and successive spring population estimates were averaged to obtain a 

 mean flow. These periods were further divided into irrigation and 

 nonirrigation seasons and a mean flow obtained for each. Between 

 successive spring estimates, April 15 through October 15 was termed 

 the irrigation season. The remaining period was the nonirrigation 

 season. Between fall estimates, April 15 to the time of the fall 

 estimate was the irrigation season. The nonirrigation season occurred 

 from the time of the preceding fall estimate through April 14. 



The WSP (Water Surface Profile) computer program of the Bureau of 

 Reclamation was used to predict hydraulic values at various flows. 

 The procedures followed are outlined by Spence (1975) and Dooley (1976). 



The percentage of willow cover along the banks of the Beaverhead 

 River was assessed using color infra-red photography and a stero- 

 comparator. Photogrammetry work was done by graduate students at 

 Colorado State University. Sinuosity and section length were determined 

 from aerial photographs using computer facilities of the Civil 

 Engineering Department, Montana State University. 



Simple and multiple linear regressions were derived on a computer 

 using the method of least squares. Unless noted, the term significant 

 refers to statistical significance at the 5% level (p <_ .05). 

 DESCRIPTION OF STUDY AREA 



Prior to the construction of Clark Canyon Reservoir in 1964, 

 the Beaverhead River originated at the confluence of the Red Rock 

 River and Horse Prairie Creek. The Beaverhead River now originates at 

 the outlet of Clark Canyon Reservoir and flows for 69 miles to its 

 confluence with the Big Hole River near Twin Bridges, Montana. Between 



