in place between sample periods until retrieved on 

 22 September. 



We also installed 3 drift fence pitfall arrays (Figure 

 1 ) on 27 May. Drift fence arrays increase the area 

 effectively sampled and generally are more pro- 

 ductive than lone pitfall traps (Com 1994, Jones et 

 al. 1 996), but they require additional materials and 

 are more time consuming to install. Drift arrays 

 were arranged with three 2.5 m long masonite fins 

 radiating out fi"om a central pitfall can (Figure 4). 

 Masonite fins were held in place with wooden lath. 

 An additional pitfall can was placed at the end of 

 each fin. Thus, 4 pitfall cans were associated with 

 each drift fence array. Drift fence arrays were 

 checked following the protocol for pitfall transects 

 and left in place until retrieved on 22 September. 



Captured mammals were identified to species, 

 when necessary using keys in Hoffmann and Pattie 

 (1968) and Clark and Stromberg(1987) as well 

 as more detailed literature. Individuals were 

 weighed and measured, and sexed if possible. 

 Vouchers of some species (especially shrews) 

 were preserved for additional study and determi- 

 nation by Dr. Kerry Foresman, and deposited in 

 the Philip L. Wright Vertebrate Museum at the 

 University of Montana. We also routinely 

 searched for regurgitated raptor pellets below 

 wooden fence posts and other potential perches in 

 the sandhills. Pellets were collected and dissected 

 for small mammal skulls and other vertebrate 

 remains. Small mammals recovered fi^om pellets 

 were identified to species based on skull charac- 

 teristics. 



Birds 



We conducted no systematic counts of birds, but 

 maintained daily lists of bird species detected 

 during other work, recording the area where they 

 where detected, and any additional striking 

 behavior or habitat notes (such as singing or 

 association with a particular habitat feature). 

 Although we conducted no nest searches, we 

 documented all nests located, recording stage of 

 nesting, nest contents, and nest position. 



Amphibians and Reptiles 



We found virtually no wetland areas in the 

 Sandhills during our visits, and so devoted little 

 time to active searches for amphibians. We 

 identified itogs by call and visually as we encoun- 

 tered them. We sought reptiles as we drove roads 

 and traversed the study area during other activi- 

 ties. 



Systematic sampling was limited to the earlier- 

 described pitfall transects and drift fence arrays, 

 which are standard sampling techniques for these 

 vertebrate groups (Com 1994), especially for 

 adult salamanders and lizards that may be over- 

 looked using other techniques. Vouchers were 

 collected and deposited in the herpetological 

 collection at the Idaho Museum of Natural His- 

 tory, Idaho State University. 



Insects 



We focused our insect sampling on 2 groups, tiger 

 beetles (genus Cicindela) and diurnal butterflies 

 (Lepidoptera: Papilionoidea and Hesperioidea). 

 We chose these groups because they are visible 

 and often tied to local habitat conditions. Cap- 

 tured individuals were subdued with ethyl acetate 

 in a killing] ar and pinned for later examination, or 

 identified in the field. 



We sampled figer beetles (Figure 5) in two ways: 

 active searching/sweep-netting, and pitfall trap- 

 ping. As with many other small terrestiial organ- 

 isms, tiger beetles are readily captiared in pitfall 

 ti-aps (Morrill etal. 1990, Clark and Blom 1992, 

 Berghe 1992), and drift fences improve capture 

 success (Knisley and Schultz 1997). Because our 

 primary objective was to document habitat use 

 and species presence, rather than relative abun- 

 dance, we did not place preservative in the pitfall 

 cans to kill insects. Nevertheless, our pitfall traps 

 were successfijl in capturing about 200 individual 

 tiger beeties. 



During our first 3 visits to the Sandhills, we docu- 

 mented relative abundance at 4 sites by walking 

 1 00 m road ti-ansects and counting numbers of 

 each species, noting habitat adjacent to the road. 



