River at the North Dakota border to 6900 feet at the summit of 

 Mount Baldy in the Bear's Paw Mountains. Except for the Bear's 

 Paw and Little Rocky Mountains, the area lies entirely within the 

 Glaciated Missouri Plateau section of the Great Plains Physio- 

 graphic Province (see Fig. 6 in Montagne et al. 1982). The 

 southern boundary of this section is defined by the southern 

 limit of continental glaciation during the last ice age. For the 

 most part, these plains consist of relatively flat to gently 

 rolling sedimentary (particularly shale) and glacial till sur- 

 faces modified by stream erosion and past glaciation (Veseth and 

 Montagne 1980) . Some areas of sharply dissected badlands topog- 

 raphy do occur, particularly along the Missouri River. 



The Bear's Paw and Little Rocky Mountains occur as isolated 

 "island" uplifts within the study area. A wide range of parent 

 materials occur within these mountain ranges although the central 

 portions of both ranges are predominantly igneous (Veseth and 

 Montagne 1980) . 



Climate 



Most of the study area experiences the extreme summer heat 

 and winter cold of a continental climate and lies directly in the 

 path of many arctic air masses from the north (Montagne et al. 

 1982). Average annual precipitation varies from over 30 inches 

 at the crest of the Bear's Paw Mountains to between 10 and 12 

 inches throughout the bulk of the study area (see sheet 2 in Ross 

 and Hunter 1976) . The average length of the freeze-free season 

 varies from less than 7 days at the crest of the Bear's Pav/ 

 Mountains to greater than 130 days along portions of the Milk 

 River (see Fig. 13 in Montagne et al. 1982). 



METHODS 



Data Collection 



To maximize the efficiency in sampling the range of vegeta- 

 tion and environmental variation, sample sites were selected 

 using a modification of the "gradsect" ( grad ient transect) method 

 described and evaluated by Gillison and Brewer (1985) and applied 

 by Austin and Heyligers (1989) . The method, as applied in the 

 present study, involved selecting a set of USGS 7.5' topographic 

 quadrangle maps containing the maximum perceived range of shrub- 

 land/grassland environmental variation in the overall study area. 

 Emphasis was placed on representing the range of moisture, 

 temperature, radiation, and soil nutrient regimes since these 

 factors likely have a primary influence on species occurrence and 

 growth . 



The following site attribute information was overlaid onto a 

 USGS quadrangle index map of the study area to select quadrangles 



