Methods 



The study area, a natural mountain grassland on Montana 

 State University's Red Bluff Ranch in southwestern Montana, 

 was part of a gently sloping broad, north-facing swale, ap- 

 proximately 6,000 feet in elevation. The soils, derived from 

 granite, are fairly deep and productive. The vegetation, mod- 

 erately grazed until the year of the study, was in good condi- 

 tion and protected from grazing during the course of the 

 study. The vegetation was dominated by Idaho fescue, blue- 

 bunch wheatgrass {Agropyron spicatum), and lupine {Lu- 

 pinus argenteus). Secondary species were more or less typical 

 of those usually associated with the above in this part of 

 Montana, and range from rose pussytoes {Antennaria rosea) to 

 rabbitbrush {Chrysothamnus nauseosus). Wherever the vege- 

 tation and soils were disturbed by rodents, cheatgrass {Bro- 

 mus tectorum) was abundant.'' Precipitation in the generad 

 area did not differ appreciably from normal during the study 

 period. 



One hundred and thirty-five vigorous, mature Idaho fes- 

 cue plants, each approximately 3 to 5 cm. in basal diameter, 

 were selected from the abundant number growing on the 

 study area. These were randomly assigned to nine separate 

 treatments, yielding 15 replications per treatment. 



The nine treatments consisted of all combinations of 

 three levels of clipping and three degrees of competition. The 

 three clipping levels for the selected fescue plants were 

 "none," "heavy," and "extreme." The "none" level was an 

 undipped control. The "heavy" level of clipping consisted of 

 75 percent of the herbage volume removed at the flowering 

 development stage (July 13), with no further treatment. The 



3 



