provide acceptable absolute minimum recommendations for four 

 of the reaches, it had no advantage over the single transect 

 method. It was costlier, more time consuming, sometimes dif- 

 ficult to interpret, and occasionally unproductive. 



Minimum flow recommendations based on a fixed percentage 

 of the mean flow of record may be valid for the trout rivers 

 of southwest Montana. The absolute minimum recommendations 

 derived from the trout-flow data for the five reaches ranged 

 from about 31-51% of the mean flow. The percentage required 

 appears to depend on the channel morphology with the wider, 

 shallower rivers such as the Madison requiring a greater per- 

 centage of the mean. The more typical rivers of the study 

 area (Beaverhead, Gallatin and Big Hole) required an absolute 

 minimum equal to about 3 3% of the mean. A minimum flow of 

 10% of the mean as recommended by the "Tennant or Montana" 

 method was unacceptable in all five reaches. Since Tennant 's 

 minimiam flow is defined as a short-term survival flow, the 

 absolute minimums derived from the trout-flow data are ex- 

 pected to exceed Tennant 's minimum recommendations. 



The acceptance of less than 50% of the optimum flow recom- 

 mendations indicates that the IFG incremental method in its 

 present state of development is not a consistent method for 

 deriving instream flow recommendations for the trout rivers 

 of Montana. Possible means for improving the present IFG 

 method for use on the relatively high gradient, boulder and 

 cobble-strewn trout rivers of the study area include (1) modi- 

 fying the existing IFG model to use bottom velocities rather 

 than the mean velocities in the water column to compute the 

 weighted usable area, (2) developing probability-of-use curves 

 from data collected for river populations of trout, and 

 (3) incorporating cover into the IFG model. 



The predictive capabilities of the IFG-4 and the Water 

 Surface Profile (WSP) hydraulic simulation models were also 

 evaluated. The IFG-4 predictions of water surface elevations, 

 velocity and depth were generally superior to those of the 

 WSP model. The IFG-4 predictions of wetted perimeter, even 

 though approximations, were judged superior to those of the 

 WSP model based on the greater accuracy of the predictions 

 of water surface elevation. Additional testing is needed to 

 clarify the reliability of the wetted perimeter predictions 

 of both models. 



99 



