cemetery. Pull in on the small two-track going east just past the cemetery. Site is on the hill to 

 the south. Aspect: ca. 220°, Slope: 30° Elevation: 5,800 ft. 



METHODS 

 Field Methods 



The Quartz Hill and Thompson's Corner exclosures were established in 1991, while the 

 Dewey Cemetery exclosure was established in 1993. Measurements were made at the three sites 

 on the following dates: 



Quartz Hill- 1 1 Jun 1 99 1 , 7 Jun 1 993 



Thompson's Corner- 11 Jun 1991, 7 Jun 1993, 17 Jun 1996 



Dewey Cemetery- 7 Jun 1993, 17 Jun 1996. 

 The Quartz Hill site was destroyed by vandals in 1995 or early 1996, before measurements could 

 be taken a third time. 



There is one exclosure plot and one control plot at each site. Each exclosure plot is 6 ft X 

 6 ft, marked at the four corners by metal stakes and completely covered by firm wire mesh. 

 Control plots are placed near by the exclosure and are identical except they are uncovered. Each 

 plot is divided into four equal subplots (3 X 3 ft) to facilitate counting. 



In each subplot estimate the basal cover of rock, moss-lichen, bare ground and 

 vegetation, and estimate the canopy cover of grass, forbs and shrubs to the nearest 5%. In each 

 subplot count the number of Arabis fecunda plants in each of three size classes: 



Seedling (S):single rosette smaller that a penny 



Rosette (R): single rosette larger than a penny or multiple rosettes 



Fruiting (F):plants with flowers or fruits 



Data Analysis 



Changes in basal or vegetation cover between exclosure and control plots can be assessed 

 by visual examination of the summary tables. 



There are no replicates of the treatments at any of the sites; thus, statistical tests based on 

 parametric distributions (e.g. t-test, analysis of variance) cannot be used. It is necessary to use 

 counts (frequency) and contingency table analysis. The ratio of Arabis fecunda plants in 

 treatment to control plots in year t can be compared to that in year t+a with a chi-square test. 

 This test assumes that the observations are independent; i.e., the plants counted in year t+a 

 cannot be the same as those counted in year t. Demographic monitoring studies from low- 



