Herbage production was measured on each area from 50 permanent 4. 8 -ft. plots. 

 These plots were distributed in a series of 10 sets, five plots per set, over a rela- 

 tively uniform portion of the grassland. Production was measured each year when the 

 vegetation appeared to have reached full growth and just before many species began to 

 dry. Generally, production data were taken between the middle and end of July at the 

 lower elevation and about 3 weeks later at the upper elevation. 



Production was determined by combining percent-weight estimates and clipping. This 

 technique is similar, but not identical, to that proposed by Hutchings and Schmautz 

 (1969) ; it is based on the concept that it is more efficient to estimate yields as a 

 percent of total yield rather than directly as weight units. Each year, one plot in 

 each set of five was protected from livestock grazing by a wire cage. This plot was 

 considered the reference plot for that particular year. The reference plot was changed 

 each year; thus, a plot served as a reference only once in 5 years. Total herbage 

 production on each of the remaining four plots in the set was estimated as a percent of 

 the total production on the set's reference plot. These estimates were adjusted to 

 compensate for current livestock utilization, which usually was minor. The weight of 

 each species on a plot was then estimated as a percent of the total production on the 

 plot; these estimates were also adjusted to compensate for utilization. The vegetation 

 on the reference plot was then clipped to ground level. Individual production of four 

 selected species served as a continual check on the accuracy of the percent-weight 

 estimates. The total production on the reference plot was used as a basis for estimat- 

 ing production on other plots in the set. Then, the green weight of individual 

 species on each plot was computed. Later, the clipped material was dried to obtain the 

 percent dry matter by important species. All production estimates were converted to 

 dry weights expressed in pounds per acre of air-dry material. 



4 



