Predicting the durability of forest recreation sites in northern Utah : preliminary results

Historic, archived document

Do not assume content reflects current
scientific knowledge, policies, or practices.

UNITED STATES DEPARTMENT OF AGRICULTURE
iy FOREST SERVICE

r¢ INTERMOUNTAIN FOREST & RANGE EXPERIMENT STATION

DEN, UTAH 84401

- USDA Forest Service U. 8. DEFT. oF

Research Note INT-117 MUTIONAL Asrounme ne vi
a IBRARY

S
PREDICTING THE DURABILITY OF FOREST FFP A 1920 res

IN NORTHERN UTAH--PRELIMI NHR AEN RESTS

Tes, L/
= Thomas J. Cieslinski and J. Alan Wagar
< ABSTRACT
r By using a special roller, trampling was simulated tn equal

amounts on 40 small plots representing potential recreation sites

throughout the Cache National Forest in northern Utah. Surviving
‘ vegetation was related to sotl and topographte factors by multtple
regression procedures. Resultant equations explained up to 64
percent of the variability in amounts of surviving vegetation,
which suggests the possibility of predicting the durabiltty of
potential recreation sites. Site factors that can be measured on
> aerial photos explained approximately as much vartability as

factors requiring on-the-ground measurements.

One of the more serious: problems facing recreation site managers today, particularly
on campground and picnic areas,: is that of maintaining adequate ground cover. This is
cee an especially difficult task in semiarid regions such as the Intermountain West, where

sites tend to support a sparse cover of ground-level vegetation. Because of different
: soil, moisture, and topographic conditions, some sites are much more durable than others
in terms of the persistence of ground-cover vegetation. If we had tools for rating the
durability of potential recreation sites, less durable sites either could be avoided or
: designed and managed in ways to increase their durability. Moreover, knowledge of the
probable level of vegetation damage to a fragile site might allow managers to estimate
any additional expenditures that may be needed if such a site is developed and used as
4 a campground or picnic area.

IThis paper is based on the senior author's master's thesis accepted at Utah State
F University in 1968. The senior author is now Supervisor of Planning and Research
for the Maine State Park and Recreation Commission. At the time this work was done,
the junior ‘author was leader of the Cooperative Recreation Research Unit maintained by
. the USDA Forest Service in cooperation with Utah State University at Logan. Hewas
currently leader of a similar unit maintained at the University of Washington in Seattle
by the Pacific Northwest Forest and Range Experiment Station, USDA Forest Service.

oe nag 4 Pat

‘ie a HAY. ict 5p

ue ea”
ee

ik 1S ue i SS Oe 7 as | 7 & Ai i 7 hil f, ai AGRE 3

va i

sha

hs arta

Figure 1.--Configurated
roller used in the
study to simulate

trampling from
recreation use.

To provide procedures for rating the durability of site vegetation, a study was
established on the Cache National Forest in northern Utah in 1965. A 100-1b. configu-
rated cement roller (fig. 1) that applied a rolling pressure of about 6 lbs./in.® was
used to simulate trampling on the 16- by 64-inch plots. Identical amounts of simulated
trampling were applied to each plot in the study. Plots then could be arranged in order,
according to resistance to damaging forces.

Whether the roller damage was greater or smaller than that resulting from human
trampling is not known but probably is unimportant. We assumed that after moderate
trampling the amount of vegetation surviving through one or more growing seasons would
express site durability. Then, if amounts of surviving vegetation could be related to
soil and topographic factors, the resulting relationships would enable us to predict
site durabitity.

Because simulated trampling had been used in only one other study (Wagar 1964a),
the 1965 growing season was used to test the effect of different intensities and timings
of simulated use. On 48 plots--24 under a cover of lodgepole pine (Pinus contorta
Dougl.) and 24 under a cover of aspen (Populus tremulotdes Michx.)--we learned that plot
vegetation responded quite differently to varying amounts of rolling, but little differ-
ently to the same amount of rolling applied with different timings. The response was
about the same whether plots received two passes of the roller on 3 days of each week
or the same amount of trampling applied as 12 passes of the roller once every other
week. This seemed to justify the once-a-week rolling treatment that was planned for
the second phase of the study.

Based on results from 1965, 40 sites were selected in May 1966. Sites represented
the range of conditions judged suitable for recreation occupancy sites on the Cache
National Forest. The sites chosen were concentrated on slopes ranging from 0 to.15
percent. To help us define the form of relationships, we included a few sites on steep-
er slopes (up to 30 percent). Sites represented overstory types ranging from lodgepole
pine to aspen to maple to willow. The 40 plots selected included the following ranges
of conditions: all aspects; elevations of from 5,100 to 8,400 ft.; soil pH values of

Rae! CEA tk ny HORT
beter .
| alga,
oe

Figure 2.--Arrangement of
16- by 64-inch plots at
each stte location.

Treatment

Control

Treatment

5.0 to 7.4; plot positions on the slope ranging from the bottom of the drainage to the

top of the slope; a variety of soil textures in both the A and B soil horizons; percent
stones in the soil (2 mm. and larger) from 6.9 to 64.8; season-long percentages of pos-
sible direct sunlight ranging from 1 to 40; and basal area of trees ranging from 26 to

313 ft.* per acre. Measurements were recorded for each of these site characteristics.

The most northerly plots and the most southerly plots were 52 airline miles apart.

The grass and forb species at the sites included the following: Polygonwn
durslasta; Aconttum colwnbtanun; Bromus marginata; Melica bulbosa; Aster engelmanntt;
Achillea lanulosa; Agropyron spicatum; Wyethia anplexicaulis; Thalietrum fendlert;
Senecto serra; Lathyrus leucanthus; Agastacha urtictfolita; Rudbeckta occidentalis;
Delphiniun barbeyt; Osmorhiza ehilensis; Stdalacea neomexicana; Luptnus laxtflorus;
Verontea ecampylopoda; Agoserts glauca; Arnica cordtfolia; Taraxacum officinale;
Hydrophyllun capttatun var. thompsontt; and several Carex species.

At each site location, four 16- by 64-inch plots were placed in stands of ground-
cover vegetation judged to be uniform. As shown in figure 2, two treatment plots were
alternated with two control plots in each plot group. Beginning the latter half of
June 1966, each treatment plot was rolled once a week with 12 passes of the roller.
Treatments continued for 11 weeks. A l-week interval between treatments provided
sufficient time for all plot locations to be treated in a.repetitive sequence and also
allowed time for measurement of site variables at each location.

2Aspect was coded as 1.0 plus the sine of the azimuth from southeast. This gave
values ranging from 0.0 on cool northeastern exposures to 1.0 on moderate southeastern
and northwestern exposures to 2.0 on hot southwestern exposures. Season-long percentage
of possible direct sunlight was measured by using an insolation grid (Wagar 1964b). The
point density procedure developed by Spurr (1962) was used to measure the basal area
Of streesi.

‘ be et A
ESS APOE eee,
aa xd RRC, Rae a Pe! aT

ee a: aN

During the growing season, plots at three locations were damaged so severely by
livestock grazing that they could not be used in the analysis. Consequently, analysis
was based on plots at only 37 locations.

At the end of the growing season (between September 10 and October 25) vegetation
within a 12- by 60-inch zone in each plot was measured in two different ways. First,
a 12- by 30-inch grid with two hundred 1.2- by 1.5-inch rectangles was placed over each
half of the measurement zone and a count was taken of the number of stocked rectangles,
i.e., rectangles having living vegetation anchored in them. This number then was ex-
pressed as a percentage of the total number of 400 rectangles and used as dependent
variables Y; and Y2. As a second measurement procedure, all plants from the measurement
zone of each plot were clipped one-half inch above their root collars. Clippings then
were ovendried and weighed for a measure of dependent variable Y3.

Two other variables also were constructed and used to express results. Vegetation
surviving on treated plots was expressed as a percentage of the vegetation surviving on
adjacent control plots. This percentage was used as dependent variable Y, when based
on clipping weights, and as dependent variable Ys when based on stocking measurements.

Multiple regression analysis procedures were used. Results for the five regression
models tested are summarized in table 1.

In the equations for Y,; and Y,, approximately 60 percent of the variability was
explained. It should be noted that the equation for Y,; is limited to variables that
either were or could have been measured from aerial photographs. From this we conclude
that the possibility of determining site durability from aerial photos is Dron aie
However, because this equation was based on measurements from only 29 plot groups,
results for this equation are not quite comparable to others.

In the equation for Y,;, most of the variability appeared to be due to topographic
factors, and studies are being continued to determine whether site durability can be
satisfactorily predicted from aerial photograph measurements.

Independent variables were examined for consistency from equation to equation and
for statistical significance (defined as 5 or less percent probability that association
with the Y variable was a chance occurrence). Variables that were both significant
and consistent included slope percent, aspect, elevation, and the interaction between
slope percent and aspect. Three additional variables, although not statistically sig-
nificant, had no inconsistencies from one equation to another and occurred in the equa-
tion that had the best predictive value (lowest mean square error) for one or more of
the models tested. These were percent clay at a soil depth of 1 to 4 inches, basal
area of trees, and percent stones (>2 mm.in diameter) at a soil depth of 1 to 4 inches.
Four variables were either inconsistent or explained too little variability to be in-
cluded in the equations. These were pH of soil at a depth of 1 to 4 inches, position
of plot on slope, season-long percentage of direct sunlight, and distance from drainage
bottom to plot.

Within the narrow range of slopes studied, the steeper slopes showed greater dura-
bility than gentle slopes. This result was not expected and may be related to soil
coarseness or some other factor associated with slope rather than slope itself.

7s

3at the time an expert photo interpreter was available, only 29 of the plot groups
had been located on aerial photos. The authors are grateful to Karl E. Moessner of the
Intermountain Forest and Range Experiment Station, USDA Forest Service, for making the
aerial photograph measurements used in this study,

Bs he J abe SAY Sikes taal es a

4 ne tps mae gh: 4b oe ee fz pe Sete
i a wy. 7 y i“ tee

r ici et) gations 9” bine. ‘epoch $3458 ness 00H o anne ploy
sere zeae Paw tb We. ny Seats 8 aby sawete ley (8 ee oe Pes

the. 2 Teva yeaaly agin estoassaey | Vw, DONG Bt Eee aan :
eo reer baits ?y “deh s Sie) Foes 3 Py me Pi oe tip: oe eh . oF
= 5h nis aaslt sredeeny 2p "heel nm err iae e ebay arn Pen \

i Shtegely ti hoz tha defend rey Wr to Ey! Cfod ait Foe
ini led 81S geet atagia Tle yoacbo ong ot SRO Brinda. we
amit eg pieags? ce tnblos: tid chem Avinia Nar. | Lagrepe th sare iS

| oil tig

ge? -ahdelrun trobrsos) To. seiPeek w o20)

rare cf thpetr Sema “ut, Pee: Sava io tour Tedag ye wr G204 Cia
(0. o0s Viwtie COL te fogey oft te gAAIhsot) & = GuersTakA Siw aaobebc ras:
Saad pote 2 ti ahdR IY, Vestuiogel ed toe Saw eyutness (- 2hiT > ety Rigi

oi OyMa par aay Yr piseda. 0G bias a et y ( & Les Fay oSter® jab se fina, 2g 2 Fai
Ti) es et 0) sete ahead) axon Be baddec Yexe vl attr pore ie !
. ( shelad ft: busi Leunsbaesres.

» Aes ae

sa CPE MaR te) “Mls OOP aduetreg 10 Ml enmeborqiny .2'f bie 1619) oan Met
ae.) aS eatevrg? S6Fsadi et hawt oc iA 40) 2099. Gosan se iets i:

WT Shigeki a Re FA] (EES 5 mong Vkrer Gest i weosrr risa Joi! Kipea> r¢9
ADS Blea pPrdeniy Vebine tyra VSi.45 py? 2 183 imtd n> Semen Lidiee
; 2tik rT sal wh Ac hi ”: 4 3g rj iee § rae 2 Slr s wae S Secs aT,
SAG, (7 SU Tod Tha Tan . Ap lates At es a
{eapoqe?7 9) Sut 400s borage oye weEl ivi is > 4 Bre i sot one Seeuee
att =F , i oo i.’ i< i™)\4> tu oT sta > a ’ \ 4 5S ts 3T° 7s rip,
“ree 3 =P a i,
2 Tilo) 7 Al [ui * » 2 bist
Borex: UG? 22. ips 62 J ule 2 Yon tsi «oe Ad tips AD = 75 4954 ith tasloa,
mh! TP. eR Bets wt dost, shosssay -29f Te S aai* es) ogitens bia) Pose eran
7 gh gis . - ae ri to a - :
¥ 9 tad | Hy ott iJrysl yes i +t has) 56
ce Whs breéeky aie ' 5 : iff a » «
i ue 0. Bercy Ah THRO os FOE ni eee eal ees rd ;
to OF 40 BIS” | IN tt), Slee es Jeswols » yr S747 915s Na wo . De
Lhitac els bor > ee iyesh Pry ae" she 'eee"s ee »tee7
PSN > ST SO 5S BUM PGS e See et) om’ S-)) 22041) grogiod gg. gee
OF 2d OF ih itar any Tiel Jap ‘bopiel yes. + TBi 6h06ni teat caw, Bo 8
mrp sasoas > ae fie {Wgdb) . 36 ites Po i eit” 2° os Lavne-eaaig
ADT ‘ WS TS RRAIE IS oe y igithvs Mpeg Fo spesresisy giol-uceese (adie a

bd

Jb we Wy Lewem chy le aggeta oui obi buy. Spg0 240 7A0AG . wovTat iltga
dice wh hase let od ne Drie, hbee sar iv tan 064, Flier; ic o> pert Sle a! $n ae nat
Sheet, Sool s" “ta i> on} ii, ol iw pera n0ras Braet yout “S207

;
—
‘adh oi sate ta

= / bef "fe x

a

syane run a %
wit Se rae
ay av Tava, 2

"66°0 3 YULSTFIUBIS xx "$6°O 38 JUBOTFIUBIS x

SUTYI0OIS
(9p°-) (92°) (1°) x(ZS") uo paseq [TATA
Os veo° 900° - IT° SSicr vIo° OvAs SK -ins jusdted
SUSTOM
? sutddt{o uo
(1z°-) *(S¢°-) (225) (ey) (62°) poseq ‘TATA
Ig° ise Pl = 60° - v0" - >3° S00° 09° ny -ins judd10d
‘4m 3utddtto
(sz*) xx(§S°-) (6z°) ean(SS2) uoseas
8c° Or’ v9°T LESAVS Sv 9s 9%* 29° PI- eX ~FO-pugq
3 xx(OS°-) (pt*-) (42°) »(¢¢"-) (91°) xx (79°)
6S° 09° SOxos Giles CSECCS 99° P- S9°88¢ 86°L CL SES OK
SUTYD0IS
(zz") xx(Z9°-) «x (ZS°-) «x (Sb°) uOSPaS
v9°0 p9°0 bo'T LT°6- 8° P8- vs*9 96°O0PL Ty ~Jo-puq
(8q) | (2a) q) 1q)
Og 6q 8q Iq Sq Tq ‘ eB
n> = poe w bn UD 5. 30 nee) ies] au 3M ep) (e)
Qe 3 r: 2s 2 pp 2 OO or or [o) oo oO or (o}
c Q ct ac n QRH » Oo mm O tr: oO SR p oO ~» Oo 3
pr ra oO mo a= AQ nN nec or < cto nd ud n
MO i= py -_ oO oO c c i) a> oO co co ing
o= ap wp nas 4 Ho aU ct 3 A 8 @
o or 30 » oc ne) ae) = Be ocd ume) oO 3
on ar oc 4 Poa} an an H- ° -h ah ah ct
Hct QF o <tn Hh: bh: 3 pf S| fe)
48 os tro m9 Pt ct oO ct oc ac ileal ou ou
[om 3 ~ fo) 3 Fr: 3 br: t= co it) 3 Fe 3 Fe
Ho < vo ° os ° ° o x af ° °
sy o xo Fh Fh O oe} oa 3 w a ° uc oq ct
=) 4 —_ rh n a K Y it) a7 He
bh >< co aa oo oo a Ao (ov ae (e)
7%) f- MW el lam (rests) ces ~< +o is
lon ~ re oO fo fe) =) fon) RY oa 3
- fe} o fo) nn an Y 30 np a
o uc) n ral lon i QF: ct
wn @ = mo mo se aH: =a
> <TD <'d ® <0 x
ies ee we y eae LD
zu ° » r u a
US TOTFJI09 = Une ars
uotssoi3z01 oa

atdrirny

*sasayzuaind ur umoys aap squaro1fface uo1ssedbad 1D24uvd pavpunig
"YD4) UlaYyZAOU 64ZSad0J 7DU01ZDN aYyoD) “vADP 996 £SaZ21S Uo1ZvadoOaA 1D14Uez0d
fo fiz111qQounp ay, zo1pead 02 fipnyg ‘szapom uorssaibad aarf aof sqqnsad fo Aapuung--"{[ eTqeL

We, ey, aii

te
r : 5 hy Gekiep ox.
, aie eg T ‘o
WE i, Brito " ul igs
t : ‘s = Se : = me I
| - ; “4 = 8
i ee Ce et
ae age! a Pt et et eee on his sa
4 ave ;
= «4
to ee oe oe
1 ge ge Re TIME te) eS he” Aes
|
«Al fee ar of pigs 2 fi
; ; o = U va eli “mn )
x i —
‘ae
7 5.4 ,
SL eane a wee

=

100

Figure 3.--The combined
effects of slope and
aspect on survival of
trampled vegetation,
1966 data,
Cache Nattonal Forest,
northern Utah. 75

50

END-OF-SEASON STOCKING (Percent)

25

5 10 15 20 25 30 35
SLOPE (Percent)

Trampled vegetation is vulnerable to severe heat and drying. Consequently, survi-
val of vegetation was greatest‘on northeast (coolest) aspects and decreased as location
approached the southwest (hottest) aspects. As shown in figure 3, this effect was
accentuated by slope steepness. ’ :

The coefficients for elevation were negative (table 1), an indication that the
amount of surviving vegetation decreased as elevation increased. Two factors may ex-
plain this decrease at higher elevations: the season was shorter than that at lower
elevations and the vegetation less well developed at the start of trampling. Apparently,
the extra precipitation at higher elevations in the area studied did not offset the
effects of a late season and low temperatures.

Although percent of clay at a soil depth of 1 to 4 inches was not significant, the
coefficient for this variable was positive in all regression models in which it was
tested. This indicates that, for the conditions studied, ground-cover vegetation sur-
vives best ‘in soils that have a relatively high clay content.

fx0g “4 war! ‘ya bind a pp 2 ie. add La ‘ie eur Tne ne 1 phe

coh tna! ede en aie Tame Spon 5 iis a
. ad Sgt Py” Aue 5! A ore

7 4 ohne ae ek: anit iby Tecoma y od Ings ie,
- ‘gh rae ae ; 3 rat bib. @ iris an

ie

. wy ‘ rab ig i

CONCLUSIONS

Although results from a single growing season may not reflect the cumulative
deterioration of vegetation in recreational areas over a period of years, several con-
clusions seem to be warranted:

1. It is possible to develop prediction equations for rating the durability of
potential recreation sites. In fact, it may be possible to develop equations that will
enable recreation managers to rate site durability from aerial photograph measurements
alone. Such measurements would be much less expensive than on-the-ground evaluations.
Lindsay (1969) reported that recreation areas could be accurately selected from aerial
photos. He used seven criteria for selection. Durability would be an important addi-
tional criterion for selection.

2. In the development of equations for recreation site durability, end-of-season
stocking seems to be a more effective dependent variable than the end-of-season weight
of vegetation or than either the stocking or the clipping weight of treated plots
expressed as a percentage of the same measurement for control plots. However, the
reader should recognize that stocking measurements can be misleading, especially after
a single season of treatment. Vegetation may be severely damaged and still give a high
stocking percentage. If damaged plants disappear in subsequent seasons, first-year
stocking measurements will not have given a valid indication of long-term site durability.

3. Use of a concrete roller to simulate trampling was effective. Simulation
procedures permit the researcher to select the range of site conditions he wishes to
examine and yet avoid the great variability associated with actual recreational use.

4. If simulated trampling is used to study areas grazed by livestock or wildlife,
plots should be fenced.

5. Finally, it would be desirable to rate site durability under natural conditions
and also under conditions of management. This is being done in a continuation of the
study reported here.

. LITERATURE CITED

Lindsay, John L. : i
1969. Locating potential outdoor recreation areas from aerial photographs. J.
Forestry 67:33-34,

Spurr, Stephen H.
1962. A measure of point density. Forest Sci. 8(1):85-96.

Wagar, J. Alan,

1964a. The carrying capacity of wild lands for recreation. Forest Sci. Monogr.
Uy COM io

1964b. The insolation grid. Ecology 45:636-639.

Lyre)

a

Afi

fi WEES Nt antie

yearcae4 it?

Me

Cs j
Cay
‘ Ligz Hy ryt :
{ SEN OTe ae:
eee. 4 + ater ss ait aes
i r ns fc c \ ‘ TO its £6, nye * Ln “YY bebe
fa \ nay ( en arth ae ao
1 { kM A i"

nk epee fae Pree
eh + z . 7 > Li Age) at
ty m9 j ' is et 2 ¥ Sif 7% m, Tika 29
fore mn} ni ey fe

Y Bowes fy
er a eS a Be

iT !
Siuow
Lf 3 fi '}2 re
; bi
if
Wh:
, q apreiap gars! ity tg Sr Pe: T
i
Oe wet ; :
td j 3D Lal ors
fi ote “4 4 tre } }
Oi MEE 7 Se hies
‘ avis

ee
om
if

fy f ig ot at eS

‘ x. Cur te tey kann (elo. 1