Society for Range Management

Can Spring Cattle Grazing among Young Bitterbrush Stimulate Shrub Growth? Author(s): Dave Ganskopp, Tony Svejcar, Fred Taylor and Jerry Farstvedt Source: Journal of Range Management, Vol. 57, No. 2 (Mar., 2004), pp. 161-168 Published by: Allen Press and Society for Range Management Stable URL: http://www.jstor.org/stable/4003914 . Accessed: 12/11/2013 06:59
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Allen Press and Society for Range Management are collaborating with JSTOR to digitize, preserve and extend access to Journal of Range Management.

http://www.jstor.org

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

J. Range Manage. 57:161-168 March 2004

Can spring cattle grazing among young bitterbrush stimulate shrub growth?
TONYSVEJCAR,FREDTAYLOR,ANDJERRYFARSTVEDT DAVEGANSKOPP,
Eastern Ore. Agr. Res. Ctr., 67826-A Hwy. 205, Burns, Ore. 97720; WildlifeBiologist, USDI-BLM,28910 Hwy. Authorsare Range Scientists, USDA-ARS, 20 W.,Hines, Ore. 97738; Habitat Biologist (retired), Oregon Dept. Fish and Wildlife,Box 8, Hines, Ore. 97738.

Abstract
Due to its palatabilityand forage quality,antelopebitterbrush (Purshia tridentata PurshDC) is a desirableshrubacrosswestern US rangelands.Becauselittle informationis availableregarding grazing managementof young bitterbrush,a study was undertaken to explore stocking pressure thresholds and quantify effectsof light and heavy springcattle grazingon shrub growth. Rates of browsing and tramplingand forage availabilitywere monitoredover 3 years in southeastOregon.Acrossyears, 29% of bitterbrushendured trampling in light-grazingtreatments, and 55% experiencedtrampling under heavy grazing. Linear modelsrelatingtime and cattle densitysuccessfully explained(r2 = 0.84-0.86) probabilities of bitterbrushbeing trampled.Forage utilizationaveraged32% and 59% in lightlyand heavilygrazed units,and 14 and 62%of bitterbrush werebrowsedin lightlyand heavily-grazed pastures, respectively. Cattle began browsing when herbaceous standing crop declined to 100-150 kg ha-'. Browsingin heavily-grazed pasturesreduceddiametersof bitterbrush by 4.5 to 9.5 cm in 1998 and 1999, but shrub height was unaffected. Lightly-grazed stands exhibited a 50% greater increasein bitterbrushdiameter,30% greaterheightincrement, and 8% longer twigs than shrubs in ungrazedpastures.At the end of the 1997and 1998growingseasons,bitterbrush in heavilygrazed pastureswere 11 cm greater in diameterthan ungrazed controlsand equalto shrubsin lightly-grazed pastures.To stimulate bitterbrushgrowth,young stands can be lightly-grazed (30 to 40% utilizationof herbaceousforage) by cattle when bitterbrushis floweringand accompanying grassesare in vegetativeto late-bootstagesof phenology.

Resumen

pronghorn (Antilocarpa americana), and livestock on western

Debidoa su gustocidady calidadde forraje el "Antelopebitterbrush"(Purshiatridentata Pursh DC) es un arbustodeseable en los pastizalesdel oestede U.S.A.Debidoa que hay poca informacion disponible respecto al manejo del apacentamientode plantas j6venes de "Bitterbrush", se condujo un estudio para explorarlos umbralesde la presionde apacentamiento y cuantificar los efectos del apacentamiento ligero y fuerte de ganado en primaverasobre el crecimientodel arbusto.Durante3 aiios en el sudestede Oregonse monitorearonlas tasas de ramoneoy pisoteoy la disponibilidad de forraje.A travesde los anlos,29% del "Bitterbrush" toleroel pisoteoen los tratamientos de apacentamientoligero y en los tratamientosde apacentamiento fuerte 55% del "Bitterbrush" sufrio el pisoteo. Modeloslinealesrelacionandotiempo y densidadde ganadoexplicaronexitosamente (r2= 0.84-0.86) las probabilidades del "Bitterbrush" de ser pisoteado.La utilizaciondel forrajepromedio32% y 59% en las unidades apacentadas ligera y fuertemente y 14 y 62% del "Bitterbrush" fue ramoneadoen los potrerosapacentados ligera y fuertemente respectivamente.El ganado inicio el ramoneo cuandola biomasade las herba&ceas diminuyoa 100-150 kg ha-'. El ramoneoen los potreroscon apacentamiento fuerteredujolos diametrosdel "Bitterbrush" de 4.5 a 9.5 cm en 1998y 1999,pero la altura del arbusto no fue afectada. Las poblaciones con apacentamiento ligero mostraronun incrementodel 50% en el diametrodel "Bitterbrush", 30% ma'sen la altura y las ramas fueron8% mas largasque las de los arbustosen los potrerossin Al final de las estacionesde crecimiento apacentamiento. de 1997 y 1998el "Bitterbrush" de los potrerosapacentados fuertemente fueron 11 cm ma'sgrandesen diametroque los arbustosde los potreroscontrolsin apacentamiento e igual a los arbustosde los potrerosligeramente apacentados. Para estimularel crecimiento Key Words:Purshiatridentata, livestock,big game,winterrange, del "Bitterbrush"las poblacionesjovenes pueden ser ligerawildlife,habitat,browse mente apacentadas (30 a 40% de utilizacion del forraje herbaiceo) por el ganadocuandoel "Bitterbrush" esta en floraciony Antelope bitterbrush (Purshia tridentata Pursh DC) is an los zacates acompafiantes estin en las etapasfonologicas de crecimportant shrub for mule deer (Odocoileus hemionus hemionus), imientovegetativo a finesde embuche.

NorthAmericarangelands (Kufeldet al. 1973, Vavraand Sneva 1978, Neal 1981, Urness 1981, Kinuthia et al. 1992). Witha sustained crude protein content above 8.0% (Hickman 1975 and Kitukuet al. 1992), bitterbrush can substantiallyenhance lateseason diet qualityof ruminants when nutritivevalue of herbaceous forageshas declinedto sub-maintenance levels (Ganskopp and Bohnert2001). Bitterbrush occurs among numerousvegetaEastern OregonAgricultural Research Centeris jointly operated by the Oregon Agr. Exp. Sta. of Oregon State Univ. and the USDA-AgriculturalResearch Service. Manuscript accepted1 Jul.03.

tion types across approximately140 million hectares(Hormay 1943) fromBritishColumbia to California andeast into Montana andNew Mexico (Cronquist et al. 1997). Senescence, wildfires, a history of excessive herbivory,and low recruitment have, however,decreasedbitterbrush abundance acrossmuchof its range(Billings 1952, TuellerandTower 1979, Winward andAlderfer-Findley 1983, Ayerset al. 1999,Clements and Young 2001), and restorationefforts have frequentlymet with limited success (Hubbard 1964, Kituku et al. 1995).

JOURNAL OF RANGE MANAGEMENT57(2) March 2004

161

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

Hubbard (1957) and Dealy (1970) showed Methods thatcompetingvegetationcm substantially reduce establishment and subsequent Study site history and description statureof bitterbrush seedlings,and plants In 1990, wildfire charredabout 30,000 mustreachan age of 60 to 75 yearsbefore ha of Bureau of Land Management maximum annual yield is realized (BLM), United States Forest Service, and (McConneland Smith 1977).Burningtyp- deededpropety northand west of Bums, ically resultsin greatermotalit of bitter- Ore. Much of the area consisted of pinebrush than does clipping or rotobeating forest/sagebrush-steppe-transition range (Muegglerand Blaisdell 1958, Clarket al. which had historically been graed by cat1982), and heavy browsing may also tle and was important winter range for reduce plant longevity (McConnel and mule deer and elk (Cervus elaphus Smith 1977). While young transplants nelsoni). Revegetation efforts began appearto benefit from grazingprotection immediately, and BLM properties were (Dealy 1970, Ferguson1968), established seeded with 'Secar' Snake River wheatbitterbrush generally responds well to grass (Elymus lanceolatus (Scribner & defoliation, and grazed plants produce J.G. Smith)Gould)at 9 kg ha, andin aeas con- where it had prviously existed, antelope more and longertwigs thanungrazed was includedat a rateof 2.2 kg trols (Tuellerand Tower 1979, Billbrough bitterbrush seed was acquired ha. bitterbrush The and Richards 1993, Kituku et al. 1994). but its collectionlocale was Ganskoppet al. (1999) found that cattle commercially, only browsed bitterbrushlightly before unknown. Soil in the ea (43 37'N 119024'W, elecompeting grasses entered anthesis, and vation was a m) montmorilfine, 1,584 that shrubsin pastureslightly-grazedby cattle early in the growing season were lonitic,frigidTypic Argixeroll.Vegetation overlargerin staturethan cohortsin ungrazed priorto the fire includeda scattered western juniper dominated story by controls. Spring mowing of competing occidentalis Hook.), a shrub vegetation also stimulates twig growth (Juniperus layer characterized by mountainbig sage(Kituku et al. 1994). Accelerated twig brush (Artemisia tridentata subspp. growthmay not occur on less productive vaseyana(Rydb.)Beetle) with a minorbitsites (Kituku et al. 1994), and in some terbrush andherbaceous vegecomponent, of twigs may also tationdominated years annualproduction by bluebunch wheatgrass be affectedby extremelylow temperatures (Agropyroi spicatum (Pursh) Scribn. & (Jensenand Umess 1979) ador fluctua- Smith) and bottlebrush squirreltail tions in annual precipitation (Garrison (Sitanionlystrix (Nutt.)Smith). 1953, Kindschy1982), In the absenceof competingwoody vegMost bitterbrushresearch has focused etation, surviving herbaceousplants and on well-establishedstandsor reclamation emergingseedlings respondedwell in the efforts, and there is little data regarding growing seasons following the fire. grazing management of young stands. Bureau of Land Management sampling Given that light, early season cattle graz- detected8,450 bitterbmsh seedlingsha-'in ing stimulatedgrowth of 3 to 5-year-old 1991 and 3,410 ha' in 1992. Crop-year which is bitterbrush et al. 1999), a study precipitation(September-June), (Ganskopp was designed to evaluate bitterbrush highly correlatedwith annualforage proresponsesto even heaviergrazingapplica- duction in the region (Sneva 1982), was tions restricted solely to the earlygrowing 91 and 86% of the long term mean (255 season. The objectives of this research mm, n = 40) for the 1991 and 1992 growwere: 1) to determinethe effects of light ing seasons, respectively (NOAA 1990and heavy early season cattle grazing on 1999, Squaw Butte Station 43029'N In accordance with BLMpolithe subsequent statureand twig growthof 119043'W). livestock grazing was not allowed in cy, 6-9 year old bitterbrushand 2) explore and 1992 to 1991 aid of vegetarecovery relationships among indices of forage and the charred remains of small tion, availability and levels of browsing on trees and shrubs provided the only evishrubs to determineif stocking pressure dence of the fire fter2 growingseasons. thresholdsfor management of young bitterbrush stands could be established, These were accomplishedby monitoring Pastures and grazing schedules Projectdesign was a randomizedcomratesof shrubuse, trampling damage,forage availabil:ity, and sta re of young bit- plete block having 3 replications and 3 terbrushin lightly, heav:ilyf and ungrazed treatmentswith individualpastures(N= 9) serving as experimental units. pastures. Treatments includedpastures thatwere: 1)
162

lightly-grazed by cattle, 2) heavily-grazed by cattle, and 3) ungrazed controls. Treatmentswere not randomizedin succeeding years, so findingsreflectcumulative effects applied across 3 consecutive growingseasons.Also, teatment designations reflect cattle applications only, even thoughsummering pronghom (Antilocarpa deerandelk had american) andwintering
free access to all pastures.

Size of grazedpastures rangedfrom0.6 to 0.9 ha and ungrazed controls were between 1.4 and 1.6 ha. In 1997 grazed pastures were stockedon 5 Maywith yearling Hereford x Angus bulls weighing from 317 to 363 kg. Lightly-grazed pastures supportedI animal, heavily-grazed pasturessupported 2, and grazingcontinued for 20 days. Given our desire to conto the boot stageof fine grazingtreatments phenologyfor the grassesandremovecattle bfore rapid elongationof bitterbrush twigs, we doubledthe ninti stockingrates for the next 2 grazing seasons. In 1998 lightly-grazedpasturessupported2 cows and heavily-grazedpasturessupported4
cows with the trial beginning on 21 May

and lasting for 12 days. Dry Herefordx Anguscows witha mea weightof 453 kg (SE = 26) were used. In 1999, 18 cows were drawnfrom a groupof dry animals scheduled for culling,andmeanweightwas slightly heavier (V= 567 kg, SE = 14.5).
Pastures were stocked on 4 June and graz-

ing continuedfor 10 days in 1999. Given the increase in stock numbers between 1997 and 1998, andthe use of successively larger animals as the trials progressed, effective stockingrate (AU ha-')progreseachyea sively increased

Shruband vegetation sampling

Endpointsof 91-m line transectswere markedwith metal staes in each pasture and the position along the tape and distance left or fight of the tape recorded for

25 randomly-selected bitterbrush in each pasture.Throughout the trials, any bitterbrushlost to completedefoliationor mortality was replacedwith the nearestavailable neighbor.Priorto each grazing session, the dimensions (greatest diameter and height) of bitterbrush were measured in all pastures, andthe ends of any recently defoliatedtwigs maked with black ink to facilitate detection of subsequent browsing.To index mass of standing crop, all herbage was clipped from ten, 1-mi plots in each pasture just before stocking oven dried at 40C, and subsequently weighed Durnug grazingtri 5 werturned evey 2 days, relocated eachbittrush, andtabl

JOURNAL OF RANGE MANAGEMENT 2004 57(2) March

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

lied evidenceof recentdefoliationor trampling. Defoliation signs included twigs displayinga clipped appearance with new yellow wood on the ends or barkskinned fromthe last few millimetersof the terminal by a cow's biting and pullingmotions. Broken stems and twigs or recently displacedbarkthat revealedbare wood were considered signs of trampling.Stocking was terminated in all pastures when at least 50% of the herbaceous forage had been utilized or 80+ % of the sampled shrubsexhibited some sign of utilization in the heavily-grazed treatment. Aftercattlewereremoved,standing crop was sampled by clipping ten, 1-m2 plots per pastureto estimateforage utilization, and the greatest diameter and height of each shrub was recorded to facilitate before and after grazing comparisonsof bitterbrush stature.Finally, shrubdimensions were measureda third time in late August to assess spring treatmenteffects on subsequent summer growth. Again, greatest heightanddiameter wererecorded for each shrub.In addition,length of currentseason'stwig growthwas talliedfor 2 randomly-selected twigs on eachplant.

Statistical procedures

A repeated measures analysis of variance was used to evaluatetreatment (N = 3), year (N = 3), date (N = 3), and 2-way (4 df) and 3-way (8 df) interaction effects on shrub diameter and height. Repeated factorswere years and dates withinyears. Analyses incorporated a first order, autoregressive covariancestructure among repeatedmeasures.This is an alternative to adjusting the degreesof freedomfor the lack of independence amongdates as sugResults gested by MillikenandJohnson(1984). A log transformation helped stabilize variPrecipitationpatterns ances among treatments,years and dates Sneva (1982) established thatcrop-year for diametermeasures,but transformation forage accumulation in the region was was unnecessaryfor shrubheight. Single most closely correlated with precipitation degreeof freedomcontrasts were used for totals for the previous Sept.-Juneperiod. mean separations betweentreatments on a The weather stationwith the longest congiven date. Withintreatments, a pairedt- tinuous recordnear our study site was on test (2 df) was used to test for changesin the Northern Great Basin Experimental shrubdimensionsbetween adjacentdates Range (referenced as the Squaw Butte (SnedecorandCochran1967). ExperimentStation in N.O.A.A. (2001) Forresponsevariablesmonitored on an documents) some 29 km southwest and annual basis (i.e., measures of standing 161 m lower than our study site. Mean crop, twig length, levels of forage utilizacrop-year precipitation (n = 64) for the tion, and cumulative number of shrubs Experimental Rangeis 26.1 cm, and accugrazed or trampled),a split-plot analysis mulations for the 1997-1999 crop-years of variance was used to evaluate treatwere 132, 196, and 106 percentof averment, year, and interaction effects. Treatments(2 df) served as whole plots, age, respectively.Given the disparitiesin elevation and vegetation between the with the block x treatmentinteraction(4 Experimental Rangeandour studysite, we df) serving as error term 1. Years were estimate our researchplots receive about
JOURNAL OF RANGE MANAGEMENT57(2) March 2004

subplotswith year (2 df) andthe treatment x year interaction(4 df) tested with the residualerrorterm (12 df). This approach was used because year effects cannot be randomized, and one is forced into an inadvertent split-plot (Cody and Smith 1997). Again, if effects were significantin analysesof variance,single degreeof freedom contrasts were used for mean separations Regression analyses were used to explore and quantifyrelationships among the numberof shrubsgrazed or trampled (dependent and severalindepenvariables) dent variables. Independent variables includedexpressionsquantifyingpassage of time, forage availability, numbersof cattle, pasturesize, and several combinations of these [i.e., days grazed,AU days, area available (ha), forage available (kg and kg ha-'), stockingrates (AU ha-', AU days ha-'), and stockingpressure(kg forage AU-', kg forageha-'AU-', and kg forage AU-' day-')]. Scattergramsrelating numbersof bitterbrushtrampledto passage of time suggested linear models might suffice. Depictions of bitterbrush grazed and passage of time suggested a curvilinear functionmightbe required. Given a high degree of variability in herbageproduction amongpastures, statistical significancefor analysesof standing cropandforageutilization was acceptedat P < 0.10. Statistical significance in all other analyses and for mean separations was acceptedat P < 0.05. Throughout the manuscript, numbers followinga "+"symbol arestandard errors(SE) of the mean.

10 cm more precipitationannually than the Experimental Range.

Herbaceousstanding crop and forage util zation

Herbaceousstandingcrop before grazing averaged652 kg ha-' and varied(P = 0.05) amongyearsbut not treatments (P = 0.28). For the 1997-1999 samplingperiods, mean herbageproduction was 777 (? 68), 596 (? 45), and 583 (? 47) kg ha-', in earlyMay. Givena history respectively, of light springuse in our grazedpastures, our samples contained a substantial amount of cured materialfrom previous growing seasons for our initial 1997 harvest. Materialswere not sorted,however, so we can not quantitatively address live:deadratios. treatment Grazing (P = 0.01) andyear(P = 0.06) effects were significantfor levels of forageutilization by cattle,butthe treatmentx yearinteraction was not (P = 0.34). Mean forageutilizationwas 32 (? 7)% in lightly-grazedpasturesand 59 (? 9)% in heavily-grazedunits. Across treatments, herbageutilization by cattleaveraged 35 + 7, 37 + 8, and65 + 14%for the 1997-1999 trials,respectively.

Defoliation and trampling of shrubs by cattle

The numberof bitterbrush browsed by cattle during the trials differed among treatments (P = 0.002) and years (P = 0.04), but no treatment x year interaction (P = 0.11) occurred. When trials ended, an averageof 14 (? 5)%of the bitterbrush werebrowsedby cattlein the light-grazing treatment, and 62 (? 9)%were browsedin heavily-grazed pastures (P = 0.05). No bitterbrushwere defoliated in the ungrazed controlswhen cattlewere on site (datanot shown). Cattle did not initially forage on bitterbrush, but beganbrowsingthe plants aboutdays 5-6 in the heavily-grazed treatmentandaboutday 8 in the lightly-grazed pastures (Fig.1). Among regressionsrelatingthe cumulative numberof shrubsgrazed (dependent variable)to the passage of time and various expressions of herbaceous standing crop or stocking pressure, best fit (r2 = 0.63, P < 0.001) was obtained with an exponentialdecay function(Fig. 2) where stockingpressure,expressedas kg forage ha-' AU-' at the close of the trials, served as the independent variable. Approximately half the bitterbrush were browsed by cattle when standingcrop was reduced to about 75 kg ha-' AU-I. The rate of browsingon bitterbrush also appearedto

163

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

90

-_

__l

-0-e 0 60 6 40 30

~ 1997

220
CO) 10

0 2
90 -

777 4 6

87070*60 o 0-eX A2 40
60

1998

2201
C,,' 2 4 6 8

90T
60
0O

~301 2201
(01 2 4 6 8

50 430

increase when standingcrop went below the 75 kg ha-'level. Light grazing and year effects were Grazingtreatment + Heavy grazing significant(P < 0.01) for tramplingdamage, but treatmentsresponded similarly amongyears (P = 0.07). Across years, 29 (? 4)% of the shrubs were trampled in light-grazing treatments,and 55 (? 5)% were trampledunder heavy-grazing. No trampling effects were noted in control 10 12 14 16 18 20 pastures.For individualyears acrossthe 2 grazed treatments,percentagesof shrubs trampledwere 28 + 5, 49 + 7, and 49 + 7%,from 1997-1999, respectively. Rates of tramplingdamage to shrubs were well correlatedwith expressionsof Lightgrazing Heavy grazing / stocking rate (AU ha-')) and time (days). With linear models and cumulativecattle days ha-' regressedagainstthe numberof shrubstrampled,coefficients of determination (r2)rangedbetween 0.84 and 0.86 10 12 14 16 18 20 (Fig. 3). Slopes of the regressionmodels ~~~1999 differed (P < 0.01) among years and increased slightly as the study advanced from 1997 through 1999. Only I shrub, located where the cattle frequentlybede0- Lightgrazing ded, diedfromtrampling damage. + Heavygrazing

10

12

14

16

18

20

Days of grazing
Fig. 1. The percent (? SE, N = 3) of bitterbrush browsed by cattle in lightly- and heavily-stocked pastures as grazing trials progressed in the spring of 1997-1999 on big game winter range in southeast Oregon.

25 z 20
e 15

y=aebx
=23.3 ~~~~~~a b =0.0082

2
,, 10
.0

r2=0.63 se est. = 5.1

.
0

0
0 100 200 300 400 Kg forage ha-1 AU-1

500

Fig. 2. The relationship between the number of bitterbrush grazed by cattle (out of a total of 25 monitored shrubs) in lightlyand heavily-stocked pastures in southeast Oregon and kg forage ha'l AU 1 during spring stocking periods of 1997-1999.

Among analyses of shrub height, all main effects (treatment, years, and dates) were significant (P < 0.03) as were the treatmentx date and year x date interactions (P < 0.01). Bitterbrush heightswere when the study similaramong treatments began(Fig. 5), andunlikeshrubdiameters, browsingby cattle did not reduceheights of bitterbrushduring any of the grazing trials.Fall andoverwinter browsingby big game, however,did reduce(P < 0.05) the in all treatments over heightof bitterbrush the 1997 and 1998 fall/winter periods. Light spring grazing among bitterbrush stimulated summer height growth compared to the other 2 treatments(Fig. 5). Across years, shrubsin the lightly-grazed pastureswere more than4 cm taller (P < 0.04) at the end of the growing season than cohorts in the heavily-grazed and ungrazed control treatments. With the exceptionof the heavilygrazedpastures in were typi1999, shrubsin all treatments cally taller in the fall than they were just before grazing started in the spring. In 1999 in the heavilygrazedpastures, bitterbrush were 37 cm tall in the spring and only 38 cm tall (P = 0.6) at the end of the growing season. Two factors likely conCanopy diameter and height of bittributed to this lack of growth.First, 1999 terbrush was the driestyear of the study, and secWith the exception of a grazing treat- ond, we observedthe highestlevels of forment x year interaction(P = 0.33), all 3 age utilization(87 + 7%)andbrowsingon main effects (treatments, years, and sam- bitterbrush in the heavilygrazedtreatment pling dates), 2-way, and 3-way interac- duringthatyear. tions had significanteffects (P < 0.05) on Whenthe trialsbeganin canopydiameter. 1997, bitterbrush in grazed treatments Current season's twig growth Significant treatment (P = 0.03) and were approximately 8 cm wider(P = 0.02) year (P < 0.01) effects occurred for current than bitterbrush in the ungrazedcontrols (Fig. 4). When cattle were removedfrom season's twig growth, but treatment the grazedpastures20 days later,howev- responses were similar across years (P = reflectedannual of shrubswere similaracross 0.28). Growthincrements er, diameters precipitation and were similar(P patterns, = all 3 treatments(P 0.07). Subsequent spring/summergrowth of shrubs in the = 0.16) for the 1997 (12.6 cm) and 1999 grazed pasturesincreaseddiametersby 7 (10.5 cm) growing seasons. Currentyear to 8 cm in 1997, while shrubs in the growthwas nearlytwice as long (24.2 cm, ungrazed controlsonly expandedby about P < 0.01), however,in 1998, the wettestof 3 cm (P < 0.01). In both 1998 and 1999, the 3 years sampled.Twig growthin the = 16 cm (lightly-grazed browsing by cattle in the heavily-grazed grazedtreatments = 16.7 cm) was similar pasturesreduced(P < 0.01) the diameter andheavily-grazed of bitterbrush suchthatthey werethe same (P = 0.20), and shrubs in both of the width as bitterbrush in the ungrazedcon- grazed treatmentsproducedlonger (P < trols. Compensatorygrowth occurred in 0.05) twigs than shrubs in the ungrazed (14.8 cm) eachyear. the heavily-grazedpasturesin both 1998 pastures and 1999, however, and bitterbrush were greaterin diameter thanungrazed controls Discussion (P < 0.01) and equal to the lightly-grazed treatment when fall dimensions were recorded. Overwinter browsing by big Earlier studies have generally estabreduced(P = 0.05) shrubdiameters lished that livestock progressively congamne in grazedpastures afterfall 1997 measure- sume morebitterbrush as the growingseaments (Fig.4), butbig gamne hadno affect son advances (Lesperance et al. 1970, on bittrbrush diameter thereafter. Stuth and Winward 1977, Neal 1981,
JOURNAL OF RANGE MANAGEMENT57(2) March 2004

164

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

Z 20 E 15
U)

Lightgrazing Heavygrazing

Q 10-

1997
0

*
0

._
73

r2 =0.84

slope =0.339

E
0

O 0

00

10

15

20

25

30

35

Z 20o E 15
10 0

- *

Lightgrazing Heavygrazing
o
*

*
0

1998

*o0

sn~~~~~ 0~~~~~

'
o

5-.o ~~00
0 o0

@0

E 0 o Z 200
r

0 0 5

6
I

*
I 1

r2=0.85 slope=0.483b
. I

10

15

20

25

30

35

Light grazing

* 0
0

E 15 10 0

Heavygrazing

1999
.
0

0
0~~~0

.
o
00
0

5
O

E 0

r =0.86 slope 0.636c 10 15 20 25 Cumulativecattle days ha-' 30 35

Fig. 3. The cumulative number of bitterbrush trampled by cattle in lightly and heavilystocked pastures monitored at 2-day intervals as grazing trials progressedin the spring of 1997-1999 on big game winter range in southeast Oregon. Slopes of regression lines sharing a common letter are not significantlydifferent (P > 0.05).

Umess 1981, and Ganskoppet al. 1999). In our earlierresearch,about 25% of the shrubs were browsed by cattle when accompanying grasseswere in the vegetative to late-bootgrowthstage and stocking rateswere between 11 to 19 AU days ha-' (Ganskoppet al. 1999). This study, with stockingratesin the light treatments ranging between 11 and 22 AU days ha-' (x =

16.5 + 1.1), support those conclusions with less than 25% of the bitterbrush browsed during 2 of the 3 periods sampled. A single exceptionoccurred in 1999 when 11 of 25 shrubsexperiencedsome defoliation in 1 replication of our lightgrazingtreatment. Withour heavierstocking rates(range=23-37 AU days ha-',(K= 28.3 ? 2), however,57 (? 7)% of the bit-

terbrush experienced some degree of browsing (Fig. 1), andmeanshrubdiameter was reducedby 4 cm in 1998 and 10 cm in 1999 (Fig. 4). Two aspects of our data suggest cattle prefer grazing to browsing within this shrub/grass communityearly in the growing season. First, in the early days of the trials, cattle in both grazing treatments essentially ignored bitterbrush(Fig. 1). Thereafter (days 6 to 12) they began foraging on bitterbrush. Second, cattle in the heavy grazingtreatment startedusing the bitterbrush about2 to 6 days sooner than cattle in lightly-grazedpastures.These 2 points suggest forage availability,or possibly foragequality,must decline to some criticalthreshold beforethe cattlebegin to browse upon bitterbrush. The negative relationshipbetween the total numberof shrubsgrazedand availableherbageat the end of the trials(Fig. 2) lends credenceto the first argumentand implies that cattle when began actively seeking bitterbrush available herbage declined to 100 tolS0 kg ha-'AU-'. BalphandMalechek(1985) investigated tramplingof grass tussocks by cattle and noted a disproportionatepreference for treadingupon interspaces and an avoidance of the most elevated tussocks. workalso showedthatas plant Subsequent stature decreased, tussocks were more likely to be trampled in short duration grazingprograms (Balphet al. 1989). The bitterbrush in this studywas relativelyuniform in age (7-9 yr) and height (26 to 43 cm). Given those dimensions, they were of the herbaceous essentiallya component canopyandwere not a seriousimpediment to livestock travel.Gutheryand Bingham (1996) discussedthe theoretical aspectsof vegetation tramplingby cattle and suggestedprobabilities were simplytime/density functionswhen grazingwas nonindependentand nonrandom (i.e., short-duration management).The size of our pastures and our stockingrates assuredrelatively uniformuse of the area. Our data supported the Guthery and Bingham (1996) hypothesis, as the independent variable of cumulative cattle days ha-' accounted for about85%of the variability in the number of shrubstrampled by cattle 3). (Fig. Lewis (1980) studied simulatedbrowsing and tramplingamong young conifers andfoundonly the most severetreatments affected seedlingheight or survival.With light andheavygrazing,we found28 (? 4) and 55 (? 5)% of our bitterbrush, respectively, enduredsome degree of trampling, but few exhibited any deleterious, long

JOURNAL OF RANGE MANAGEMENT57(2) March 2004

165

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

3DJ.55

.o

1999

Grazingtreatmient

Ugae

Fig. 4. Canopy diameters of bitterbrush before (pregrazing) and after (postgrazing) cattle turnout and at the end of the growing season (fall) during 1997-1999 on big game winter range in southeast Oregon. Treatment means within a year and sampling period (row) sharing a common upper case letter on the side of a bar are not significantlydifferent (P > 0.05). Means for a given treatment (column) sharing a common lower case letter between adjacentsamplingperiods are not significantlydifferent (P > 0.05). term effects. In most instances, only 1 or 2 twigs were affected, but some shrubs experienced fractured main stems and persisted for the duration of the study with little evidence of stress. We suspect bitterbrush might escape trampling when its lowest branches originate from sources higher than the surrounding herbaceous layer. Grazing of herbaceous vegetation among shrubs removes competing leaf area. This is thought to slow the extraction of limited moisture and nutrients from the soil leaving more resources available for the shrubs (Garrison 1953, Hubbard 1957, Ferguson and Basile 1966, McConnel and Smith 1977, Neal 1981, Reiner and Umess 1982). This was probably the mechanism that stimulated shrub growth in our light-grazing treatment, as light cattle grazing had little if any direct effect on bitterbrush dimensions. Bitterbrush responded favorably over the growing season to the partial removal of the herbaceous component and were wider, taller, and supported longer twigs (P < 0.05) in the fall than shrubs in the ungrazed controls. Overwinter and early spring browsing of bitterbrush also stimulates subsequent twig growth (Garrison 1953, McConnel and Smith 1977, Kituku et al. 1994). This response is attributed to 2 mechanisms: 1) the absence of apical dominance (Tueller and Tower 1979) and 2) altered resource allocation patterns within the shrubs (Billbrough and Richards 1993). Both of these modes of action and a release from competition probably came into play in our heavily-grazed pastures.

Much of bitterbrush's positive response to spring cattle grazing is likely related to its seasonal pattems of growth. With the first warming temperatures in spring, bitterbrush initiates flowering, and little if any twig elongation occurs. As bitterbrush flowers, competing grasses like bluebunch wheatgrass and Idaho fescue (Festuca idahoensis Elmer) pass through vegetative stages of growth and into the boot stage of development. Grazing of the grasses at this time removes both transpiring leaves and apical meristems and retards subsequent root growth and herbage development (Ganskopp 1988). This lets bitterbrush exploit resources that might typically be taken up by grasses at the exact time that twigs are starting to grow. In practice, bitterbrush responses to grazing treatments may vary substantially from year to year and across its range. Kindschy (1982) noted that up to 40 percent of the annual variation in twig growth of bitterbrush can be explained by cropyear precipitation dynamics. Garrison (1953) also advised that managers might not see shrub responses to grazing on arid sites or during dry years when all components of the plant community are affected by limited soil moisture supplies.

E45

40'

230'BA
B raze

A
Light grazing

1999

Ungrazed

PmGraz.

1997

Fig. 5. Canopy height of bitterbrush before (pregrazing) and after (postgrazing) cattle turnout and at the end of the growing season (fall) during 1997-1999on big game winter range in southeast Oregon. Treatment means within a year and sampling period (row) sharing a common upper case letter on the side of a bar are not significantlydifferent (P > 0.05). Means for a given treatment (column) sharing a common lower case letter between adjacentsampling periods are not significantlydifferent (P > 0.05).

166

JOURNAL OF RANGE MANAGEMENT57(2) March 2004

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

Conclusions
To stimulate bitterbrush growth in grass/shrubcommunities we suggest the stand be lightly-grazed (about 30 to 40 percentutilizationof the herbaceous component)by cattlewhenbitterbrush is floweringand accompanying grassesare in the vegetativeto late-bootstages of phenology. Bitterbrushin lightly-grazed stands exhibited about a 50 percent greater increase in diameter, 30 percent greater increase in height, and about 8 percent longertwigs thancohortsin ungrazed pastures.On a conservativenote, most coolseason grasses are sensitive to and easily harmed at this time of year by overgrazing (Ganskopp 1988). Becausemost managers want to sustain both the herbaceousand woody components of theirpastures, grazing to stimulatebitterbrush growthshould probablybe appliedin a deferredrotation program to assurean overallhealthyplant community. Heaviergrazingapplications,with utilization levels approaching 60 percentof the herbaceous component, can have immediateimpact on bitterbrush stature. Compensatory growthoccurred,however, and shrubs in heavily-grazed pastures exhibitedabouta 30 percentgreaterdiameter incrementat the end of the growing season and about 13 percentlonger twigs thanbitterbrush in ungrazed unitsfor 2 out of 3 years.At the end of the growingseason in the thirdyear of our study,however, shrubsin the heavily-grazedpastures were not as tall as cohorts in ungrazed areas. That year was the driest of the 3 sampled, herbage and browse incurred theirhighestratesof utilization,and those 2 factorsprobably combinedto reducethe lengthof the effective growingseasonand abilityof bitterbrush to recover. Cattle in the sagebrush/steppe typically use herbaceous forages in the spring months(Vavraand Sneva 1978) andavoid bittertbrush until the grassesbegin flowering (Ganskoppet al. 1999). Measuresof residualstandingcrop among our grazing treatments suggestedcattle began foraging on bitterbrush in the springmonthswhen standing cropdeclinedtolOO to 150 kg ha-'. In ourenvironment, thatoccurred at abouta 60 percentutilization level, by which time the cattle had essentiallygrazed and then regrazedall availableherbage.We do not recommendthis degree of forage utilization, but should it occur, bitterbrush can respondby generating as much growthas shrubs in ungrazedpasturesif cattle are removedwhile thereis still sufficientsoil moisture for twig growth.

LiteratureCited
Ayers, D.M., D.J. Bedunah, and M.G. Harrington. 1999. Antelopebitterbrush and Scouler's willow responseto a shelterwood harvest and prescribed burn in western Montana. W. J. Appl.For. 14:137-143. Balph, D.F. and Malechek, J.C. 1985. Cattle of crestedwheatgrass trampling undershortduration grazing. J. Range Manage. 38:226-227. Balph, D.F., M.H. Balph, and J.C. Malechek. 1989. Cues cattle use to avoid stepping on crested wheatgrass tussocks. J. Range Manage.42:376-377. Billbrough, C.J. and J.H. Richards. 1993. Growthof sagebrush and bitterbrush following simulatedwinterbrowsing:mechanisms of tolerance. Ecol. 74:481-492. Billings, W.D. 1952. The environmental complex in relationto plantgrowthand distribution.Quart. Rev. Biol. 27:251-265. Clark, R.G., C.M. Britton, and F.A. Sneva. 1982. Mortalityof bitterbrush after burning and clipping in eastern Oregon. J. Range Manage.35:711-714. Clements, C.D. and J.A. Young. 2001. Antelope bitterbrushseed production and standage. J. RangeManage.54:269-273. Cody, R.P. and J.K. Smith. 1997. Applied statistics and the SAS programminglanguage.Prentice Hall,N.J. Cronquist, A., N.H. Holmgren, and P.K Holmgren. 1997. Intermountain flora:vascular plants of the intermountain west, USA, vol. 3. New York Botanical Garden,New
York.

Dealy, J.E. 1970. Survivaland growthof bitterbrush on the SilverLakedeerwinterrange in central Oregon. USDA For. Serv. Res. Note PNW-133, PNW Forest and Range Exp. Sta.,Portland, Ore. Ferguson, R.B. 1968. Survivaland growthof young bitterbrush browsedby deer.J. Wildl. Manage.32:769-772. Ferguson, R.B. and J.V. Basile. 1966. Topping stimulates bitterbrushgrowth. J. Wildl.Manage.30:839-841. Ganskopp, D. 1988. Defoliation of Thurber needlegrass:herbageand root responses.J. RangeManage.41:472-476. Ganskopp, D. and D. Bohnert. 2001. Nutritional dynamics of 7 northernGreat Basin grasses. J. Range Manage. 54:640-647. Ganskopp, D, T. Svejcar, F. Taylor, J. Farstvedt,and K. Paintner. 1999. Seasonal cattle management in 3 to 5 year old bitterbrushstands.J. RangeManage.52:166-173. Garrison, G.A. 1953. Annual fluctuationin production of some eastern Oregon and Washington shrubs. J. Range Manage. 6:117-121. Guthery, F.S. and R.L. Bingham. 1996. A theoreticalbasis for study and management of trampling by cattle. J. Range Manage. 49:264-269. Hickman, O.E. 1975. Seasonal trendsin the content of importantrange forage nultritive

species near Silver Lake, Oregon. USDA For.Serv.Res. PaperPNW-187,PNWForest andRangeExp. Sta.,Portland, Ore. in California. Hormay, A.L. 1943. Bitterbrush US For. Serv. Calif. For. and Range Exp Sta.,.Res. Note-34,Berkeley,Calif. Hubbard,R.L. 1957. The effects of plantcompetitionon the growthand survivalof bitterbrush seedlings. J. Range Manage. 10:135-137. Hubbard, R.L. 1964. A guide to bitterbrush seedingin California. USDA For.Serv.PSW Exp. Sta., Berkeley, and Calif. Dept. Fish andGameResour. Jensen, C.H. and P.J. Urness. 1979. Winter cold damageto bitterbrush relatedto spring sheep grazing. J. Range. Manage. 32:214-216. Kindschy, R.R. 1982. Effects of precipitation varianceon annualgrowthof 14 species of browse shrubs in southeastern Oregon. J. RangeManage.35:265-266. Kinuthia, R.N., J. Powell, F.C. Hinds, and R.A. Olson. 1992. Range animaldiet compositionin southcentral Wyoming.J. Range Manage.45:542-545. Kituku, V.M., J. Powell, and R.A. Olson. 1994. Sites, mowing, 2,4-D, and seasons affect bitterbrush twig morphology. J. Range Manage. 47:200-205. Kituku, V.M., W.A. Laycock, J. Powell, and A.A. Beetle. 1995. Propagating bitterbrush twigs for restoringshrublands, p. 327-328. In: B.A. Roundy,E.D. McArthur, J.S. Haley, and D.K Mann (comps.), Proc. wildland shruband arid land restoration symposium. USDA-For. Serv., Intermountain Res. Sta., Gen. Tech. Report INT-GTR-315.Ogden, Ut. Kituku, V.M., J. Powell, M.A. Smith, and R.A. Olson. 1992. Increasing bitterbrush nutrient quality with 2,4-D, mowing, and burningin southcentralWyoming.J. Range Manage.488-493. Kufeld, R.C., O.C. Wallmo, and C. Feddema. 1973. Foods of the Rocky Mountain mule deer. USDA For. Serv. Res. PaperRM-l1 1, FortCollins,Colo. Lesperance, A.L., T.T. Tueller, and V.R. Bohman. 1970. Competitive use of the range forageresource. J. Anim.Sci. 30:115-121. Lewis, C.E. 1980. Simulatedcattle injury to plantedslash pine: combinations of defoliation, browsing and trampling. J. Range Manage.35:340-345. McConnel, B.R. and J.G. Smith. 1977. Influenceof grazingon age-yieldinteractions in bitterbrush. J. RangeManage.20:91-93. Milliken, G.A. and Johnson, D.E. 1992. Analysis of messy data: designed experiments.Chapman andHall,New York. Mueggler, W.R. and J.P. Blaisdell. 1958. Effects on associated species of burning, rotobeating,sprayingand railing sagebrush. J. RangeManage.11:61-66. National Oceanic and Atmospheric Administration (NOAA). 1990-2001. Climatological data annual summary, Oregon.95-106:(13).

JOURNAL OF RANGE MANAGEMENT57(2) March 2004

167

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions

of GreatBasin Stuth, J.W. and A.H. Winward. 1977. Vavra, M. and F. Sneva. 1978. Seasonaldiets Neal, D.L. 1981. Improvement deer winterrange with livestock grazing,p. of five ungulates grazing the cold desert Livestock-deer relations in the lodgepole relation61-73. In: Proc. Wildlife-livestock biome. In: D.N. Hyder (ed.), Proc. 1" Int. pine-pumice region of central Oregon. J. ships symposium.Forest,Wildlife & Range Rangeland Congress Soc. Range Manage. RangeManage.30:110-116. Exper.Sta.,Univ. Idaho,Moscow,Ida. Denver,Colo. Tueller, P.T. and J.D. Tower. 1979. Reiner, RJ. and PJ. Urness. 1982. Effect of Vegetation stagnation in three-phase big Winward, A.H. and J. Alderfer-Findley. of grazing horses managedas manipulators 1983. Taxonomic variationsof bitterbrush game exclosures. J. Range Manage. big game winter range. J. Range. Manage. in Oregon,p. 25-30. In: (Purshia tridentata) 32:258-263. 35:567-571. A.R. and K.L. Johnson(comps.) Tiedemann, Urness, PJ. 1981. Livestockas tools for manSnedecor, G.W. and W.C. Cochran. 1967. Proc. Research and managementof bitteraging big game winter range in the interStatisticalmethods.Iowa State Univ. Press. brush and cliffrose in western North mountain west, p. 20-31. In: Proc.WildlifeAmes, Iowa. America. USDA-For. Serv., Intermountain Forest, relationships symposium. livestock and Sneva, F.A. 1982. Relationof precipitation Forestand RangeExp. Sta. Gen. Tech. Rep. Wildlife & RangeExp. Sta., Univ. of Idaho, with yield of herbaceousplants temperature INT-152.Ogden,Ut. Moscow,Ida. in eastern Oregon. Int. J. Biometeor. 26:263-276.

168

JOURNAL OF RANGE MANAGEMENT57(2) March 2004

This content downloaded from 181.15.183.213 on Tue, 12 Nov 2013 06:59:43 AM All use subject to JSTOR Terms and Conditions