Experimental evaluations of two leave-no-trace techniques : removing boots and using geotextile groundcloths (scrim)

14 Pages·1997·1.1 MB·English

by  Cole, David N

Si 7
I
I
document
Historic, archived
Do not assume content reflects current
scientific l<nowledge, policies, or practices.

3d States Experimental Evaluations of
LJttpartment
of Agriculture
Two
Leave-No-Trace Techniques:
Forest Service
Intermountain Removing Boots and Using
Research Station
Research Paper Geotextile Groundcloths (Scrim)
INT-RP-497
November 1997
'3
David N. Cole
—
The Author causedwhenthegroundclothwasnotinplace. However,
1 year after trampling, magnitude of cover loss did not
David N. Cole is a Research Biologist with the Aldo differ. Footwear had no effect on vegetation height.
Leopold Wilderness Research Institute, Missoula, MT. These results suggest that these two practices have
He received his B.A. degree in geography from the smallshort-termbenefitsbutnolong-termbenefits.There-
University of California, Berkeley, in 1972. He received fore, they are unlikely to contribute to a meaningful
his Ph.D. degree, also in geography, fromthe University reduction in resource impact. However, they are not
of Oregon in 1977. He has written many papers on harmful.
wilderness management, particularly on the ecological
effects of recreation use.
Acknowledgments
Research Summary am grateful to the many people who assisted with
I
thesestudies.ThankstoDaveSpildieforfieldassistance
Experiments were designed to evaluate the effective- andto Ralph Swain and Chris Monzforcomments on an
ness oftwo recommended Leave-No-Trace practices early draft of this manuscript.
removing boots and using a geotextile groundcloth, re-
ferredtocolloquiallyas"scrim,"incamp.Oneexperiment
compared trampling impacts on vegetation infourdiffer- Contents
ent vegetation types when hikers wore lug-soled boots Page
and when they wore lightweight running shoes. About Introduction 1
6 percent morevegetation coverlossoccurredwhen the Footwear Experiment 1
tramplers wore lug-soled boots than when they wore Methods 1
running shoes. However, 1 yearaftertrampling, the mag- Results 2
nitudeofcoverlossdidnotdiffer. Moreover,footwearhad Discussion 2
no effect on vegetation height. Geotextile Groundcloth Experiment 3
In the second experiment, conducted on two different Methods 3
vegetation types, a geotextile groundcloth was placed Results 4
over the ground cover vegetation before trampling took Discussion 6
place. When the groundcloth was in place, trampling Conclusions 6
caused only about one-half the vegetation cover loss it References 7
You may order additional copies of this publication by sending your
mailing information in label form through one of the following media.
Please specify the publication title and Research Paper number.
Telephone [phone]
DG message Pubs:S22A
FAX [phone], Attn: Publications
E-mail /s=pubs/[email protected]
—
IVIailing Address Publications Ogden Service Center
Rocky Mountain Research Station
(formerly Intermountain Research Station)
324 25th Street
Ogden, UT 84401
Rocky Mountain Research Station
(formerly Intermountain Research Station)
324 25th Street
Ogden, UT 84401
—
Experimental Evaluations of
Two
Leave-No-Trace Techniques:
Removing Boots and Using
Geotextile Groundcloths (Scrim)
David N. Cole
Introduction types, named for the most abundant ground cover
species,were:(1)Sitkavalerian{Valerianasitchensis),
Recreationaluseofwilderness altersnaturalcondi- lush subalpine herb meadows in the Cascade Moun-
tions intended for preservation. The magnitude of tains, WA; (2) grouse whortleberry (Vaccinium
recreationimpactishighlyinfluencedbythebehavior scoparium), subalpine coniferousforests inthe Rocky
of recreationists. Consequently, considerable effort Mountains,CO;(3)Canadamayflower{Maianthemum
has gone into development of the "Leave-No-Trace" canadensis), low-elevation hardwood forests with a
educational program. Leave-No-Trace is designed to herbaceousgroundcoverintheWhiteMountains,NH;
promote responsible use ofwildlands by visitors par- and (4) hog-peanut (Amphircarpa bracteata), low-
ticipating in nonmotorized recreational activities elevation cove hardwood forests with a herbaceous
(Swain 1996). The program unites four Federal land- ground cover in the Great Smoky Mountains, NC.
—
managingagencies theForestService,NationalPark
Service, Bureau ofLand Management, and Fish and Methods
—
Wildlife Service with nongovernmental organiza-
tions, manufacturers, user groups, and individuals. Details ofthe exerimental design follow the proto-
The program attempts to build awareness, apprecia- colsofCole and Bayfield (1993). Fourreplicate sets of
tion, and respectforthe wildnessin places. Itteaches experimental trampling lanes were established on
visitors how to minimize their impact on wildlands levelsitesineachofthefourvegemtationtypes.Emachset
by recommending specific low-impact techniques consistedofninelanes,each0.5 wideand 1.5 long.
(HamptonandCole 1995). Onetechniquethatisoften Treatments were randomly assigned to lanes. One
suggestedis toremovebootswhenyou arrive atcamp lanewasacontrolandreceivednotrampling.Theother
and put on soft-soled footwear. The rationale for this lanes received either 25, 75, 200, or 500 passes, either
suggestionisthatlug-soledbootshavemorepotential byahikerinlug-soledbootsorahikerinrunningshoes.
to cause erosion (Harlow 1977) or to impact plants. A pass was a one-waywalk down the lane.
Another recent trend, particularly with packstock Measurements were taken on each lane in two
groups, is to lay down geotextile groundcloths (com- adjacent 30 by 50 cm subplots. Visual estimates of
monlyreferredtoasscrim)incamp(Stonerandothers coverwererecorded, bycoverclass, foreachspeciesof
1993). The rationale for their use is that they shield vascularplant and for mosses and lichens. Total veg-
vegetation and soil from the direct abrasive effects of etation cover was the sum ofthe coverages for indi-
trampling,andthereforereducedamage.Experiments vidual species, mosses, and lichens. Mean vegetation
were undertaken to evaluate these two techniques heightwas recorded with a point quadrat frame with
removing boots and laying down scrim—as effective five pins. The frame was placed 10 times in each
means ofreducing trampling impacts on vegetation. subplot. Pins were dropped to the ground. When pins
hit live vegetation, the height of the pin strike was
recorded.Meanheightwasthemeanofthepinstrikes.
Footwear Experiment
Measurements were taken immediately prior to
The investigation offootwear effects was part ofa trampling, 2 weeks after trampling, and 1 year after
trampling. The primary measures of vegetation re-
largerinvestigationoftramplingimpactsinfourmoun-
sponse to trampling were relative vegetation cover
tainous regions ofthe country (Cole 1993). It was con-
ductedin onevegetationtypein eachoffourmountain- and relative vegetation height, 2 weeks and 1 year
after trampling. Relative vegetation cover was calcu-
ous regions in the United States. The four vegetation
lated as:
1
.
Survivingcoverontrampledsubplots interactionbetweenvegetationtypeandfootwearwas
X cf X 100 percent
Initial coverontrampled subplots significant, however. Lug-soled boots caused more
where: impact than running shoes in two vegetation types
cf= Initial coveron control subplots andlessimpactintheothertwotypes (fig. 2). None of
Survivingcoveron control subplots theseindividual differenceswere statistically signifi-
Relative height was calculated in a similar manner, cant. Overall, mean relative cover 1 year after tram-
substitutingheight for cover. Both relative cover and pling was 68 percent when the trampler wore lug-
soled boots and 67 percent when the trampler wore
height would be 100 percent in the absence of any
trampling effect. Lower relative cover and height running shoes.
measures indicate greater trampling disturbance. Neither relative vegetation height after trampling
The significance of footwear effects was assessed norheight 1 year aftertramplingvaried significantly
within a three-way analysis of variance, with foot- with type of footwear (p > 0.1). Two weeks after
wear, trampling intensity, and vegetation type as trampling, the effect offootwear interacted with the
main factors. This permitted identification ofthe ef- effectofvegetationtype. Lug-soledbootscausedmore
impact than running shoes in three vegetation types
fect of footwear, while statistically controlling the
effects of trampling intensity and vegetation type. and less impact in the fourth type. None of these
Where interactions between footwear and the other differences were statistically significant, however.
main factors were significant, simple effects were Mean relative height after trampling was 29 percent
examined in more detail. when the trampler wore lug-soled boots and 34 per-
cent when the trampler wore running shoes. Mean
Results relativeheight 1 year aftertramplingwas 69 percent
when the trampler wore lug-soled boots and 72 per-
Relative cover after trampling varied significantly cent when the trampler wore running shoes.
withallthreemainfactors(table 1). Footwearwasthe
factor with the least effect on relative cover. None of Discussion
the interactions of other factors with footwear were
This experiment shows that the type of footwear
significant. Althoughthetype offootwearwornhad a
statistically significanteffect, the magnitude ofeffect worn can affect magnitude ofvegetation disturbance
wtraasmnpoltinpgrwoanosu3n1cpeedr(cfeign.t1w).heMneathnerterlaatmipvleecrowvoerrealfutge-r blausttitnhga.tWdhifefneretnhceestarraempnleeirthwerorseubbsotoatnst,iaslhonrotr-tleornmg
soled boots and 37 percent when the trampler wore vegetationcoverlosswasgreaterthanwhenthetram-
plerworerunningshoes.However,themeandifference
running shoes.
Oneyearaftertrampling,thetypeoffootwearworn in coverloss was only6 percent, and differences were
hadnoeffectonrelativevegetationcover(table1).The nonexistent 1 year after trampling. Footwear had no
—
Table 1 Analysisofvariancefortheeffectoffootwear(bootorrunningshoe), trampling intensity, andvegetationtype
on relative vegetation cover2 weeks and 1 year aftertrampling.
Two weeks aftertrampling One yearaftertrampling
Source of variation df Mean square F Mean square F
Model 34 2,643 18.10^ 2,787 8.50^
Error 93 146 328
Main effects
Footwear 1 1,249 8.55^ 3 0.01
Trampling intensity 3 18,988 130.05^ 11,033 33.64^
Vegetation type 3 7,186 49.22^ 8,995 27.42^
Blocks 3 1,050 7.19^ 5,013 15.28^
Interactions
Footwearx intensity 3 138 0.95 237 0.72
Footwearx vegetation 3 333 2.28 929 2.83"
Intensity x vegetation 9 532 3.64^ 1,408 4.29^
Footwearx intensity x vegetation 9 81 0.55 384 1.17
^Significance: <0.01
''Significance: <0.05.
2
100 -, Typeoffootwearappearstohave evenless effecton
the magnitude ofsoil impacts. Two studies assessed
the effect offootwear on variables related to erosion
(percenl 80 - Lug-soled boot potential, sedimentyield, and amountofsoil adhered
I I
^^^^^s^ Running shoe to soles (Kuss 1983; Saunders and others 1980). In
- both studies, no significant differences between boots
Cover 60
with lug soles and boots without lugsoles were found.
co Kuss(1983)conductedhis studies atseveral soil mois-
40 — ture levels and in several soil types, suggesting his
Vegetal findings may be representative of a range of trail
conditions. Saundersandothers(1980)reportedthata
> 20 - lug-solebootcollectedmoresoilduringtramplingthan
a tennis shoe, but since they did not report their data,
(U the magnitude ofdifference cannot be determined.
a:
n Although changing footwear is unlikely to contrib-
— ute to meaningful reductions in resource impacts,
Figure 1 Mean relative vegetation cover
there appear to be no negative consequences ofthis
2 weeks after trampling by hikers in lug- behavior. Manycampers change tolighter shoes sim-
soledbootsandinlightweightrunningshoes.
plybecause they are more comfortable. There maybe
Bars indicate 1 standard error.
benefitsfromthispracticethatcouldnotbeevaluated
in our experiments. For example, an effective way to
minimizetramplingdamagetocampsitevegetationis
Lug-soled boot to avoid stepping on plants. Campers in light shoes
_100 ^ I|>C\NI Running shoe maybemorelikelytowatchwheretheyplacetheirfeet
c thancampersinheavyboots. Ifso,theywillcauseless
8 T vegetation damage.
80
2.
Geotextile Groundcloth
> 1
o
O 60 Experiment
c X
o
The investigation of the effects of a geotextile
2
40 groundcloth(commonlyreferredto asscrim)waspart
of a larger investigation of trampling impacts by
a> 20 - hikers, llamas, and horses (Cole and Spildie, submit-
.> ted). It was conducted in two montane coniferous
m
forestvegetationtypesinwesternMontana. Onetype
0)
a: (denotedtheEquisetumtype)hadanunderstorydomi-
WA CO NH NC nated by forbs, particularly horsetail {Equisetum
arvense) and queencup beadlily (Clintonia uniflora).
—
Figure 2 Mean relative vegetation cover The other type (denoted the Vaccinium type) had an
1 yearaftertramplingbyhikersinlug-soled understory dominated by the low shrub grouse
boots and lightweight running shoes in whortleberry (Vaccinium scoparium).
Washington (WA), Colorado (CO), New
Hampshire(NH), and North Carolina(NC). Methods
Bars indicate 1 standard error.
Fourreplicatesetsofexperimentaltramplinglanes
were established in each ofthe two vegetation types.
m m
Each set consisted offive lanes 0.5 wide and 3.0
effect on vegetation height. In a study conducted in
Great Smoky Mountains National Park, Whittaker long. Treatments were randomly assigned to lanes.
Onelanewasacontrollaneandreceivednotrampling.
(1978) also reported no difference in magnitude of
The other lanes received either 25 or 150 passes by a
vegetation height reduction, depending on whether
hikerinboots,withorwithoutageotextilegroundcloth
thetramplerworelug-soledbootsorlightweightshoes
coveringthevegetation Thegroundclothwasremoved
with soft soles. immediatelyaftertramp.lingceased.Apasswasaone-
way walk down the lane.
3
Measurements andthecalculationofresponsevari- thegroundcloth, aswell astramplingintensity. Over-
ables were identical to those for the footwear experi- all, relative cover ofvascular plants was 66 percent
ment, with three exceptions. First, subplots were lo- after trampling without the groundcloth and 85 per-
m
cated 1 apart, instead of adjacent to each other. cent after trampling with the groundcloth in place.
Second,vegetationcoverwascalculatedasthepropor- The interaction between groundcloth and vegetation
tion ofthe 100 pins (50 in each subplot) that hit live type was moderately strong {p = 0.07). The positive
vegetation. Field observations suggested that the re- effectsofthegroundclothweremorepronouncedinthe
sponses ofvascular plants might differ from those of Vaccinium vegetation type (fig. 3).
mosses and lichens. Relative cover ofvascular plants Thelow shrubunderstoryofthe Vaccinium vegeta-
was calculated in addition to relative cover of all tiontypeisgenerallymoreresistanttotramplingthan
vegetation. theerectforbsoftheEquisetumtype(ColeandSpildie,
The significance of the groundcloth's effects was submitted). Most cover loss in the Vaccinium type
assessedwithinathree-wayanalysisofvariance,with resulted from breakage of the woody stems and
groundcloth (presence or absence), trampling inten- branches of Vaccinium scoparium. Flattening of
sity,andvegetationtypeasmainfactors.Thispermit- Vaccinium plants was insufficient to kill them. Field
tedidentificationoftheeffectofthegroundcloth,while observations suggested that the groundcloth was ef-
statisticallycontrollingtheeffectsoftramplinginten- fective in reducing breakage of woody stems and
sity and vegetation type. Where interactions between branches. However, the trampling experiments were
thegroundclothandtheothermainfactorsweresignifi- conductedinmidsummerbefore thewoodystems and
cant, simple effects were examined in more detail. branches of Vaccinium were highly brittle. Later in
the season, whenwoodyparts are brittle, the effect of
Results the groundcloth maybe different. The fragile forbs of
the Equisetum type were usually killed ifthey were
The three main factors (groundcloth, tramplingin- flattened;consequently,thegroundclothofferedthem
tensity, and vegetation type) and their interactions relativelylittle protection.
didnotexplainmuchofthevariationinrelativecover One year after trampling, the model (groundcloth,
ofall vegetation after trampling (F - 1.38,p = 0.26). trampling intensity, and vegetation type and their
Nevertheless,theeffectofgroundclothwassignificant interactions) explainedmuchmoreofthevariation in
{F = 5.64, p = 0.03). Without the groundcloth, mean vascularplantcover(F= 10.00,/?<O.OODthanintotal
relative cover after trampling was 73 percent. With vegetation cover {F= 2.24,p = 0.07). For both depen-
the groundcloth in place, mean relative cover after dent variables, trampling intensity and vegetation
trampling was 85 percent. typehadthemostpronouncedeffect,whilepresenceor
The model was much more effective in explaining absence of the groundcloth did not explain much
variationintherelativecoverofvascularplants(table2). variation in vegetation response (table 2). Overall,
Relative cover ofvascular plants 2 weeks after tram- mean relative cover of vascular plants 1 year after
pling varied significantly with presence or absence of trampling was 78 percent without the groundcloth
—
Table2 Analysisofvariancefortheeffectofageotextilegroundcloth "scrim,"trampling intensity, andvegetationtype
on relative cover of vascular plants 2 weeks and 1 year aftertrampling.
Two weeks aftertrampling One yearaftertrampling
Source of variation df Mean square F IVIean square F
Model 7 782 4.68^ 2,521 10.00^
Error 24 167 764
Main effects
Groundcloth 1 2,810 16.80^ 20 0.08
Trampling intensity 1 1,810 10.83^ 5,022 20.00^
Vegetation type 1 32 0.19 5,863 23.25^
Interactions
Groundcloth x intensity 1 144 0.86 64 0.25
Groundcloth x vegetation 1 626 3.75 1,014 4.02
Intensity x vegetation 1 49 0.29 5,629 22.32^
Groundcloth x intensity x vegetation 1 5 0.03 35 0.14
^Significance: <0.01.
4
Scrim
Scrim
100 - I I
cc- K^^^^ No scrim
k 80
>
o -
O 60
c
o
-
•2 40
D)
>0)
20
I
(U
Vaccinium Equisetum
Vaccinium Equisetum
—
Figure 3 Mean relative cover of vascular —
plants in two vegetation types—grouse Figure 4 Mean relative cover of—vascular
whortleberry (Vaccinium) and horsetail plants in two vegetation types grouse
— whortleberry (Vaccinium) and horsetail
(Equisetum) 2 weeks after trampling by —
hikers, with and without a geotextile (Equisetum) 1 yearaftertramplingbyhikers,
with and without a geotextile groundcloth
groundcloth (scrim) to shield the vegetation.
(scrim)toshieldthevegetation.Barsindicate
Bars indicatel standard error.
1 standard error.
and 77 percent with the groundcloth in place. The (fig. 4). Neither ofthese differences was statistically
interaction between groundcloth and vegetation type significant,suggestingthattheeffectofthegroundcloth
was moderately strong (p = 0.06). In the Vaccinium onvegetationcoverisnegligible 1yearaftertrampling.
type, more vascular cover survived when the Neither vegetation height 2 weeks after trampling
groundcloth was in place; in theEquisetum type, less norheight 1 year aftertramplingvaried significantly
cover survived when the groundcloth was in place with presence or absence ofthe groundcloth (table 3).
—
Table 3 Analysis of variance for the effect of a geotextile groundcloth "scrim," trampling intensity, and vegetation
type on relative vegetation height, 2 weeks and 1 yearaftertrampling.
Two weeks aftertrampling One yearaftertrampling
Source of variation df Mean square F Mean square F
Model 7 5,182 13.74^ 2,588 3.81^
Error 24 377 679
Main effects
Groundcloth 1 168 0.45 106 0.16
Trampling intensity 1 7,503 19.90^ 3,051 4.49*^
Vegetation type 1 25,153 66.70^ 3,948 5.81^
Interactions
Groundcloth x intensity 1 1,490 3.95 3,661 5.39"
Groundcloth x vegetation 1 1,581 4.19" 95 0.14
Intensity x vegetation 1 77 0.20 7,171 10.56^
Groundcloth x intensity x vegetation 1 298 0.79 85 0.12
^Significance: <0.01.
''Significance; <0.05.
5
results suggest that the groundcloth may be more
effective in avoiding height reduction in vegetation
that is not readily flattened and at relatively high
trampling intensities. At low trampling intensities,
particularlyinvegetationthatisreadilyflattened,the
groundcloth can accentuate the flattening effect of
trampling. Plants that are not stepped on can be
flattened by the weight of the groundcloth as it is
trampled.
Discussion
This experiment shows that the use ofa geotextile
groundcloth in camp can reduce the magnitude of
vegetation disturbance, but that differences are not
long lasting. When the groundcloth was in place,
Vaccinium Equisetum
short-term vegetation cover loss and vascular plant
Figure 5—Mean relative vegetation height coverloss were 12 percent and 19 percent lowerthan
in two vegetation types 2 weeks after when the groundcloth was not in place. However,
trampling by hikers, with and without a differences were nonexistent 1 year after trampling.
geotextile groundcloth (scrim) to shield the Use of the groundcloth had no consistent effect on
vegetation. Bars indicate 1 standard error. vegetation height.
This suggests thatuseofageotextilegroundclothis
unlikely to reduce impact to any meaningful degree.
However, it is not likely to aggravate impact either.
Twoweeksaftertrampling,meanrelativeheightwas Anecdotal evidencegatheredfrom outfitters suggests
slightly lower on the groundcloth plots (60 percent) that the primary benefits of using a groundcloth in
than on the plots that were trampled without the camp are to avoid dust and mud problems and to
groundcloth (65 percent). One year after trampling, facilitate camp cleanup.
meanrelativeheightwasslightlyhigheronthelanes
that had been covered with the groundcloth (79 per-
Conclusions
cent)thanonthelanesthathadnotbeencovered (75
percent).Again,thesedift"erencesarenotstatistically
Neither changing footwear nor laying dovm a geo-
significant.
textilegroundcloth(commonlyreferredtoas scrim)is
Interactions complicate the picture slightly. For
vegetationheight 2weeks aftertrampling, theinter- likely to contribute to a meaningful reduction in re-
source impact. Both recommended behaviors have
actions between groundcloth and both trampling in-
small short-term positive effects but no long-term
tensity (p = 0.06) and vegetation type (p = 0.05) are
moderately strong. Mean relative height aftertram- benefits. Conversely, neither behavior appears likely
pling was higher when the groundcloth was in place to aggravate impact problems. Therefore, they are
behaviors that can be recommended.
intheVacciniumtypeandlowerwhenthegroundcloth
Theprimarymanagementimplicationofthesefind-
was in place in the Equisetum type (fig. 5). Mean
relative height aftertrampling was higherwhen the ingsconcernstheprominencetheserecommendations
should be given in "Leave-No-Trace" messages. Wil-
groundcloth was in place at the higher trampling
intensity but lower when the ground cover was in derness visitors, like all humans, have limits to the
attention and cognitive capacitythey are willing and
place at the lower trampling intensity. This latter
responsepatternwasrepeatedforrelativevegetation able to devote to information. For example, a recent
study found that typical wilderness visitors will only
height 1 year after trampling.
allocate enough attention to low-impact messages
The findings that (1) none ofthese differences are
posted on bulletin boards to comprehend two of the
statistically significant and (2) the direction of re-
behaviors recommended in the messages (Cole and
sponse differs between vegetation types and tram-
others 1997).
pling intensities support a general conclusion that
Because these two recommended behaviors are not
the use ofthe groundcloth has no consistent or sub-
likelytohavesubstantialbenefits, theyshouldnotbe
stantial effect on resultant vegetation height. The
high priorities in educational campaigns. The use of
6