consisted of 40 trapping stations systematically 

 spaced in 10 rows and four columns. The rectangu- 

 lar grids were positioned lengthwise along Summit 

 Creek and straddled the stream channel. The trap- 

 ping grid sampling the beaver pond area was 

 centered on the main pond. 



Grid dimensions were surveyed and trapping 

 stations marked with numbered stakes at 25-m 

 intervals. Two Museum Special mouse traps and 

 one Victor rat trap were located within a 2-m radius 

 of each measured gprid point. Traps were baited 

 with a mixture of peanut butter and rolled oats and 

 examined for 5 consecutive days from August 3-7, 

 1988, and from August 17-21, 1989. Traps on both 

 grids were examined once each day to minimize the 

 influence of weather on comparative catches. 

 Species, sex, weight, and station number were 

 recorded for each captured animal. The reciprocal 

 of Simpson's index (D = where Pi is the pro- 



portion of the sample belonging to the ith species) 

 was used to calculate species diversity (Hill 1973). 



Plant nomenclature is from Hitchcock and 

 Cronquist (1973). Scientific and common names 

 of mammals follow Jones and others (1986). 



RESULTS AND DISCUSSION 



A total of 2,400 trap nights yielded 203 individual 

 small mammals in the two seasons of study at 

 Summit Creek. Eight species of rodents and shrews 

 were caught; these included montane voles {Micro- 

 tus montanus), 46.3 percent; shrews iSorex spp.), 

 32.0 percent; deer mice (Peromycus maniculatus), 

 15.8 percent; water shrews (S. palustris), 2.5 per- 

 cent; western jumping mice (Zapus princeps), 1.5 

 percent; long-tailed voles (Af. longicaudus), 1.0 

 percent; Great Basin pocket mice (Perognathus 

 parvus), 0.5 percent; and northern pocket gophers 

 (Thomomys talpoides), 0.5 percent. Montane voles, 

 deer mice, and shrews were caught on both the 

 beaver pond area and the adjacent nonwillow 

 habitat. Other species were caught irregularly and 

 in smaller numbers. 



The total number of small mammals trapped was 

 much larger (153) in the beaver pond habitat 

 compared to the nonponded habitat (50). Relative 

 density in the beaver pond habitat was about three 

 times that of the adjacent area in each of the 2 years 

 of study (table 1). Similarly, small mammal stand- 

 ing crop biomass was substantially higher in the 

 beaver pond ecosystem. There was no pronounced 

 difference in the two habitats in either species 

 richness or our estimates of species diversity. 

 Six small mammal species were recorded in each 

 habitat (table 1). 



Montane vole populations differed dramatically 

 between the two habitats (table 1). Almost 80 

 percent of the montane voles trapped were caught 

 in the beaver pond complex. Of these, most were 



trapped in streamside habitats, with the frequency 

 of capture highest in thick stands of grasses, sedges, 

 and rushes that formed the ground layer of willow/ 

 mesic herbaceous communities. None were caught 

 in sagebrush/upland communities. In Idaho, mon- 

 tane voles occur most commonly in moist, weedy, 

 or brushy areas near water (Larrison and Johnson 

 1981). The importance of vegetative cover to the 

 montane vole has been well documented (for ex- 

 ample. Brown 1967a; O'Farrell and Clark 1986). 



Shrews other than the water shrew were trapped 

 almost exclusively in the willow/mesic herbaceous 

 communities that encircled the beaver pond. Only 

 a single specimen was caught in the nonwillow 

 study area. Shrews, along with montane voles, 

 were the most commonly trapped small mammals 

 in the beaver pond ecosystem (table 1). They were 

 frequently caught in runways of the montane vole. 

 In Idaho, most shrews are foimd in moist, grassy 

 habitats (Larrison and Johnson 1981), but they 

 occur in a variety of other habitats including forests 

 and shrublands (Brown 1967b). The association of 

 vagrant shrew (Sorex vagrans) populations with 

 densely vegetated and marshy habitats has been 

 reported in California (Ingles 1961), Colorado 

 (Spencer and Pettus 1966), Oregon (Hooven and 

 others 1975), Montana (Clothier 1955), and 

 Washington (Newman 1976). 



Captures of water shrews were infrequent and 

 consisted of only two or three animals in each trap- 

 ping period. They were trapped in both the beaver 

 pond and adjacent nonponded habitats. All were 

 caught near swifl-flowing sections of the stream 

 in willow or mesic herbaceous communities. In 

 Manitoba, a close correlation existed between the 

 local distribution water shrews and beaver pond 

 habitats (Wrigley and others 1979). 



Deer mice were trapped in both the beaver pond 

 habitat and the nonwillow area, and in each of the 

 plant community types described for the Summit 

 Creek study site. Overall, relative density in the 

 nonponded area was three times that found in the 

 beaver pond ecosystem (table 1). Most deer mice 

 were trapped in sagebrush/upland communities 

 that bordered the riparian zone. They were caught 

 infrequently in the thickly vegetated, willow- 

 dominated beaver pond habitat and in other stream- 

 side habitats. The deer mouse is found in diverse 

 habitats including swamps, waterways, forests, 

 grasslands, and deserts, and among rocks and cliffs 

 (Larrison and Johnson 1981). Higher relative 

 densities in the drier and less densely vegetated 

 sagebrush community type at Summit Creek 

 suggests an affinity by the deer mouse for low-shrub 

 habitats with a sparse herbaceous understory. 



Other species of small mammals were either 

 trapped or observed on the Summit Creek study 

 site. The Great Basin pocket mouse, a species 

 normally associated with arid and semiarid habitats 



2 



