lanes (Bayfield 1979). What I have called "estimated 

 relative cover" is a less precise measure of change; in 

 this case, original cover was not reported, so estimated 

 relative cover is the surviving cover on the trampled 

 lane expressed as a percentage of cover on the control 

 lane. To be as accurate as relative cover, the original 

 cover on controls and all trampled lanes would have to 

 be identical. Although this is unlikely to be the case, 

 differences are seldom substantial enough to create a 

 large problem. For some studies, I reanalyzed original 

 data to convert them to these two consistently defined 

 variables. They are generally arranged from forested to 

 nonforested communities and, within these classes, from 

 low to high elevation. 



Almost without exception, these graphs illustrate the 

 curvilinear relationship between amount of use and 

 amount of vegetation loss. Vegetation is lost most 

 rapidly following the initial application of trampling; suc- 

 cessive incremental increases in amount of trampling 

 have progressively less effect on vegetation cover. This 

 corroborates the conclusions of several studies of the 

 relationship between use and impact on existing trails 

 and campsites (for example. Dale and Weaver 1974; Cole 

 and Fichtler 1983). Only the most resistant of these 

 vegetation types, the Poa pratensis-Festuca idahoensis 

 grassland, does not lose cover most rapidly at low levels 

 of trampling. 



Relative Resistance of Different Plant 

 Community Types 



Of the many potential means available for evaluating 

 the resistance of different plant communities, three seem 

 particularly useful: 



1. Liddle (1975a) suggested evaluating resistance on 

 the basis of the number of passes required to reduce 

 cover to 50 percent of its original value. This provides a 

 useful index of how rapidly cover is lost after the intro- 

 duction of trampling. Such an index does not, however, 

 indicate how well a community type will stand up to 

 either prolonged or heavy amounts of trampling. 



2. Although none of these 1-year studies can be used 

 to estimate the effects of prolonged trampling, most 

 studies report the effects of at least 800 passes. There- 

 fore relative cover after 800 passes can be used as an 

 index for resistance to heavy trampling. 



3. The most useful index would reflect resistance to a 

 broad range of trampling intensities. The curves in fig- 

 ure 1 can be used to provide such an index. To facihtate 

 comparison, an index was calculated that expressed the 

 area under each curve (up to 800 passes) as a percentage 

 of the entire graphed area. In effect, this integrates the 

 effects of trampling across the range from to 800 

 passes, expressing surviving cover as a mean percentage 

 of control conditions. Table 1 categorizes 32 plant com- 

 munities on the basis of these three alternative indices. 



Several conclusions can be drawn from table 1. First, 

 the relative resistance of a community type to heavy 

 trampling is similar to its resistance to light trampling. 

 Although some changes in order exist (for example, eri- 

 caceous undergrowth is usually more sensitive to heavy 

 than to light trampling), the rankings based on these 

 three indices of resistance are generally similar. Thus, 



any single index (such as the number of passes it takes 

 to reduce cover to 50 percent of its original value) can be 

 used as a rough index of initial resistance. It is impor- 

 tant to remember that resistance to prolonged trampling 

 and ability to recover are not reflected in any of these 

 indices. 



Second, there is little relationship between elevational 

 zone and resistance. Alpine communities, even within the 

 same general area, range from very susceptible (Phyl- 

 lodoce glanduliflora) to very resistant [Carex nigricans); 

 the same is true of low elevation types. 



Third, most of the very resistant types are dominated 

 by grasses, sedges, and rushes. But not all grasslands 

 are highly resistant. Some have only a moderate level of 

 resistance, and the sedge marsh (type 6) is one of the 

 least resistant vegetation types. This resistance gradient 

 may reflect, in part, the abundance of moisture. As men- 

 tioned, the initial effects of trampling are generally most 

 profound on moist sites and the sedge marsh is inun- 

 dated by water for much of the year (Nagy and Scotter 

 1974). 



Finally, with one exception, the forested sites exam- 

 ined are quite susceptible to trampling damage. Most of 

 these forests occupy mesic sites and have a ground cover 

 dominated by lush forbs. The whitebark pine forest, with a 

 ground cover dominated by the ericaceous shrub 

 Vaccinium scoparium (grouse whortleberry), is considera- 

 bly more resistant. Some of this apparent resistance 

 may be a result of study method. In this study, tram- 

 pling treatments and measurements were completed all 

 at once. Vaccinium spp. often die slowly after trampling 

 (Bayfield 1979), so by irrmediately measuring survival, 

 cover loss of Vaccinium scoparium may have been under- 

 estimated. In studies of existing recreation sites, most 

 researchers have found Vaccinium scoparium to be 

 a fragile ground cover (see, for example, Dale and 

 Weaver 1974; Cole 1982a,b). 



Relative Resistance of Different Growth 

 Forms 



Morphological characteristics of plants that influence 

 resistance to trampling include location, size, erectness 

 and strength of leaves, strength of petioles and nodes, 

 woodiness and erectness of stems, extent and type of 

 root system, and bud location. Characteristics found to 

 be common to resistant species in experimental studies 

 are short, procumbent, and/or flexible stems (Schreiner 

 1974), a low-growing, matted or tufted growth form 

 (Singer 1971; Holmes and Dobson 1976), basal leaves 

 (Holmes and Dobson 1976; Rogova 1976), and leaves 

 that are pliable, folded, and either thick or narrow and 

 wiry (Schreiner 1974; Holmes and Dobson 1976). Particu- 

 larly fragile plants usually have tall, caulescent (obvi- 

 ously leafy), succulent stems and thin leaves, or brittle, 

 low-growing woody stems and branches (Singer 1971; 

 Holmes and Dobson 1976; Rogova 1976). 



Studies of campsites and trails have consistently 

 found graminoids (grasses and grasslike plants) to be 

 particularly resistant to trampling (see, for example. 

 Cole 1982a). This has also been confirmed in experimen- 

 tal studies (Wagar 1964; Nagy and Scotter 1974; 

 Schreiner 1980). Forbs and shrubs are more variable in 



4 



