In September (table 16) ■ were again abundant, as were 

 Chironomidae. Totals were comparable for Glendive (2964) and Intake (3638). 



and Chironomidae again dominated in the October samples 

 (table 17). Number of taxa and total number of individuals greatly increased 

 at both stations. 



November samples showed Hydropsyche and Chironomidae dominant (table 18). 

 Totals were high at Glendive (20,245) but considerably reduced from October 

 at Intake (8988). 



All 48 samples taken each month were pooled to illustrate which orders 

 dominate the fauna (table 19). The fauna was dominated by Trichoptera and 

 Ephemeroptera with Diptera third. Ephemeroptera monthly percentages ranged 

 from 11.7 to 73.6 while Trichoptera percentages varied from 21.1 to 56.3 

 percent of the total. The October and November samples contained more infor- 

 mation than the August-September samples, probably due to summer emergence 

 losses and the presence in August and September of very small larvae and 

 nymphs, most of which passed through the collecting net. Mean population 

 estimates varied from 138/m2 (August) to 681 /m2 (October). Percentage com- 

 position of orders at each station is shown in table 20. 



Results obtained with the kick net were compared with results of the 

 Water's sampler (figures 44 and 45). The Water's sampler is 19.5 in high; 

 thus only kick samples taken in depths less than 19.5 in were compared. 

 Results were similar, but the number of organisms obtained with the kick net 

 was always lower than numbers obtained with the Water's sampler. Several kick 

 samples were taken at the water's edge in water too shallow to sample with 

 the Water's sampler, tending to expand the range and reduce the mean. Results 

 from the two samplers followed the same trend over time at both stations, and 

 a line joining the means of both methods is almost parallel. 



ENVIRONMENTAL REQUIREMENTS 



Multiple regression analyses were performed on the current-depth data with 

 current and depth as independent variables and number of taxa and number of 

 individuals as dependent variables. Three models were applied: 1) untransformed; 

 2) semilog transformation (of dependent variables); and 3) log-log transformation. 

 The detailed results of these analyses, for all three models, are reported in 

 Newell 1976 and are on file with the Montana DNRC. The general results are 

 given in tables 21 and 22. 



Number of taxa and number of individuals yield similar results when 

 regressed against current velocity. Figures 46-48 show how these regression 

 equations can be used to predict the numbers of individuals at any particular 

 current or depth. The deviation of the data from the regression line is 

 demonstrated in figure 48, for example, where the regression coefficients (r) 

 are 0.774 for current and 0.808 for depth. 



75 



