set of samples was taken for the main channel 

 and a second set for all tributary channels 

 for each drainage. Computations of precision 

 were then made for (1) the main channel, 

 (2) the tributary channels, and (3) the total 

 drainage (1 and 2 combined). Because tribu- 

 taries generally were smaller and steeper than 

 the main channels, they were expected to have 

 different characteristics. 



Limits of accuracy for proportions of bank 

 stability and vegetation types were easy to 

 compute because they constituted a binomial 

 distribution. The following formulas were 

 used to calculate the confidence intervals.' 



' Freese Frank. Elementary Forest Sampling, pp. 

 62-63. USDA Handbook 232. Wash.. D.C. 1962. 



where 



CIgg = the confidence interval that has 

 a 66-percent probability of con- 

 taining the true value of the 

 stream characteristic described. 



_ number of samples in class/type 



p — ^ 



total number nf samples 



For example, the proportion of unstable banks 

 ranged from 4 percent on the Main Fork of 

 the Bear River to 10 percent on Hayden Fork, 

 excluding tributaries (table 2). Computations 



Table 2 — Precision of bank stability estimates 



Channels 



Number of 

 samples 

 (n) 



Percent 

 unstable 

 banks 



(p) 



Standard Confidence 

 error interval 

 (Sp) (66% level) 



— Percent — 

 HENRY'S FORK 



Main 136 7.4 -2 5-9 



Trib 102 2^0 



238 5.0 =1 4-6 



MAIN FORK OF BEAR RIVER 



Main 100 4.0 =2 2-6 



Trib 



100 4.0 =2 2-6 



HAYDEN FORK 



Main 90 10.0 ^3 7-13 



Trib 86 3.5 ±2 2-6 



176 el =2 sT" 



All drainages 514 5.4 —1 4-6 



10 



