three-tip sagebmsh, A. tripartita; common 

 rabbitbrush, Chrysothamnus nauseosus; green 

 rabbitbrush, C viscidiflorus), forbs (silvery 

 lupine, Lupinus argenteus; silky lupine, L 

 sericeus; brittle prickly-pear, Opuntiafragilis; 

 slender-flowered scurf-pea, Psoralea tenuiflora), 

 and grasses (thick-spiked wheatgrass, ^gro/7yra« 

 dasystachyum; Idaho fescue, Festuca 

 idahoensis; prairie junegrass, Koeleria cristata; 

 needle-and-thread, Stipa comata). Sandhill sites 

 occur in 3 classes based on topography and 

 effects of sand movement: lower-slope erosion 

 blowouts, (Figure 2), upper-slope deposition, and 

 stabilized (both upper and lower slopes), (Figure 

 3). Lesicaand Cooper (1999) identified three 

 successional stages (early-, mid-, and late-seral) 

 from the previous classes. 



During our 1 999 field work we used the Lesica- 

 Cooper succession classification as a reference for 

 documenting animal detections. We found that 

 four physiognomic types (inter-dune trough, dune 

 crest, low dunes, and sand flats) were also usellil 

 for our purposes in categorizing trapline and 

 driftfence locations (see Appendix 3 ). Inter-dune 

 troughs were elongate depressions between 

 parallel dune ridges of the West Hills. They were 

 equivalent to the stabilized class but often sup- 

 ported denser stands of sagebrush than were 

 found on dune slopes; troughs also covered more 

 extensive and relatively flat terrain. Dime crests 

 were equivalent to either deposition or stabilized 

 classes, but sometimes occurred as lengthy and 

 wide ridge tops that gently sloped along the main 

 dune axis. The low dunes category was generally 

 equivalent to the stabilized class, but contained 

 small sparsely vegetated sandy depressions 

 intermixed among dune swales; this category was 

 most extensive east of Tepee Creek. Sand flats 

 were extensive expanses of low-relief terrain 

 supporting sagebrush and grasslands in various 

 degrees of cover; an area roughly 6 km wide east- 

 to-west bracketing Tepee Creek fit into this 

 physiognomic category. 



Survey Methods 



We visited the Centennial Sandhills four times 

 during the summer of 1999: 24-28 May, 28 June- 

 3 July, 26-30 July, and 20-23 September Daily 

 temperature minima and maxima were recorded 

 with a Taylor minimum-maximum thermometer to 

 characterize weather conditions during the periods 

 of daily sampling. We employed a variety of 

 techniques to sample a diverse array of animal 

 groups. In some cases we actively searched for 

 animals or their spoor, while in other cases we 

 used a variety of passive trapping methods. These 

 are described below for each animal group. In 

 many cases we attempted to systematize sampling 

 procedures, but opportunistic sampling was also 

 employed to increase our survey coverage of the 

 Sandhills. We also searched the Heritage data- 

 bases, pubhshed and gray literature, and museum 

 records for documented historical occurrences of 

 target species, especially for mammals, amphib- 

 ians, reptiles, and tiger beetles. Tiger beetle 

 collections at the University of Idaho and Wash- 

 ington State University were examined for speci- 

 mens collected in Beaverhead County and adja- 

 cent localities in Idaho. 



Common and/or scientific names in tables and text 

 throughout follow Jones et al. ( 1 986) for mam- 

 mals, American Ornithologists' Union (1998) for 

 birds, Stebbins ( 1 985) for amphibians and rep- 

 tiles, Pearson et al. ( 1 997) for tiger beetles, and 

 Opler ( 1 999) for diurnal butterflies. Statistical 

 analyses, where used, follow standard procedures 

 presented in Sokal and Rohlf ( 1 98 1 ). Statistical 

 significance of tests is assumed when p < 0.05, 

 although we recognize that statistical significance is 

 not the same thing as biological significance. 

 Throughout the text "West Hills refers to that 

 portion of the sandhills west of Tepee Creek and 

 "East Hills" is that portion of the sandhills east of 

 Tepee Creek (see Figure 1 ). 



