the phenological stages at which trampHng occurred 

 were also different. Hylgaard and Liddle's (1981) one- 

 time treatment occurred at a late date (August 21), when 

 vulnerability may have been low and, as they acknow- 

 ledge, the time between treatment and measurement was 

 not sufficient to observe the delayed damage that occurs 

 to the dominant plant, Empetrum nigrum. 



Where recovery of vegetation trampled at different fre- 

 quencies has been followed for a year or more, making 

 recovery periods approximately equivalent, differences 

 related to frequency of trampling have been minimal 

 (Campbell and Scotter 1975; Douglas and others 1975; 

 Bayfield 1979). Type of vegetation and intensity of tram- 

 pling had a much greater effect on vegetation cover. 



It is difficult, then, to form conclusions about the 

 effect of trampling frequency. Effects are probably 

 different in different vegetation types and they certainly 

 vary with differences in number of passes and the period 

 of time over which use is spread. When the number of 

 passes per period of time is high, the effects of tram- 

 pling frequency diminish because alteration reaches max- 

 imum limits. Common sense suggests that at the other 

 extreme— where a small number of tramples are spread 

 over a very long period of time— concentrated trampling 

 would be more destructive than trampling spread over 

 time. Yet, all studies of trampling frequency have found 

 the opposite. Spreading the trampling out over a grow- 

 ing season causes more damage than concentrating it at 

 one time, even at trampling levels as low as 25 passes 

 per year. Even here, however, the consistent effect of 

 frequency is much less important than other factors. 



The effect of seasonality of trampling is also complex 

 and difficult to determine. Again the effect probably 

 varies with such factors as number of passes, morphol- 

 ogy and phenology of the constituent plant species, and 

 the seasonality of soil moisture. For example, trampling 

 grasses late in the season, after they have completed 

 their annual growth and reproduction, may have little ef- 

 fect compared to early trampling. Low-growing shrubs, 

 in contrast, are often more susceptible to late-season 

 trampling, because they become increasingly brittle as 

 the season progresses. 



Singer (1971) found that vegetation cover was not 

 significantly affected by the time of summer (July 3- 

 August 13) at which trampling occurred. Nagy and 

 Scotter (1974) also found no consistent difference 

 between trampling in early season (early June-early July) 

 and midseason (late July-early August). If pretreatment 

 measurements had been taken and differences in recov- 

 ery periods had been accounted for, however, it might 

 have been possible to identify meaningful differences for 

 individual species. Holmes and Dobson (1976) found that 

 late season trampling (early September) caused much 

 more cover loss than midseason trampling (early August) 

 for nine of the 14 species that they studied. They attrib- 

 ute this higher vulnerability in late season to reduced 

 plant vigor and drier, more brittle, plant parts. There 

 may, however, be little relationship between cover loss 

 immediately after trampling and cover at some time in 

 the future. Early-season trampling may, for example. 



have a more profound effect on the stored reserves and 

 reproductive success of the plants, changes that will 

 have a greater effect on future conditions. 



WESTERN MONTANA STUDY AREA 



The study reported here attempts to extend our know- 

 ledge of the relationships discussed in the preceding re- 

 view. Originally five habitat types were selected for 

 study. The types chosen were those most commonly 

 used for camping at low elevations to midelevations in 

 the Bob Marshall Wilderness. Four of these are forested: 

 Abies lasiocarpa/Clintonia uniflora (subalpine fir/queen- 

 cup beadlily), Abies lasiocarpa/Vaccinium caespitosum 

 (subalpine fir/dwarf huckleberry), Abies lasiocarpa/ 

 Xerophyllum tenax (subalpine fir/beargrass), and 

 Pseudotsuga menziesii/Symphoricarpos albus (Douglas- 

 fir/snowberry) (Pfister and others 1977); the other is a 

 grassland: Festuca scabrella-F. idahoensis (rough fescue- 

 Idaho fescue) (Mueggler and Stewart 1980). Two repli- 

 cate experimental plots were established in each type. 

 Following initial sampling, it became clear that the two 

 Abies lasiocarpa/Clintonia uniflora plots were different. 

 The ground cover of one was predominantly lush forbs, 

 particularly Thalictrum occidentale. This is the Clintonia 

 uniflora phase of the type, called ABLA/CLUN here- 

 after. The other was predominantly low shrubs, such as 

 Linnaea borealis and Arctostaphylos uva-ursi. This is the 

 Vaccinium caespitosum phase of the type, called ABLA/ 

 CLUN-VACA hereafter. 



All of the experimental plots were located close to 

 roaded access on the fringes of the Bob Marshall Wilder- 

 ness (fig. 2). The forested plots were all on the Seeley 

 Lake District of the Lolo National Forest; the grassland 

 was on the Montana Department of Fish, Wildlife and 

 Parks' Blackfoot-Clearwater Game Range. All of the 

 plots are located at elevations of 4,200 to 4,400 ft (1 260 

 to 1 320 m). They are all essentially flat (slopes of to 

 7 percent), except for the Pseudotsuga menziesii/ 

 Symphoricarpos albus (PSME/SYAL) plots, which are on 

 a 20 to 25 percent southwest slope. 



The ABLA/CLUN plot is located on a stream terrace 

 above Dunham Creek, in a slight concavity at the base 

 of a toe slope. The soil is a Eutric Glossoboralf, with 

 thick organic horizons— 4.3 inches (11 cm), a silty loam 

 surface mineral horizon, and an absence of coarse frag- 

 ments throughout the soil profile. Soil profiles for this 

 and the other habitat types are described in appendix 1. 



The overstory on the plot is a dense stand, 

 predominantly of large Larix occidentalis and Picea 

 engelmannii. Pinus contorta, Abies lasiocarpa, and 

 Pseudotsuga menziesii are also present. Spherical den- 

 sitometer readings indicate a canopy cover of about 90 

 percent. The ground cover is dominated by lush forbs, 

 particularly Thalictrum occidentale, Arnica latifolia. 

 Smilacina stellata, and Clintonia uniflora (fig. 3). Various 

 shrubs, of which Berberis repens is most abundant, are 

 also present. Tables giving cover and frequency for each 

 species on this and other habitat types are in appendix 2. 



6 



