 year at Bonneville was not significant, nor was a significant correlation 

 coefficient found between size of run at Bonneville and size of commercial 

 catch each year. Since the commercial catch of shad on the Columbia is 

 incidental and supplemental to the salmon fishery (Ore[:on Fish Commission 

 1950), the size of catch probably does not represent the population present. 



The catch of shad above Bonneville Dam, most of which is caught 

 at Celilo Falls, 60 miles above, is directly proportional to the numbers 

 passing Bonneville D-tm. A significant correlation was obtained between 

 the shad catch for zone 6 (iirhich includes Celilo Falls) (Oregon Fish 

 Conmission 19^0) and the Bonneville shad counts for each year from 1938 

 through 19ii9. 



It can be concluded from the above correlation analyses that flow 

 and temperature do affect the times of entry of the shad into the river 

 and of passage through the Bonneville fishways. The river discharge 

 appears to be the more important of the two factors, 



Bradford Island Ladder Counts Compared 

 with Washington Shore Ladder Counts 



The total number of shad counted in the Washington Shore ladder 

 for all years from 1938 through 19^0 was 90,937; during the same period 

 167,285 were counted through the Bradford Island ladder. The Bradford 

 ladder has entrances in both channels of the river (fig. 1). If these 

 two entrances are considered as separate ladders and equal shares of the 

 Bradford count are assigned to each, their counts do not differ greatly 

 from that of the Washington Shore ladder. 



In some years (1939> 19U0, and I9U6) more shad were counted in 

 the Washington Shore ladder than in the Bradford ladder. No reason for 

 this has been ascertained. Several factors may influence the counts at 

 each ladder. When the forebay level changes, it is necessary to adjust 

 the weir heights between the counting gate and the water level behind the 

 dam. If the forebay level fluctuates rapidly the efficiency of the ladder 

 may be impaired for several days. Other things, such as repairs to the 

 ladders, rain-squalls, people visiting, and amount of sunlight and shadow, 

 may also affect the operation of the fishways and the movements of the fish. 



Ih 



