impacts of hiking and camping on soils and vegetation: a...

Impacts of Hiking and Camping on Soilsand Vegetation: a Review

David N. ColeAldo Leopold Wilderness Research Institute, Forest Service, Missoula, Montana

USA

Introduction

Ecotourism effects local cMivironments in manyways. Although some ot the most dramatic environmental changes result from development ofthe infrastructure to support tourism, morewidespread impacts result from the recreationalactivities that tourists engage in. For ecotouristsengaged in adventurous pursuits, hiking andcamping are perhaps the most common activities that can have profound ecological impacts.This is particularly true in more remote places,protected as parks or wilderness.

Of the many environmental effects ofhiking and camping, impacts on soil and vegetation have been most thoroughly explored.Consequently, the literature on this subject isvoluminous and is a challenge to review thoroughly. The strategy of this chapter is to providean historical context for the development of thisliterature, discuss the types of studies that havebeen employed (cvich with inherent strengthsand weaknesses) and briefly assess the geographical distribution of research. Emphasis isplaced on development of generalities from theliterature and identification of critical knowl

edge ga[:)s, rather than a comprehensive reviewof many site- and context-specific descriptivestudies. I try to identify th(^ early papers thatprovided the genesis of ideas and conc epts, aswell as recent papers that extend earlier workconceptually and geograi)hically. Inevitably I

havedrawn moreexamples from my ownworkthan might be representative because 1am mostfamiliar with their details. Additional sources

can be found in several textbooks (Liddle,1997; Hammitt and Cole, 1998; Newsome eta/., 2002) and reviews of the literature (Cole,1987, 2002; Leung and Marion, 2000).

In this chapter, I do not distinguish between recreation and tourism. Fromthe point ofview of impacts to soils and vegetation, differences between the two seem negligible.Ecotourism suggests environments characterized by near-natural conditions, low levels ofdevelopment and crowding. Fortuitously, mostc:)f the literature on recreation impacts hasbeen conducted in such environments, makingapplication to ecotourism straightforward.

Hiking and Camping as Activities

Humans have walked and camped for as longas they have existed. Only in recent centuries,particularly in developed countries, has thereheen little need for large portions of the population to walk from place to place. In the pasthalfcentury, this trend hasreversed. As theproportion of people with substantial leisure timehas increased, people areturning to hiking andcamping as recreational activities (Fig. 4.1). Inthe USA, for example, two-thirds of the population engages in walking for pleasure and

''c, ( _AB Inti'rnation.il J0()4. tin ironnicnhil Imp.u ts ot tcotouri^m led. R. Buckley) 41

42 D.N. Cole

Rg.4.1. Recreational hiking and backpacking have increased dramaticaiiy in the past few decades.

about one-quarter hikes and camps (Cordelland Super, 2000). Increased Interest in ecotour-ism reflects this trend and its disseminationaround the globe.

Hiking has always been more ubiquitousthan camping, particularly in more developedand less remote places. In road-accessibleplaces, with well-developed infrastructure,most hiking may occur on highly engineeredtrails designed to absorb the impacts of hikingand to confine those impacts to the designedtrail system and nodes of activity (e.g. viewpoints, picnic sites, etc.). Most hiking is of shortduration, less than 1 day and often for just anhour or two, with tourists staying the night insome sort of lodging. In addition to staying inovernight lodging, many people camp in road-accessible developed campgrounds, whichideally are designed to confine traffic to surfaces that are hardened to absorb use. In thesesituations, impacts to soils and vegetation canbe limited despite very high visitation levels.Where people venture offthe trail system, however, impacts can be pronounced.

Less-developed and more remote areas areused in more variable ways. Day hiking onengineered trails still occurs, but overnighthiking on less-developed trails and even off-trail travel also occurs. In certain parts of the

world (e.g. much of Europe, Nepal and NewZealand), long-distance trekkers usually overnight in lodges or shelters, but, in many places,the tradition involves overnight camping.Camping may occur on designated campsites;informal, long-established sites; and even onplaces that have never been camped on betore.

The value of research on recreation impacts to scjils and vegetation seems generallygreater in less-developed and more remotelands. This has nothing to do with the relativeamount or importance of recreation in theseplaces. In less-developed and more remoteplaces, management is more complex, and theknowledge required to manage eftectively isgreater. Management relies less on engineeringand on separating the natural environmentfrom recreational use. Therefore, it is more critical to understand the inherent durability ol thenatural environment, and bow much of whattypes of use the environment can su/iport. Thestandards for acceptable levels ot impact arealso likely to he more stringent, and concernabout the obtrusiveness of management islikely tcj he greater. This management complexity, I think, explains the fact that although mostvisitation occurs c^n more developed lands,most research has been conducted in less-

developed parks and wilderness <ireas.

Impacts of Hiking and Camping on Soils and Vegetation 43

Historical Context of Research

Research on tlie ecol()t;it al impacts nt recreation has a short histors. Althoiij^li there were afew isolated earU studies ol the ecolot^icalimpacts of toLirists iMeinecke, 1^28) and ofvegetation sulijected to trampling (Bates,1935), the 19()()s was the dec ade when interest

in recreation impacts first decelo[")ed widely.Not coincidentally, it was the l')(it)s when thedemand for outdoor recreation first exidoded inmuch of the developed world. This earliestwork was descriptive, highK' site-S|X^cific,seldom published, and largeK' ccMifined to theUSA and western Europe. Few researchers everconducted more than one stuck'.

By the early 197()s, intcMcst had grownenough I'or collaborative and cumulativeresearch to be supported. The term 'recreationeccdogy', the most commoti descrifitor ofresearch on the c>nvironmental ellec ts of recre-

atic3n, was probably coined about this time. B\'1973, in Great Britain, the Rec reation EcologyResearch Grou|:) was convening regularlv toshare information. The lirst pionc>ers in recrcM-tion ecology also began work in the earlv1970s. Neil Baytield (1 971, 197 1, 1979) developed the first sustained [Kogramme of recreation ecology research, a 20-year [programme ofgovernment-lunded work on trampling andfootpath impacts in the mountains of Scotlandand England. He was among the first to [proposemethods for monitoring trail im[iacts and toinvestigate means ol restoring damaged recreation sites. Michael Licldle began a liletime (4'work in academia on rec reation impacts, first inGreat Britain (Liddle and Greig-Smith, 1975)and later in Australia (Liddlc> and Kay, 1987).Notably, Liddle (1975a,b) was among the firstto search for generalities about recreationimpacts and his carec>r culminated in a comprehensive textbook on recreation ecologv(Liddle, 1997).

The earliest students of rec reation ecologvin the USA did not pursue c<ireers in thc» field.Nevertheless, their c ontributions were vital. Al

Wagar conduc ted the lirst simulated tramplingexperiments, and |)rovided initial conceptualflevelopmeni ol the carrying capacitx cone c^jitfWagar, 19()4). Sid Irissell conduc tcnl the lirststudy of campsites that rec eived diflering levelsof use (Erissell and Uuncan, l*)()Si. This

research showed that impact occurs whereveruse occurs, leading Frissell to suggest that thedecision facing recreation managers is howmuch impact is acceptable - not whether or notto allow impact. This obserx'ation provided theconceptual foundation for planning processessuch as the Limits of Acceptable Change(Stankey et a/., 1985). Frissell's data also Illustrated the curvilinear nature of the relationshipbetween amount of use and amount of impact,although it was another 15 years before thegenerality of this finding and its significance torecreation management was articulated (Cole,1981a). Frissell (1978) was also among the firstto publish suggested methods for monitoringwilderness campsites.

Efforts to develop generalities and the management implications of recreation ecologywere substantially increased when governmental research institutions hired recreation ecolo-

gists. Since the late 1970s (Cole, 1978), myposition with the US Forest Service has allowedme to focus my professional work on recreationecology, jetfMarion has held a similar positionwith the National Park Service (now the USGeological Survey) since the mid-1980s. Thishas provided the opportunity for more rigorousstudy of recreation ecology. Ithas been possibleto use multiple methodologies to examineimpacts (Marion and Cole, 1996), to developmodels of factors that influence impacts (Cole,1987, 1992), to search for generality across different environments (Cole, 1995a), to studytrends over time (Cole, 1993) and to work at

multiple spatial scales (Cole, 1996). It has alsoprovided more opportunity to apply researchresults to the development of management strategies (Cole, 1987, 2002; Hammitt and Cole,1998; Leung and Marion, 2000) and monitoringtechniques (Cole, 1989a; Marion and Leung,2001).

The geographic distribution of recreationecology research has also expanded. Prior tothe 1980s, recreation ecology research waslargely confined to North America and Europe.Research continues to be conducted throughout Europe, but nowhere is recreation ecologyan established discipline. Occasional studieshave been conducted in lapan since at least thelate 19(g)s (Tachibana, 1969) and that tradi

tional continues todav (Yoda and Watanabe,2000) - there and in Hong Kong (jim, 1987;

44 DM Cole

^ .V

rWit »• eV*

w ^

Fig. 4.2. Much of the research in recreation ecology has been conducted in mountainous environments.

Leung and Neller, 1995). In the 1980s, researchexpanded in developed countries around theworld, most notably in South Africa (Garland,1987) and Australia. Notable in Australia is the

work of Liddle and his students (Liddle andThyer, 1986; Sun and Liddle, 1993a,b) andresearch related to management of WorldHeritage Areas in Tasmania (Whinam et al.,1994; Whinam and Chilcott, 1999) and theGreat Barrier Reef(Liddle and Kay, 1987).

In the 1990s, perhaps in response toincreased ecotourism and recognition of itspotential environmental consequences, recreation ecology research has expanded into developing countries and ecotourism destinationsaround the globe. Recent studies have beenconducted in the Middle East - in Israel (Kutieland Zhevelev, 2001) and Egypt (Hawkins andRoberts, 1993) - as well as in the tropics - inCentral and South America (Boucher et al.,1991; Earrell and Marion, 2001a), Africa (Obuaand Harding, 1997) and South-East Asia (Jusoff,1989). It has expanded throughout the temperate lands of the southern hemisphere - in NewZealand (Stewart and Cameron, 1992) and inChile (Earrell and Marion, 2001b) - and eventhe sub-Antarctic (Scott and Kirkpatrick, 1994).Much of this generation of research has drawndirectly from the research techniques andprotocols developed bythe original generationof recreation ecologists. Buckley and Pannell

(1990) applied the findings of recreation ecology to ecotourism and Tracy Earrell applied JeffMarion's impact monitrjring procedures inCentral and South America (Earrell and Marion,2001 a,bj.

The ecosystems in which recreation ecology research has been conducted has expandedalong with the geographical distribution ofstudies. The earliest work occurred in moun

tainous and coastal environments, due to theattraction of tourists to these locations (Eig. 4.2).To this day, the preponderance of work is stillconducted in the mcjuntains and, to a lesserdegree, along coasts. Although the earliest workin the mountains was typically in the alpine andsubalpine zones, recently more research hasbeen conducted at lower elevations (e.g. EHalland Kuss, 1989; Leung and Marion, 1999a).Much of the recent coastal work has shifted to

recreational impacts on reefs and intertidalareas (Liddle and Kay, 1987; Elawkins andRoberts, 1993; Rou[3bael and Inglis, 2002).Other environments recently studied includeriparian (Marion and (a)le, 199()) and desertenvironments (Cole, 19tU)).

Research Designs

Eour different research designs have beenemployed as a means of studyifig rec reatiotial


impacts (Cole, 1MH71. I ach ot these (iesie,ns hasstrengths and weaknessi'^. The \akiahle perspective of each clesigti is ri'lietted in the factthat each was used in earl\ rei realion t>c c)log\research and each continues to hc> used toda\'.

The most common design, [larticularK inhighly a[:)pli(>d research, designed to assessimpacts to an entire park, campgroimd or trailsystem, is the desc ripti\'e field sur\e\. \'egeta-tion and soil paramcMers on rec reation sites aremeasured for the [Hir|)()sc> of assessing ccirreritconditions. Environmental acid usc> c haracteris

tics are otten simultaneousK' assessc>d and then

correlated with varitition in impac ts to soil andvegetation. E.xamples of tliis afif^roach includeBayfield's (1 97T) work on Scottish trails, as wellas the work ot Marion and his students on trailsand campsites in the eastern USA and inCentral and South America da'ung and Marion,1999a,b; Farrell and Marion, 2t)0]a,hi. Thevalue of this approach is that im|xu t conditionscan be surveyed over large areas rapidK' andwith minimal training. Surveys provide a snapshot of conditions at a point in time and, whenrepeated, can be used to assess trcaids ovcytime. ConsecjLiently, such studies can providemuch ot the toundational intormation neededto guide day-to-day management. However, ifone's goal is to understand cause-and-eftect,this is the least useful of the researc h designs.One can speculate about cause and effect fromcorrelational analyses, but apparent relationships can be s|nirious and true relationshipscan be missed due to (be contounding of intervening variables.

A common variant of the descriptivesurvey is the addition of measures taken onundisturbed control sites that, when comparedwith recreation sites, provide an estimate ofchange resulting I'rom recreation use. Thisamounts to using spatial dilTerences (usedversus unused) to inter temporal change (pre-versus post-use). In sue h studic^s, it is commontfj compare im|)ac ts on catc^gories of sites thatvary either in use or environmcmlal charac tcM is-tics. An early example is Erissell atui Duncan's(196S) study of variation in impac t, related toamc:)unt of use, on canoe campsites. Thisapprc^ach, though more time-consuming thanthe simple descriptive survey, has the advantage cjf providing an estimate* of the extent towhich conditions rellect rec reatiotial use*.

However, control sites are never perfect replicates of pre-existing conditions and, in somesituations, the difficulty of finding good controls makes it impossible to use this approach.

A further variant of the descriptive fieldsur\'ey is the before-and-after natural experiment. This design involves assessing conditionsbefore and after recreational use occurs, or

before and after a change in managementregime. Ideally, identical measures are taken oncontrol sites that are not subjected to use or achatige in management. In this case, changeresulting from management is measureddirectly. An early example of this approach isMerriam and Smith's (1974) study of impactsresulting from initial use of newly openedcampsites. Spildie et a/. (2000) used this designto assess the effectiveness of a managementprogramme designed to confine and reducecampsite impacts associated with packstock.Typically, such studies are conducted in oneplace at one point in time. Consequently, it canbe difficult to assess the general applicabilityofresults.

The three variants of the descriptive fieldsurvey have the advantage of realism and providing highly relevant site-specific information,but they all suffer, to varying degrees, in theirability to identify cause-and-effect and to contribute to general knowledge. Thealternative isthe simulated experimental approach. With thisapproach, researchers carefully control useandenvironmental factors in a replicated designthat maximizes insights into cause and effect.Ba)'field (1971) was perhaps the first to employexperimental trampling by humans, althoughWagar (19(il) trampled vegetation using anartificial 'tamp'. More recently, Cole and Bay-field (1993) developed a standard protocol torconducting trampling experiments. This protocol has been applied in many different vegetation types, from mountainous areas of the USA(Cole, 1995a) to such places as Arctic tundra(Monz, 2002), sand dunes in France (Lemauvieland Roze, 2003) and forested communities inLtganda (Pratt, 1997). Widespread applicationof similar field techniques increases the abilityto develop broad generalizations and to understand the causes of variability.

Each of these re.search designs has inherent strengths and weaknesses. The most appro-priatc* approach to take will depend on the

46 D.N. Cole

goals of the study. Maximum insight can begained byutilizing several approaches simultaneously. For example, Marion and Cole (1996)combined; (i) descriptive field surveys ofcampsites, stratified according to amount of use andvegetation type, along with measures taken onadjacent controls; (ii) natural experiments onpreviously undisturbed sites, before and afterbeing opened for camping; (iii) natural experiments on established campsites, before andafter being closed to use, as well as before andafter management actions designed to reducecampsite size; and (iv) trampling experiments.

Progress in recreation ecology is hampered by minimal attention given to conceptualand theoretical development. Early exceptionsinclude Liddle's (1975a,b) conceptual model oftrampling processes and his hypothesis thattrampling tolerance is related to primary productivity. Cole's (1992) simplified model ofcampsites represents one of the few attempts touse analytical models to build foundationalconcepts regarding how various factors operate'u etermining impact magnitude. Rigorousana yses of the efficiency of impact assessments

'̂̂ '̂ ough Leung and Marion9c) is a notable exception.

Research Results

Descriptive information about recreationalimpacts can bedivided into information aboutthe nature and magnitude of impacts caused by• erent recreational activities, spatial aspects

0 impacts, and temporal patterns of impact.1here is also an extensive body of informationaDout use and environmental characteristicshat influence the nature and magnitude ofimpacts. This knowledge provides the basis forinsight into management actions that mighteffectively control impacts. Finally, a substantial amount of work has developed regardingthe effectiveness of impact management techniques, as well as efficient ways to monitorimpacts.

The nature and magnitude of impacts

Much of the research into hikingand campingimpacts on soil and vegetation is focused on

either linear travel rcjutes, osLjally trails, ornodes of concentrated rise, osoally cam[asitesbut also picnic sites and viewpoints. The othertradition has been tf) stody the etiects ot trampling, which occurs on trails and campsites butalso away from these places ot concentrateduse.

Trampling has at least three eltects: abrasion ofvegetation, abrasion ot organic soil horizons and compaction ot soil (Fig. 4. 5i. Plantscan be bruised, crushed, sheered ott cmd evenuprooted by trampling. Tram[Dling ettects in-clucJe reductions in plant height, stem lengthand leaf area, as well as in the numl;)er ot [plantsthat flower, the number ot tlower heads [oer plantand seed prcjcluction (Liddle, 1*^19/). Reducedheight and leaf area decrease thephotosyntheticarea of plants, resulting in de[)leted carbohydrate reserves (FHartley, 1999). These changestypically result in reductions itn plant vigour andreproduction. Many plants are killed i^y trampling. At moderate levels ot tram[)ling, however,some species increase in abundance, otten as aresult of decreased competition or a change inmicrohabitat. Generally, where tramfiling isintense, plant cover and hiomass are low, mostplants are short, species richness is reduced andspeciescompcjsiticjn has shitted.

Trampling compacts soils, reducing porosity, particularly the volume ot macro[3ores(Monti and Mackintosh, 1979). This r(»duces thewater-holding capacity ot soil, except in somecoarse-textured soils, (aimi^action reduceswater infiltration rates, leading to increasedrunoff and erosion potential. These physicalsoil changes alter soil chemistry and biota,although such changes are poorly understood.Compactedsoils can also inhibit seed germination and plant growth. Alessa and Earnhart(2000) have shown that plants in compactedsoils may be less able to utilize available nritri-ents because they grow lewer lateral roots androot hairs and because cytoplasmit streamingwithin root hairs is reduced. Soil tomfiac tioneffects are exacerbated l)y abrasion and loss otorganic soil horizons, which shiekl underlyingmineral soil horizons from excessive compaction and erosion.

Loss of organic litter direc tly atlec ts plantand animal populations, both above and belowthe grcjund. Since certain [dant spec ies germinate mcjst frec|uently on organic soil surtac es.


Trampling

Abrasion of

vegetation

Abrasion of

organic matterCompaction

of soil

Reduction in

litter cover

Reduction in

air and water

permeability

Reduction inReduction in

plantreproduction

plant vigour

YY Y

Change in Reduction in

species vegetativecomposition cover

Cfiange insoil biota

Increase in

runoff and

erosion

Fig. 4.3. A conceptual model of trampling impacts. Note tfie numerous reciprocal and cyclicrelationships.

loss of liller can cause s[)(h i(>s composition toshift towards s|)ecies that j^erminate most fre-c|uently on mineral soil. Loss of organic matterfrom the soil typically reduces the water-holding capac ity ol thc> soil and has an adverseeffect on soil microhial [lopiil.itions, whichdepend on soil organic matter and root e\u-dates from ahove-groimd plants tor theirenergy. Z.ihinski .ind Ciannon (l')<)7) reportsuhsttintial reductions in tht> funttional diver

sity of mi( rohial populations on a hackcountrycampsite. Microhial popukitions contrihute toecosystem func tioning h\' metal^olizing nutrients, transforming soil organic matt(>r, [producing phytohormones and contributing to soilfcpod webs.

The imp.u Is ol cam|)ing inc ludc^ all theeffects of tr.impling, as vvc>ll ,is some unir|uc>

impacts. Numerous studies have quantified themagnitude of soil and vegetation impact oncampsites. The data in Table 4.1 are typical.They describe vegetation and soil conditionson 29 paired canoe-accessible campsites andundisturbed control sites in low-elevation riparian forests in the eastern USA (Marion and

Cole, 199()). On mcpst camfpsites, most of thevc^getation has been eliminated and the vegetation that remains consists primarily of grami-noids. Forbs dominate undisturbed control

sites. Organic horizons on campsites are onlyalioLit one-third as thick as on controls; mineral

soil is exposed over most of the campsite. Thesemineral soils are compacted - exhibitingincrcMsed bulk density and penetration resistance. Substantial numbers ol trees have beendamaged (c cit branches or scarred trunks) or

48 D.N. Cole

Table 4.1. Vegetation and soil conditions on 29 campsites and undisturbed control sites at DelawareWater Gap National Recreation Area, 1986 (from Marion and Cole, 1996).

Campsite Control

Impact parameter Mean Range Mean Range P

Ground vegetation cover (%) 15 0-63 72 1-95 0.001

Floristic dissimilarity (%) 75 23-100 Not applicableGraminoid cover (%) 58 0-100 26 0-92 0.023

Forb cover (%) 23 0-78 59 5-100 0.001

Mineral soil cover (%) 61 21-94 1 0-15 0.001

Organic horizon thickness (cm) 0.5 0-1.4 1.5 0.2-3.1 0.002

Soil bulk density (g/cm^) 1.26 1.0-1.4 1.06 0.7-1.4 0.001

Soil penetration resistance (kPa)^ 275 137-382 49 0-226 0.001

Soil moisture (g/cm^) 18 8-32 17 8-31 0.710

Felled trees (%) 19 0-53 Not applicable

Damaged trees (%) 77 25-100 Not applicable

Tree reproduction (stems/ha) 936 0-6275 10,090 0-56,400 0.001

Non-vegetated area (m^) 181 0-696 0 0-15 0.001

Campsite area (m^) 269 51-731 Not applicableShoreline disturbance (m) 9 0-20 Not applicable

^ 1 kPa=the pressure corresponding to 1.01971 x 10 ^ kg/cm^

felled, and tree reproduction has been dramatically reduced. Along with the felling of treesaplings, lack of tree reproductionsuggests thatoverstorey trees will not be replaced on campsiteswhen they eventually die.

Camping also can cause off-site impacts.The most common off-site impacts are informaltrailing (between the campsite and watersources, other campsites or the main trail) andimpacts caused by the collection of wood to beburned in campfires. Hall and Farrell (2001)documented 25-63% reductions (dependingon size class) in abundance of woody materialon and around campsites. Taylor (1997) foundthat the density of saplings around campsiteswas reduced within an area that extended 45 mon average from thecentre ofthecampsite. Themost pronounced off-site impacts are oftenthose associated with the confinement of

horses and other packanimals used to transportpeople and gear (see Newsome ct a!., Chapter5, this volume).

Impacts on trails have also been studied.However, it is difficult to separate the impactsof hiking on trails from the impacts associatedwith trail construction and maintenance, andthe impacts that would occur on trails in theabsence of hiking (e.g. erosion by rainwaterchannelled down a trail tread). Major impacts

of trail construction and maintenance include

opening up tree and shrub canopies, the building of a barren, compacted trail tread that mayalter drainage patterns, and the creation of avariety of new habitats, including cut slopesabove the trail and fill below (Cole, 1981b).Except where hiking use is extremely high, it isprobably rare for the impacts of hiking on trailsto exceed the impacts caused by trail construction. However, these rare cases of profoundhiking impact can be highly problematic. Forexample, the deep, peaty soil of tracks in muchof the Tasmanian Wilderness World FHeritageArea can be churned into deep quagmires by asmall number of hikers (Calais and Kirkpatrick,1986; Whinam and Chilcott, 1999).

Impacts adjacent to trails are similar tothose caused by trampling. Although tramplingadjacent to trails can reduce vegetation cover(Cole, 1978; Boucher ct a!., 1991), it is common for vegetation cover to be greater adjacentto trails than on undisturbed sites (Hall and

Kuss, 1989), presumably due to increasedlight, water and nutrients there. Organic mattercan decrease and soil compaction increase(Adkison and Jackson, 1996). Vegetation composition adjacent to trails is usually very different from undisturbed site controls. It c an be less

diverse (Boucher etai, 1991 i, but often is more


diverse (Hall and Kuss, 198')i, [Dartially due tothe invasion of exotic species that use trails asconduits for movement (Benninger-Truax ef a/.,1992).

Of more practical significance and concern is the impact of hiking on the constructedand maintained trail surl'ace. Constructed trails

are barren and compacted by design. So, theinterest here is not impacts on native soil andvegetation but impacts on the trail itself.This isa concern because hikers can increase soil ero

sion from trails, either by detaching or transporting soil particles. Two recent experimentalstudies provide insight into the process bywhich this occurs. They show that sedimentyield and trail erosion is detachment-limitedrather than transport-limited (Wilson andSeney, 1994; DeLuca ct a/., 1998). Trail useloosens soil particles, making them easier todetach and, therefore, available to be trans

ported by such erosive agents as running water.Most trail-impact studies document trail

characteristics, such as width and depth, without regard for the comfjlex factors (of use, environment and management) that combine toinfluence these characteristics. Bayfield andLloyd (1973) developed survey techniques forperiodically assessing trail width and depth, aswell as censLising the presence or absence of'detracting' features, such as rutting and baddrainage. Coleman (1977) developed a technique for measuring trail cross-sectional area.More recent assessments of trail conditions, in

such places as Guadalupe Mountains NationalPark (Fish et dl., 1981), the Selway-BitterrcwtWilderness, (Cole, 1983) and Great SmokyMountains National Park (Leung and Marion,1999b), are largely extensions of this earlywork. These studies provide descriptive statistics (means and ranges) tor such metrics as trailwidth and depth, as well as frequency andextent of trail [iroblems (Bayfield's 'detracting'features). For example, mean trail width anddepth were 115 cm and 10cm, respectively, ontrails in the Selway-Bilterroot Wilderness (Cc^le,1991). On trails in Great Smoky MountainsNational Park (Leung and Marion, 1999b) therewere 470 occurrent es of multi|;)le tread. A totalof 10.3 km of trail (1.8"/., ot the trail system) hadmultiple treads. These studies typically searchfor correlations between trail conditions andcharacteristics ot use, environtnent and man

agement. For example, in Great Britain, Bay-field (1973) found that trailwidth was positivelycorrelated with soil wetness, roughness andsteepness, and Coleman (1981) found that trailwidth was positively related to recreation use.

The most significant impacts of hiking onnative soils and vegetation are probably thoseassociated with proliferation of user-createdtrails along hiking routes where a trail tread isnever constructed. Lance etal. (1989), describethis process in Scotland, noting that trail development usually starts with formation of a singletrack. As this path widens and erodes, secondary paths are created. These widen and mergewith other paths, ultimately creating a braided,eroding web (Fig. 4.4). On the tallest peaks inColorado, user-created trails to the summits

have eroded so severely that they are nowbeing replaced by constructed trails. Restoration of abandoned sections of user-created trail,which are often steep and eroding, is difficult(Ebersole ef a/., 2002).

Spatial patterns of impact

Most studies of impact report the intensity ofparticular types of impact - the amount ofimpact per unit area (e.g. the campsitelost 50%of its vegetation cover). Assessments of magnitude of impact must also consider the area overwhich this impact occurs. The magnitude of a50% cover loss on a 1000 m- campsite is twicethat of a 50% cover loss on a 500 m- campsite- although the intensity of impact is the same.Magnitude of impact (sometimes referred to asaggregate impact) is minimized when both thearea of impact and the intensity of impact perunit area are minimized (Cole, 1981a). Certainimpact parameters only describe impact intensity (e.g. vegetation cover loss), while othersonly describe area of impact (e.g. campsitearea). A few parameters describe both. Forexample, the area of vegetation losson a campsite (Cole, 1989b) expresses vegetation loss, inm-', as the product of campsite area and the difference between vegetationcoveron the campsite and an adjacent control site. This metric-makes it possible to compare the magnitude olvegetation impact on sites that vary greatly insize (e.g. Marion and Farrell, 2002).

Spatial aspects of impact have received

50 D.N. Cole

' - " • - ^ ' ' '

V •fc;r.<-r-w; -^V...:

• .^.i«5*2i*. J.V : ' I

little attention, beyond recognition that assessments of the magnitude of impact must consider the area that has been impacted, as wellas the intensity of impact. In addition to theintensity and aggregate area (magnitude) ofimpact, other potentially important descriptorsof impact include the size of impacts and thespatial distribution (pattern) of impacts. Given aconstantaggregate area of impact, there may hemany small impacts or a few large impacts.Theoretically, these impacts can he distributedin a pattern that is either more clumped (aggregated orunderdispersed) or more regular (over-dispersed) than a random pattern. In reality,spatial impact patterns are almost always moreclumped than random. Campsites are clusteredin campgrounds or around lakes and in placesaccessed by trails. Hiking impacts are concentrated along trail corridors, with little impact offtrail.

Quantitative descriptions of impact varywith the spatial scale of analysis that isselected. For example, vegetation loss may he1007o at the centre of a campsite hut only 507owhen the entire campsite is surveyed. At thescale of a lake basin, vegetation loss associatedwith camping might amount to only 1 or 27oand, at the scale of the park or wilderness, lessthan ]"/(, of the vegetation is likely to he lost(Cole, 1981h). Impacts might he consideredfew and large at a 10 ha scale hut many and

Fig. 4.4. Trail braiding is a common type of trail impact in certain environments.

small at the scale of 10,000ha. They may beregularly distributed at a lOha scale butclumped at the scale of 10,000ha. What thissuggests is that any quantification ol im|)acts isonly valid at the chosen scale of analysis.

Although generally ignored, spatial descriptors of impact and scaling issues are important considerations, particularly in assessinghow much of a problem impacts are, and indevising strategies for managing them. Cole(1981 b) noted that hiking and camping impactson soil and vegetation, while severe whenmeasured at small scales, are minimal at largespatial scales. This suggests that while recreation impacts can be serious for individualplants and animals, and perhaps localized rarepopulations, they are generally of little signiti-cance to landscape integrity or regional bioticdiversity. Moreover, unless much of a population is impacted by a single im[aacted site, tbeintensity, size and distribution ol im[)acts arenot relevant to the significance ot imfaactsassessed at large spatial scales. II animal populations are considered, however, spatial [)at-terns in which imfjacts are clustered, leavinglarge expanses undisturbed, might l:>e th(> ideal.

Recreation impacts on soil arid vegetationare highly significant at the stale of human perception - the scale humans can re.idily observe.Studies r)f wilderness campers show th«it mostcampers view small areas of impac t as 'posi-


Effective closure of campsites Low resilience

environment

o

E<

Several yearsafter initial use

Campsite is first used

High resilienceenvironment

Time

Fig. 4.5. The typical life history of a campsite, from initial use through a period of closure and recovery.

five', 'pretty natural, healthy' (Farrell ct a/.,2001), liecause ihey make the site lunctlon wellas a temporary dwelling lor humans. Perhapsfrom the human perspective, many smallimpacts are [3referal)le to a fcnv large impacts,because small imfiacts are perceived as'healthy' dwelling sites, while large impactedareas (several hectares or more) suggest abuse,damage and unhealthy conditions. Moreover,dispersal ol im|:)acts at this scale provides moresolitude and privacy lor tourisls. This line ofthinking leads to the conclusion that, whenimpacts cjn soils, v(>getaiion, animals andhumans are all considered, they are least problematic when; (i) aggregate^ impact (intensity incombination with area) is minimized; and (ii)impacts are concentrated al the site scale, dispersed at intermediate scales (within a destina-ticjn area such as a lake basin) and c luslered at

larger scales (within a [xirk or wilderness)(Hammitt and dole, 1008). Although little attention has l)een devoted to these spatial issues,Leung and Marion (1 OOOd) suggest some Sfiatialstrategies lor managing impacts.

Temporal patterns of Impact

The tendenc y to sludv impac Is at on(> point intime has contributed to a lac k ot data on tem|-)o-

ral patterns of impact, much as the tendency toconduct studies at just one spatial scale leavesus with little insight into spatial patterns.Available studies suggest that individual campsites have a typical 'life history', movingsuccessively through stages of development, dynamicequilibrium and recover)' (Fig. 4.5). Impactoccurs rapidly during the development phase,shortly after a campsite is first used. For example, on newly established canoe campsites,most of the impact that occurred over the 6years following creation of the campsiteoccurred during the first year of use (Marion andCole, 1996). Impact did increase over the first 3years, but at a decelerating rate. This phase isfollowed by a more stable phase in whichimpacts change little unless there are dramaticchanges in amount of use. For example, onlong-established campsites in the Eagle CapWilderness, mean vegetation cover was 15% in1979, 127n in 1984 and 19% in 1990 (Cole and

Flail, 1992). Vegetation cover on these campsites might be expected to fluctuate betweenabout 10"/o and 20%, as long as use characteristics are relatively stable. These patterns are relatively consistent across diverse ecosystemtypes and typesof recreation, although impactsoccur more rapidly (the development phase isshorter) as amount ot use increases and site dur-abilitv decreases. Moreover, aberrant behaviour

52 D.N. Cole

Frequencyof use

Type andbefiaviour

of use

Season

of use

Environmental

conditions

Spatial distributionof use

Intensity of impact Area of impact

Total impact

Fig. 4.6. Factors that influence the intensity and area of impact and, therefore, the total amount ofimpact.

(e.g. someone cutting down a tree) can causedramatic spikes in impact at any time.

The recovery phase is almost invariablylonger than the development phase, becausedeterioration occurs more rapidly than recovery. Recovery rates also varygreatly with kindsof impact, magnitude of impact and environment. Variation in the resilience of different

ecosystem types is pronounced. Hartley (1999)reports residual effects of trampling after 30years, in alpine meadows in Glacier NationalPark, while most evidence of camping onclosed riparian campsites disappeared within 6years (Marion and Cole, 1996). Cole and Monz(2002) report that an alpine grassland trampled1000 times recovered more rapidly than aneighbouring forest, withan understorey of lowshrubs, that was trampled just 75 times. Giventhe same environmental setting, sites thatreceive more use and that are more heavilyimpacted will take longerto recover.

Temporal patterns at larger spatial scaleshave generally been ignored. They are particularly important, however, because impacts tendto proliferate and spread across the landscapewhere use distribution is not tightly controlled.Forexample, in two drainages in the Eagle CapWilderness, the number of campsites increasedfrom 336 in 1975 to 748 in 1990 (Cole, 1993),even though the condition of most of the sitesthat existed in 1975 changed little between1975 and 1990. Site proliferation occursbecause, as use shifts across the landscape,new campsites appear more rapidly than oldcampsites disappear.

Temporal patterns on trails and hikingroutes are likely to be similar, though they haveseldom been studied. Trail impacts occur rapidly; most segments on established trail systemsare generally stable (Fish ct a/., 1981; Cole,1991); and recovery of closed trails is typicallyslow, except where it is assisted (Eagen et al.,2000). However, trail segments that are poorlylocated or inadequately designed and maintained may deteriorate substantially. At largespatial scales, impacts have increased over timedue to: (i) lack of recovery on re-routed trailsegments; and (ii) the pioneering of routes intotrailless places. This latter trend can be particularly problematic because development of atrail makes access easier, which can lead to acycle of ever-increasing use and impact.

Factors that influence magnitude ofimpact

The types of research that have probably beenmost useful to management are studies of thefactors that influence the magnitude ot impacts-why impacts are minor in some situations andsevere in others. The principal factcirs that influence intensity of impact (Fig. 4.6) are: (i) frequency of use; (ii) type and behaviour of use;(iii) season of use; and (iv) environmental conditions, while area of impact is primarily aresult of the spatial distribution of recreationuse (Cole, 1981a, 1987). An understanding ofeach of these influential varial")les suggestsstrategies for managing the inqiac ts ot hiking


Amount of use

Fig. 4.7. The relationship between amount of useand amount of impact is curvilinear (asymptotic).

and camping on soils and vegetation (Cole etal., 1987; Marion and Leung, Chapter 13, thisvolume).

The relationship between frequency of useand intensity of impact is generally asymptotic(Fig. 4.7). At first, small increases in use frequency cause pronounced increases in impact;however, the rate of increase in impact decreases as use intensity increases. Where use islight, sites that receive even small differences inamount of impact can have very differentimpact levels. However, where use is heavy,sites that receive substantiallydifferent amountsof use may have similar impact levels. Frisselland Duncan (1 965), the first researchers to document this relationship in a field situation, concluded that 'if any use is to be allowed in thewilderness areas, some immediate loss of thenatural vegetation will have to be tolerated' (p.258). Similar results have been found in numerous field surveys of recreation sites and inexperimental studies. The further implication ofthis relationship is that the magnitude ofimpacts can usual ly be minimized by encouraging the repetitive use of as small a number ofsites as possible (i.e. concentrating use) (Cole,1981a). This strategy involves accepting a slightincrease in the intensity ot impact to realize thebenefits of a large decrease in the area ofimpact.

The type and behaviour of use can alsohave a profound elfect on both the type andmagnitude of impact. For example, camperswhfj build fires cause both more and differenttypes of impact than campers who do not huildfires. Several studies have compared the im

pacts of hikers with those of groups who usehorses or llamas for transport. Generally, thesestudies have found that horses cause more

impact than hikers or llamas, which causeequivalent levels of impact (Cole and Spildie,1998; DeLuca etal., 1998). Recreationecologyresearch has provided the scientific foundationfor minimum-impact educational programmes(Cole, 1989c). These programmes teach techniques of trip planning, route selection, hikingbehaviour, campsite selection and campingbehaviour that minimize the per capita impactsof use.

Season of use is a less critical factor for

hikers than it is for horses and heavy pack animals that can cause severe damage to trailsandmeadows when soils are water-saturated and

plants are growing rapidly. During seasonswhen snow banksare melting, hikers also needto avoid walking offtrail andon water-saturatedsoils.

A substantial body of research has developed regarding characteristics that make different environments more or less durable ascampsites or as trail locations. Experimentalapplications of both trampling (e.g. Bayfield,1979; Cole, 1995b) and camping (Cole, 1995c)have been particularly insightful in buildingthis knowledge. Field surveys of trails andcampsites that develop correlations betweenimpact parameters and environmental variables have also been helpful (e.g. Leung andMarion, 1999a,b). Experimental studies showthat some vegetation types can tolerate morethan 30 times as much use as others, with nomore damage (Cole, 1995a).

Experimental studies suggest that there isan important difference between a site's resistance (its ability to tolerate use without beingdamaged) and its resilience (its ability torecover from damage). Cole (1995b) hasshown, for groundcover plants, that resistancedecreases with erectness and that broadleavedherbs are typically less resistant than grass-likeplants and shrubs. Herbs growing in shade areparticularly intolerant of trampling becauseadaptations to shading - possession of large,thin leaves and tall stems - make these plantsvulnerable when trampled. This explains thecommon finding that trampling offorested sitesgenerally results in more rapid loss of vegetation than trampling of open woodlands or

54 DM Cole

m

' v-""v:* ',-v7r--r', '^' •*

-;^d

Fig. 4.8.The areaofvegetation loss on this campsite is small, due to the durability of the graminoidvegetation cover.

meadows. Low shrubs, such as heather, are relatively resistant to trampling stress, but theirresilience is low. Once damaged, they recoverslowly. Grass-like plants are most tolerant oftrampling.

At the risk ofovergeneralizing about a verycomplex subject (refer to reviews in Cole,1987; Liddle, 1997; Hammitt and Cole, 1998;and Leung and Marion, 2000, for furtherdetails), a few conclusions about site durabilityseem warranted. Characteristics of durable

campsites and other nodes of concentrated useinclude: (i) either lack of groundcover vegetation or presence of resistant vegetation (Fig.4.8); (ii) an open, rather than closed, treecanopy; (iii) thick organic soil horizons; or (iv)a relatively flat but well-drained site. Marionand Farrell (2002) also note the importance ofdesigning campsites to confine impacts to asmall area, in the absence of natural featuressuch as rocky terrain that serve this purpose.

Leungand Marion (1996) provide a usefuloverview of knowledge regarding how environmental characteristics influence trail condition.

Terrain and topography have a major influenceon trail conditions. Steep trail slopes, steep sideslopes and trail alignments in which the traildirectly ascends slopes all tend to be moredegraded, usually because more water is chan

nelled, with more force, clown the trail tread.Trail problems are also common where soils arefine-textured, stone-tree and hrjmogeneous, orhighly organic and where soils are poorlydrained or have high water tables. Trails alsotend to widen where the grcjunci surface is wetor rough (Bayfield, 1973).

Management and monitoring

Management and mcjnitcjring ol trails andcampsites are covered in detail in Leung andMarion (Chapter 14 this volume). The scientificfoundation for knowledge about etiective management strategies was derived from hundredsof studies of the nature and magnitude ofimpacts, and how they are influenced by characteristics of use and the environment. Alongwith the experiential knowledge^ developedfrom decades of implementing ret reation management programmes, a wide array ot etiectivemanagement strategies has evolved (Hammittand Cole, 1998). Similarly, decades ot recrea-ticjn ecology research, developing methods otmeasuring impact, have caintributed to tht^campsite and trail monitoring methods employed today (Cole, 1983, 1989a; M.trion,1991; Leung and Marion, H)99b).


Conclusions and Future Directions

Although the field of recreation ecology is onlyabout 30 years old, somewhere around 1000studies have been conducted. A majority ofthese have focused on the impacts of hikingand camping on recreation and soils. Specificdetails about the nature, magnitude and spatialaspects of impact vary with the context of everysituation (with amount and type of use, environment, management, etc.). In addition, themanagement objectives of every park, wilderness or other tourist destination also vary.Therefore, in every place where recreationimpacts are a concern, it is worthwhile to haverecreation ecology studies conducted in thatarea, so results can be interpreted in referenceto the specific context and management objectives of the area. However, in the absence of

site-specific studies and intbrmation, muchinsight can be gleaned from generalizationssuggested by the recreation ecology literature.

Since the late 1 970s, there have been sev

eral attempts to synthesize the recreation ecology literature. Each attempt, including this one,is somewhat unique but there is substantialconsensus as well. The lollowing five generalizations are among the most important andgenerally agreed upon.

1. Impact is inevitable with repetitive use.Numerous studies have shown that even verylow levels of repetitive use cause impact.Therefore, avoiding impact is not an optionunless all recreation use is curtailed. Managersmust decide on acceptable levels of impact andthen implement actions capable of keeping useto these levels.

2. Impact occurs rapidly, while recoveryoccurs more slowly. This underscores theimportance of proactive management, since itis much easier to avoid im[oact than to restoreimpacted sites. It also suggests that relativelypristine places should receive substantialmanagement attention, in contrast to thecommon situation of focusing most resourcesin heavily used and impactcb [daces. Finally,it indicates that rest-rotation of sites (periodically clcjsing damaged sites, to allow recovery,before re-opening them to use) is likely to beineffective.

3. In many situations, im|)act increases more

as a result of new places being disturbed thanfrom the deterioration of places that have beendisturbed for a long time. Thisalso emphasizesthe need to be attentive to relatively pristineplaces and to focus attention on the spatial distribution of use. It suggests that periodic inventories of all impacted sites is often moreimportant than monitoring change on a sampleof established sites.

4. Magnitude of impact is a function of frequency of use, the type and behaviour of use,season of use, environmental conditions, andthe spatial distribution of use. Therefore, theprimary management tools involve manipulation of these factors.

5. The relationship between amount of useand amount of impact is usually curvilinear(asymptotic). This has numerous managementimplications and is also fundamental to manyminimum impact educational messages. Itsuggests that it is best to concentrate use andimpact in popular places and to disperse useand impact in relatively pristine places.

New insights into recreation ecology havebeen generated as researchers have adoptedmultiple methodologies and expanded both thetemporal and spatial scales of analyses.However, further progress is hampered by alack of theory and conceptual thinking. Nowthat the field is 30 years old, the time seems ripefor conceptual and theoretical work that canbuild a framework for organizing the knowledge gained from the multitudeof idiosyncraticfield studies that have been conducted.

Two critical gaps in knowledge also limitmaturation of the field. First, research needs tomove beyond the easily observable and measurable effects of recreation. In particular, weneed to better understand relationships between the physical, chemical and biologicaleffects of recreation on soil, and how these soilimpacts affect, and are affected by, plants. In theabsence of such knowledge, attempts to restoredamaged sites often fail. Plants are placed insoil that has not held plants for a half-centuryand the plants die (Moritsch and Muir, 1993).Soil amendments are needed before plants cansurvive (Cole and Spildie, 2000; Zabinski et a/.,2002). Restoration has been called the acid testof our ecological knowledge (Jordan ct a/.,1987) because our ability to restore ecosystems

56 D.N. Cole

will be dependent on the depth of our understanding and insight into howecosystemswork.By this definition, our understanding of recreation ecology is still wanting.

The lack of attention that recreation ecolo-

gists have given to the spatial aspects of recreationimpacts isalso problematic. Impacts havealmost always been evaluated at the meso- orsite-scale. Populations and communities ofplants and soil pedons have been the primaryunit of analysis. We have generally done a goodjob of describing impacts that occur at thehuman scale. As mentioned above, lack ofresearch atsmaller scales hampers ourability torestore damaged sites. Lack ofresearch at largerspatial scales - regarding how landscapes andregions are impacted by recreation - limits ourinsight into the significance of recreationimpacts. How do we answer the 'so what' questions? Hiking and camping impacts on soil andvegetation are generally severe but localizeddisturbances. Evaluations of these impacts atlarger spatial scales would result in wiserjudgements about how much of a problemthese impacts are, and the most appropriatebalance between impacts and access for recreation and tourism.

References

Adkison, G.P. and Jackson, M.T. (1996) Changes inground-layer vegetation cover near trails inMidwestern U.S. forests. Natural Areas Journal16, 14-23.

Alessa, L. and Earnhart, C.G. (2000) Effects of soilcompaction on root and root hair morphology;implications for campsite rehabilitation. In:Cole, D.N., McCool, S.F., Borrie, W.T. andO'LoughIin, ]. (comps) Wilderness Science inaTime ofChange Conference. Vol. 5. WildernessEcosystems, Threats and Management Proceedings RMRS-P-15-VOL-5. U.S. Departmentof Agriculture, Forest Service, Rocky MountainResearch Station, Ogden, Utah, pp. 99-104.

Bates, G.Ft. (1935) Vegetation of footpaths, sidewalks, cart-tracks, and gateways, journal ofEcology 23, 470-487.

Bayfield, N.G. (1971) Some effects of walking andskiing on vegetation at Cairngorm. In: Duffey, E.and Watt, A.S. (eds) The Scientific Managementof Plant and Animal Communities for Conservation. Blackwell, Oxford, UK, pp. 469-485.

Bayfield, N.G. (1973) Use and deterioratiran of some

Scottish hill paths. Journal of Applied Ecology10, 639-648.

Bayfield, N.G. (1979) Recovery f)f four montaneheath communities on Cairngorm, Sccjtland,from disturbance by trampling. HiologicalConservation 1 5, 1 65-1 79.

Bayfield, N.G. and Lloyd, K.j. (1973) An a|3proach toassessing the impact of use on a long distancefootpath —the Pennine Way. Recreation NewsSupplement 1973, 11-17.

Benninger-Truax, M., Vankat, j.L. and Shaefer, K.L.(1992)Trail corridors as habitat and conduits formovement of plant species in Rocky McauntainNational Park, Colorado. Landscapi' Ecolrygyh,269-278.

Boucher, D.H., Aviles, ]., Chepote, R., 19ominguez-Gil, O.E. and Vilchez, B. (1991) Recovery oftrailsicle vegetation from trampling in a tropicalrain forest. Environmental Management 15,257-262.

Buckley, R.C. and Pannell, ). (1990) Environmentalimpacts of tourism and recreation in nationalparks and conservation reserves. Journal otTourism StucJies 1, 24-32.

Calais, S.S. and Kirkpatrick, J.B. (1986) Impactof trampling on natural ecosystc>ms in theCradle Mountain-Lake St Clair National Park.Australian Geographer 17, 6-1 5.

Cole, D.N. (1978) Estimating the susceptibility ofwildland vegetation to trailsicle alteration.Journal ofApplied Ecology 15, 281-286.

Cole, D.N. (1981a) Managing ecological impacts atwilcJerness campsites: an evaluation ot tech-nicjues. Journal of Forestry79, 86-89.

Cole, D.N. (1981 b) Vegetational changes assoc iatedwith recreational use and fire sup[iression in theEagle Cap WilcJerness, Oregon: some management implications. Biological Cr^nservation 20,247-270.

Cole, D.N. (1983) Assessing and Monitoring Back-countryTrail Conditions. Research Pafier INT-303.US Department of Agriculture, Forest Service,Intermountain Research Station, Ogden, Utah.

Cole, D.N. (1986) Recreational impacts on back-country campsites in Grand Canyon NationalPark, Arizona, USA. Environmental Management I 0, 651-659.

Cole, D.N. (1987) Research on soil and vegetationin wilcJerness: a state-ol-knowleclge revicwv.In: Lucas, R.C. (comp.) Proceedings NationalWilderness Researc h Conference: Issues, State-

of-knowledge. Future Hirer tions. Ciener.il Icx h-nical Report INT-22(). US (department ofAgriculture, Forest Service, IntermountainResearch Station, Ogdc^n, Utah, |ip. M5-1 77.

Ode, D.N. (1989a) Wilderness Campsite NtonitoringMethods: a Sourcehook. Gener.il Technical


Report INT-25'^). US |])e[iartnient ol Agriculture,Forest Service. Intermountain Research Station,

Ogden, Utah.Cole, D.N. (1989b) Area ol VogctKition Loss: a New

Index of Campsite Impact. Research Note INT-389. US De|)artment ol Agriculture, ForestService, Intermountain Research Station,

Ogden, Utah.Cole, D.N. (1989c) Low-impact Recreational

Practices tor Wilderness and Backcountrw

General Technical Report INT-2b.S. USDepartment erf Agriculture, Forest Service,Intermountain Research Station, Ogden, Utah.

Cole, D.N. (1991) Changes on Trails in the Selway-Bittcrroot Wilderness, Montana, 1978-89.

Research Paper INT-450. US Department otAgriculture, Forest Service, IntermountainResearch Station, Ogden, Utah.

Cole, D.N. (1992) Modeling wilderness campsites;factors that influence amount of impact.Environmental Management 16, 255-264.

Cole, D.N. (1993) Campsites in Three WesternWildernesses: Proliferation and Changes inCondition over 12 to 18 years. Research PaperINT-463. US Department of Agriculture, ForestService, Intermountain Research Station,

Ogden, Utah, USA.Cole, D.N. (1995a) Experimental tram|)ling of vege

tation. I. Relationshi[) between trampling intensity and vegetation response, lournal ofAppliedEcology 32, 203-2 14.

Cole, D.N. (1995b) Experimental trampling of vegetation. II. Predictors of resistance and resilience.

Journal ofApplied Ecology 32, 21 5-224.Cole, D.N. (1995c) Disturbance of natural vegetation

by camping; experimental applications of lowlevel stress. Environmental Management 19,405-416.

Cole, D.N. (1996) Wilderness recreation in theUnited States-trends in use, users, and impacts.International Journal of Wilderness 2, 14-18.

Cole, D.N. (2002) Ecological imf^acts of wildernessrecreation and their management. In: Hendee,J.P. and [i)awson, C.P. (eds) WildernessManagement: Stewardship and Protection ofResources and Values. Fulcrum Publishing,Gfdden, Colorado, USA, [ip. 413-459.

Cole, D.N. and Baytield, N.G. (1993) Recreationaltrampling ot vegetation: standard experimental procedures. Biologiial Cc^nservation 6 1,209-2 1 5.

Cole, D.N. and Hall, I.E. (1992) Trends in CampsiteCondition: Lagle Cap Wilderness, R(^b MarshallWilderness and Crand Canycni National Park.Research Paper INT-451. US Department otAgri( ulture. Forest Servir e, IntermountainResearc h Station, ()gden, Utah.

Cole, D.N. and Monz, C.A. (2002) Trampling disturbance of high-elevation vegetation. Wind RiverMountains, Wyoming, U.S.A. Arctic, Antarctic,and Alpine Research 34, 365-376.

Cole, D.N., Petersen, M.E. and Lucas, R.C. (1987)

Managing Wilderness Recreation Use: Common Problems and Potential Solutions. General

Technical Report INT-259. US Departmentof Agriculture, Forest Ser\'ice, IntermountainResearch Station, Ogden, Utah.

Cole, D.N.and Spildie, D.R. (1998) Hiker, horse, andllama trampling effects on native vegetation inMontana, USA. Journal of EnvironmentalManagement 53, 61-71.

Cole, D.N. and Spildie, D.R. (2000) Soil amendmentsand planting techniques: campsite restoration inthe Eagle Cap Wilderness, Oregon. In: Cole,D.N., McCool, S.E., Borrie,W.T. andO'Loughlin,]. (comps) Wilderness Science in a Time ofChange Conference. Vol. 5. Wilderness Ecosystems, Threats and Management. ProceedingsRMRS-P-15-VOL-5. US Department of Agriculture, Forest Service, Rocky MountainResearch Station, Ogden, Utah, pp. 181-187.

Coleman, R.A. (1977) Simple techniques for monitoring footpath erosion in mountain areas of northwest England. Environmental Conservation 4,145-148.

Coleman, R. (1981) Footpath erosion in the EnglishLakes District. Applied Geography 1, 121-131.

Cordell, H.K. and Super, G.R. (2000) Trends inAmerican's outdoor recreation. In: Gartner,

W.C. and Lime, D.W. (eds) Trends in OutdoorRecreation, Leisure and Tourism. CABInternational, Wallingford, UK, pp. 133-144.

DeLuca, T.H., Patterson, W.A., Freimund, W.A. andCole, D.N. (1998) Influence of llamas, horses,and hikers on soil erosion from established recreation trails in western Montana, USA.

Environmental Management 22, 255-262.Eagen, S., Newman, P., Fritzke, S. and Johnson, L.

(2000) Restoration of multiple-rut trails inTuolumne Meadows ofYosemite National Fbrk.In: Cole, D.N., McCool, S.F., Borrie, W.T. andO'LoughIin, j. (comps) Wilderness Science inTime of Change Conference.Vol. 5. WildernessEcosystems, Threats and Management. Proceedings RMRS-P-15-VOL-5. US Department ofAgriculture, Forest Service, Rocky MountainResearch Station, Ogden, Utah, pp. 188-192.

Ebersole, |.|., Bay, R.F. and Conlin, D. (2002)Restoring high-alpine social trails on theColorado Fourteeners. In: Perrow, M.R. andDavy, A.), (eds) Handbook ot EcologicalRestoration. Vol. 2, Restoration in Practice.

Cambridge University Press, Cambridge, UK,pp. 389-391.

58 D.N. Cole

Farrell, T.A. and Marion, J.L. (2001a) Identifying andassessing ecotourism visitor impacts at eightprotected areas in Costa Rica and Belize.Environmental Conservation 28, 215-225.

Farrell,T.A. and Marion, J.L. (2001 bJTrail impacts andtrail impact management related to visitation atTorres del F^ine National F^rk, Chile. Leisure26, 31-59.

Farrell, T., Hall, T.E. and White, D.D. (2001)Wilderness campers' perception and evaluationofcampsite impacts. Journal ofLeisure Research33, 229-250.

Fish, E.B., Brothers, G.L. and Lewis, R.B. (1981)Erosional impacts of trails in GuadalupeMountains National Park. Landscape Planning8, 387-398.

Frissell, S.S. (1978) judging recreational impacts onwilderness campsites. Journal of Forestry 76,481-483.

Frissell, S.S. and Duncan, D.P. (1965) Campsite preference and deterioration in the Quetico-Superiorcanoecountry. Journal of Forestry 65, 256-260.

Garland, G.G. (1987) Rates of soil loss from mountain footpaths: an experimental study in theDrakensberg Mountains, South Africa. AppliedGeography?, 41-54.

Hall, C.N. and Kuss, F.R. (1989) Vegetation alterationalong trails in Shenandoah National Park,Virginia. Biological Conservation 48, 211-227.

Hall, T.E. and Farrell, T.A. (2001) Fuelwood depletionat wilderness campsites: extent and potentialecological significance. Environmental Conservation 2Q, 241-247.

Hammitt, W.E. and Cole, D.H. (1998) WildlandRecreation: Ecology and Management, 2ndedn. John Wiley &Sons, New York.

Hartley, E.A. (1999) Visitor impact at Logan Pass,Glacier National Ftark: a thirty-year vegetationstudy. In: Harmon, D. (ed.) On the Frontiers ofConservation. George Wright Society, Hancock,Michigan, pp. 297-305.

Hawkins, j.P. and Roberts, C.M. (1993) Effects of recreational scuba diving on coral reefs: tramplingon reef-flat communities. Journal of AppliedEcology30, 25-30.

Jim, C.Y. (1987) Trampling impacts of recreationistson picnic sites in a Hong Kong country park.Environmental Conservation 14, 117-127.

Jordan, W.R., Gilpin, M.E. and Aber, J.D. (eels) (1987)Restoration Ecology. Cambridge UniversityPress, Cambridge, UK.

Jusoff, K. (1989) Physical soil-properties associatedwith recreational use of a forested reserve areain Malaysia. Environmental Conservation 16,339-342.

Kutiel, P. and Zhevelev, Y. (2001) Recreational use

impact on soil and vegetation at picnic sites in

Aleppo pine forests on Mount Carmel, Israel.Israel Journal of Plant Sc iem cs 49, 49-56.

Lance, A.N., Baugh, 1.(9. and Lovc», J.A. (1989)Continued footpath widening in the CairngormMountains, Scf)tland. Biological Conservation49, 201-214.

Lemauviel, S. and Roze, F. (200 3) Kes[oonse of threeplant communities to tramfding in a sand dunesystem in Brittany (France). EnvironmentalManagement 31, 22 7-2 35.

Leung, Y. and Marion, J.L. (1996) Frail degradation asinfluenced hy environmental factors: a state-of-knowlecige review. Journal ot Soil and WaterConservation 51, 1 30-1 3 6.

Leung, Y. and Marif)n, J.L. (1999a) Characterizingbackcountry camping imfiacts in Great SmokyMountains National Park, USA. Journal otEnvironmental Management 57, 193-203.

Leung, Y. and Marion, J.L. (1 999h)Assessing trail conditions in protected areas: a[)plicati(3n ofa problem assessment method in Great SmokyMountains National Park, USA. EnvironmentalConservation 26, 270-270.

Leung, Y. and Marion, J.L. (1999c) The inlluence ofsampling interval on the accuracy of trail impactassessment. Landscape and Urban Planning 43.167-179.

Leung, Y. and Marion, J.L. (1999d) S[:)atial strategiesfor managing visitor im[)acts in national parks.Journal of Park and Recreation Administration17, 20-38.

Leung, Y. and Marion, J.L. (2000) Recreation impactand management in wilderness: a state-ot-knowledge review. In: Cole, D.N., McCool, S.F.,Borrie, W.T. and O'Loughlin, J. (comps) Wilderness Science in a Time ot Change Conference.Vol. 5. Wilderness Ecosystetns, Threats andManagement. Proceedings RMRS-P-1 5-VOL-5.US Department of Agriculture, Forest Service,Rocky Mountain Research Station, Ogden,Utah, pp. 23-48.

Leung, Y. and Neller, R.J. (1995) Trail degradationalong the PatSin Range: an example ot environmental geomorphokjgy. Hong Kong Geologist1, 79-87.

Liddle, M.J. (1975a) A selective review ol the ecological effects of human trampling on natural ecosystems. Biological Conservation 7, 17—36.

Liddle, M.J. (1975b) A theoretical relationshipbetween the primary [jroduc tivity of vc^getationand its ability to tolerate Ir.impling. BiologicalConservation 8, 2 51-255.

Liddle, M.J. (1997) Recreation Ecology. Chafiman <KHall, London, UK.

Liddle, M.J. and Greig-Smith, P.J. (1975) A survey ottracks and paths in a sand dune ecosystem.Journal cjfApplicH Ecology 12, 89 1-9 10.


Liddle, M.J. and Kay, A.W. (I'587i Resistance, survival and recovers- ot Irani[:)led corals on theGreat Barrier RetM. Hiolopicjl Condonation A2,1-18.

Liddle, M.J. and Thyer, N. (1988) Tram[3ling and tirein a subtropical dry sc lerophs-ll forest. En\ir(}n-mental Condor\ati(yn 11.

Marion, j.L. (1991) Do\ eloping a Natural RodourceInventory and Monitoring Prc\yrani for ViditorImpacts on Rccrcatiim Sitod: a ProceduralManual. Natural Resources Report NPS/NRV'T/NRR-91/()8. US [ê[iartnient of Interior,National Park Service, Denver, Colorado.

Marion, J.L. and Cole, U.N. ( I9<^)bi Spatial and temporal variation in soil and vegetation imfâctsoncampsites. Ecological Applicationd b, 520-530.

Marion, J.L. and Farrell, T.A. (2002) Managementpractices that conccMitrate visitor activities:camping im|)act management at Isle RoyaleNational lârk, LISA, journal of En\ ironmental

Management 86, 201 -2 12.Marion, J.L. and Leung, Y. (2001) Trail rc^source

impacts and an examination of alternativeassessment technicjues. Journal of Park andRecreation Administration 19, 17-17.

Meinecke, E.P. (1928) The Effect of E\cesdi\. e Tourist

Travel on the California Redwood Parks. Cali

fornia 13e|)artmeMit of Natural R(>sourcc»s, Division of Parks, Sacramento, California.

Merriam, L.C. and Smith, C.K. (1 974) Visitor impacton newly developed campsites in the BoundaryWaters Canoe Area. l(Hirnal of Eoredtr\ 72,

627-830.

Monti, P. and Mackintosh, E.t. (1979) Effect of camping on surface soil pro|:)erties in the bcrrealfc:)rest region ot northwestern Ontario, Canada.Soil Science Society of America Journal 43,1024-1029.

Mcjnz, C.A. (2002) The response of two arctic tundraplant communities to human trampling disturbance. Journal ot Environmental Management64, 207-217.

Moritsch, B.J. and Muir, P.S. (199?) Subalpine reveg-etation in Yosemite National Park, California:

changes in vegetation alter three vears. NaturalAreas journal 1 ?, I 5 5- H) ?.

Newsome, I)., Moore, S.A. and Dowling, R.K. (2002)Natural Area Touri.'̂ m: Ecolop\. Impacts andManagement. ( hanncd Vi(>w Publications,Clevedon, UK.

Obua, J. and I larding, D.M. (1997) Environmentalim|ia(t of ecotoLirism in Kibale Nation<il Park,Uganda, journal ot Sustainahle Tourism 5,213 22 5.

Pratt, J. (1 997) The ellrx t ot human tramplingon threevegetation ( ommunities in Mount ElgonNational Park. Available at website hlt[i:/'

www.env.leeds.ac.uk/elgon/trampling.html(verified 30 January 2003).

Rouphael, A.B. and Inglis, G.J. (2002) Increasedspatial and temporal variabilitc' in coral damagecaused by recreational scuba diving. EcologicalApplications 12, 427-440.

Scott, J.J. and Kirkpatrick, J.B. (1994) Effects ofhuman trampling on the sub-Antarctic vegetation of Macquarie Island. Polar Record 30,207-220.

Spildie, D.R„ Cole, D.N. and Walker, S.C. (2000)Effectiveness of a confinementstrategy in reducing pack stock impacts at campsites in theSelway-Bitferroot Wilderness, Idaho. In: Cole,D.N., McCool, 5.P., Borrie,W.T. and O'Loughlin,J. (comps) Wilderness Science in a Time ofChange Conference. Vol. 5. Wilderness Eco-svstems, Threats and Management. ProceedingsRMRS-P-15-VOL-5. US Department of Agriculture, Forest Ser\'ice, Rocky MountainResearch Station, Ogden, Utah, pp. 199-208.

Stankey, G.H., Cole, D.N., Lucas, R.C., Petersen,M.E. and Frissell, S.S. (1985) The Limits ofAcceptable Change iLACi System forWilderness Planning. Research Paper INT-176.US Department of Agriculture, Forest Service,Ogden, Utah.

Stewart, D.P.C. and Cameron, K.C. (1992) Effect oftrampling on the soils of the St. James Walkway,New Zealand. Soil Use and Management 8,30-36.

Sun, D. and Liddle, M.J. (1993a) A survey of tramplingeffects on vegetation and soil ineight tropical and subtropical sites. EnvironmentalManagement 17, 497-510.

Sun, D. ancJ Liddle, M.J. (1993b) Plant morphologicalcharacteristics and resistance to simulated tram

pling. Environmental Management17, 511-521.Tachibana, H. (1969) Vegetation changes of a moor

in Mt. Hakkoda caused by human treading.Ecological Review 17, 177-188.

Taylor, J.Y. (1997) Leave only footprints? How back-country campsite use affects forest structure.Yellowstone Science 5, 14-17.

Wagar, J.A. (1961) How to predict which vegetatedareas will stand up best under 'active' recreation. American Recreation journal 1, 20-21.

Wagar, |.A. (1964)Thecarryingcapacity ofwild landsfor recreation. Forest Science Monograph 7.Society of American Foresters, Washington, DC.

Whinam, j. and Chilcott, N. (1999) Impacts of trampling on alpine environments in centralTasmania, journal of Environmental Management 57, 205-220.

Whinam, I., Cannell, E.J., Kirkpatrick, j.B. and(Omfort, M. (1994) Studies on the potentialim[")act of recreational horseriding on some

60 D.N. Cole

alpine environments of the Central Plateau,Tasmania. Journal of Environmental Management 40. 103-117.

Wilson, J.P. and Seney, ).P. (1994) Erosional impact ofhikers, horses, motorcycles, and off-road bicycles on mountain trails in Montana. Mountain

Research and Development 14, 77-88.Yoda, A. and Watanabe, T. (2000) Erosion of moun

tain hiking trail over a seven-year period inDaisetsuzan National Park, central Hokkaido,Japan, in; Cole, D.N., McCool, S.F., Borrie, W.T.and O'Loughlin, j. (comps) Wilderness Sciencein a Time of Change Conference. Vol. 5.

Wilderness Ecosystems, Threats and Management. Proceedings RMKS-P-1 3-VC)L-5. USDepartment of Agriculture, Forest Service,Rocky Mountain Research Station, Ogden,Utah, pp. 172-178.

Zabinski, C.A. and Cannon, j.F. (1 907) F.llec ts ot recreational impacts on soil microbial communities.Environmental Management 21,2 13-238.

Zabinski, C.A., DeLuca, T.F1., Cole, D.N. andMoynahan, O.S. (2002) Restoration ot highlyimpacted campsites in the EagleCap Wilderness,Oregon. Restoration Ecology 10, 27.S-281.

impacts of hiking and camping on soils and vegetation: a...

Documents