STUDY AREA AND METHODS 



The area of study lies mostly in Petroleum County, but includes a small part of adjacent 

 Fergus County to the west and north. It encompasses Tegan, Winnett, and Cat Creek and 

 extends south to Flatwillow Creek and north to the Wilder Trail readjust south of Charles M. 

 Russell National Wildlife Refuge, bounded on the west near Highway 19 and east by the 

 Musselshell River. Most of the study area falls within the Montana Shale Plains subsection 

 (Nesser et al. 1997) with portions of the Montana Glaciated Plains subsection to the north. 

 Elevation averages about 3000-3500 ft, encompassing dissected plains, hills, ridges, and 

 floodplains formed in shale, siltstone, and sandstone. Mean annual precipitation ranges from 10- 

 14 inches, about one third falling as snow, with dry summers. The area is used predominantly 

 for livestock grazing with some dryland farming. 



Vegetation coverage in the area includes four course-level categories (Jorgensen 1979); 

 shrub-grasslands (="shrubsteppe"), wetlands, mixed grasslands, and coniferous forest (= 

 ponderosa pine woodland and savanna) Vegetation classification used at roadside point counts 

 follows the gross categories used by the US Forest Service Northern Region Landbird 

 Monitoring Program (Hutto and Young 1999), with some additional refinements incorporating 

 shrub and grassland structure (J. Hoffland personal communication) suggested by Montana PIF. 

 Because most shrubsteppe vegetation was low-stature (< 0.5 m tall) sagebrush {Artemisia 

 tndentata), and points dominated by other cover types were too few to analyze in finer detail, 

 only three major "categories" are discussed in this report; low-stature sagebrush of various 

 densities, mixed grassland/dryland agriculture/pasture, and ponderosa pine {Pinus ponderosa) 

 savanna/woodland. Pine occurs on parallel east-west tending ridges in the study area, and near 

 C M. Russell National Wildlife Reflige. Route habitat descriptions include mention of habitat 

 cover at specific stops when different fi^om the predominant type; at a few points the bird fauna 

 was obviously influenced by wetlands or stream-side riparian vegetation (see route descriptions), 

 and this, too, is noted for specific stops. 



Point count routes were selected to pass through as much land under BLM ownership as 

 possible (based on BLM ownership maps), given accessibility and road conditions, all routes 

 were on secondary or tertiary unpaved roads. In a few cases (such as the Flatwillow South route) 

 maps failed to reflect current ownership, and routes were run largely adjacent to private lands. 

 No attempt was made to randomly place routes and points or to stratify these relative to habitat 

 availability. The primary goal with these counts was to conduct a rapid assessment of bird 

 species in the area, and to establish routes that can be run in future years when deemed desirable 

 to do so Roadside counts could provide a biased representation of bird species presence and 

 relative abundance on smaller landscape scales, due to alteration of habitat near road corridors 

 and how birds respond to the presence of roads. However, analyses of paired point counts on 

 roads and adjacent off-road sites in the same habitat in forest (Hutto et al. 1995) and shrubsteppe 

 (Rotenberry and Knick 1995) showed that road-side counts provide a good representation of the 

 bird fauna of a study area, especially if counts are conducted from secondary and tertiary roads. 



Each route consisted of 10 stops spaced 0.5 miles apart (one route consisted of only 7 

 stops) where birds were counted. In general, methodology at count points followed procedures 

 discussed in detail elsewhere (e.g., Hutto et al. 1986, Ralph et al. 1993) Counts were conducted 

 for 3 min (standard Breeding Bird Survey length) to allow for some comparison with BBS data. 

 This shorter count period (others suggest 5-10 min counts, depending on conditions) was chosen 



