August) were significantly and positively correlated with trees and tall shrubs, 

 snowberry-rose and silver sagebrush (p<.001), and were significantly and nega- g 

 tively correlated with big sagebrush badlands (p<.001) and grassland (p<.01). f 

 Dense clumps of silver buffaloberry along the heads or edges of deep coulees 

 v;ere preferred as nest sites; 8 of 10 nests located in this study were thus 

 situated (table 10). Clutch size averaged 3.1 for seven nests. Most birds 

 had left the study area by mid-September. 



Horned Lark . The horned lark was abundant yearlong throughout the study 

 area and was the most characteristic resident bird in grassland and cropland 

 habitats. Sample abundances for all routes combined reached a low in February, 

 peaked in March as the spring movement was under way and singing activity in- 

 creased, then showed a general decline throughout the summer (figure 12). A 

 September peak coincides with the beginning of the fall movement and a fall 

 increase in singing activity as the length of day becomes similar to that of 

 March. Breeding season sample abundances (April -June) were significantly and 

 positively correlated with both grassland and cultivated habitats (p<.001), 

 which predominate on the Circle West route (table 5). Significant negative 

 correlations were noted for silver sagebrush, trees and tall shrubs, big sage- 

 brush badlands, horizontal juniper, and bluebunch wheatgrass (p<.001). The 

 Circle route provided the highest sample abundances of all routes during all 

 but three of the monthly replicates, and this species was dominant in all runs 

 of this route. This is consistent with the findings of Owens and Myers (1973) 

 that horned lark abundances are highest on heavily overgrazed grasslands, which 

 were abundant along this route. Sample abundances were lowest for the Missouri 

 River and Flowing Well routes and intermediate for the Dreyer Ranch and Prairie 

 Elk routes. Three late nests were observed during this study (table 10). 

 Breeding densities for the five census plots averaged 22 pairs/km^ (9 pairs per ^ 

 square mile.) 



Black-billed Magpie . The black-billed magpie showed a strong association 

 yearlong with trees and tall shrubs, which were also preferred nesting habitat; 

 sample abundances were significantly and positively correlated with trees and 

 tall shrubs and with snowberry-rose (p<.001) and significantly and negatively 

 correlated with grassland (p<.001). It is not surprising, then, that sample 

 abundances were greatest for routes having highest representation of tree and 

 tall shrub habitats, i.e. the Dreyer Ranch, Missouri River and Prairie Elk 

 routes (figure 13). Sample abundances for the four standard routes combined 

 peaked in March, declined drastically during the nesting season as the birds 

 became more secretive, and increased to a peak in October as conspicuous flocks 

 of adults and young were seen frequently. The Circle route apparently traversed 

 poor yearlong habitat, as only a few individuals were observed in fall and winter. 



Rock Wren . The rock wren was restricted to rocky badlands, scoria, steep 

 coulee walls, and eroded mudstone and shale slopes. These habitats were most 

 abundant in the western and southwestern portions of the study area and most 

 rock wrens were encountered on the Flowing Well route, which traversed the edge 

 of these rocky badlands (figure 14). In fact, this route accounted for more 

 than 80% of all observations recorded for this species on the four standard 

 routes at all seasons. Sample abundances were significantly and positively cor- 

 related with big sagebrush badlands, little bluestem, and rocky mountain juniper 

 habitats (p<.001) which were most abundant on the Flowing Well route (table 5). 



80 



