Herbaceous Succession After Burning of Cut Western Juniper Trees

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers,academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.

Herbaceous Succession After Burning of Cut Western Juniper TreesAuthor(s): Jon D. Bates and Tony J. SvejcarSource: Western North American Naturalist, 69(1):9-25. 2009.Published By: Monte L. Bean Life Science Museum, Brigham Young UniversityDOI: http://dx.doi.org/10.3398/064.069.0120URL: http://www.bioone.org/doi/full/10.3398/064.069.0120

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological,ecological, and environmental sciences. BioOne provides a sustainable online platform for over170 journals and books published by nonprofit societies, associations, museums, institutions,and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicatesyour acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use.Commercial inquiries or rights and permissions requests should be directed to the individualpublisher as copyright holder.

http://dx.doi.org/10.3398/064.069.0120

http://www.bioone.org/doi/full/10.3398/064.069.0120

http://www.bioone.org

http://www.bioone.org/page/terms_of_use

The expansion of western juniper ( Junipe-rus occidentalis spp. occidentalis Hook.) in thenorthern Great Basin has resulted in land-scape-level conversions of sagebrush-steppe,riparian, and aspen communities to juniper-dominated woodlands (Miller et al. 1999, 2000,Wall et al. 2001). Western juniper woodlandshave increased nearly 90% since settlement ofthe region began in the 1860s (Miller et al.2005). Woodlands now occupy 3.5 million hain eastern Oregon, northeastern California,northern Nevada, and southwestern Idaho.The main cause of the expansion has beenattributed to reductions in fire disturbances asa consequence of historic grazing and fire sup -pression (Burkhardt and Tisdale 1969, Millerand Rose 1995). Juniper invasion reducesshrub steppe productivity (Vaitkus and Eddle -man 1987, Bates et al. 2000, Miller et al.2000), alters hydrologic and nutrient cycles(Buckhouse and Mattison 1980, Josiatis 1990,Bates et al. 2000, 2002, Roberts and Jones2000, Pierson et al. 2007), and ultimatelycauses decreases in wildlife habitat (Miller et

al. 2005, Noson et al. 2006). Efforts to controlwestern juniper and restore sagebrush-steppe,riparian, and aspen communities have becomea major focus of ecosystem management in theregion.

Prescribed fire, mechanical cutting, and acombination of these treatments have been themain methods used to control western juniper(Miller et al. 2005). Fire remains a viable man-agement option for western juniper removal inwoodlands that are in early (Phase I) to mid-successional (Phase II) stages, when sufficientand continuous surface (0–1 m) fuels are pre-sent (Miller et al. 2005). In late successional(Phase III) woodlands, surface fuels are typi-cally not adequate to sustain fire and removetrees; therefore, these woodlands have mainlybeen cut with chainsaws (Miller et al. 2005).Cutting treatments have commonly prescribedleaving cut western juniper on site. Evidencesuggests that retaining cut western juniper orother pinyon-juniper species stabilizes sitehydrology by reducing runoff and erosion,encouraging the establishment of perennial

Western North American Naturalist 69(1), © 2009, pp. 9–25

HERBACEOUS SUCCESSION AFTER BURNING OFCUT WESTERN JUNIPER TREES

Jon D. Bates1,2 and Tony J. Svejcar1

ABSTRACT.—The expansion of western juniper ( Juniperus occidentalis spp. occidentalis Hook.) in the northern GreatBasin has resulted in the wide-scale conversion of sagebrush-steppe communities to juniper woodlands. Prescribed fireand mechanical cutting are the 2 main methods used to remove juniper and restore sagebrush steppe. Mechanical treat-ments commonly leave cut juniper on site. Disadvantages of leaving cut juniper are the increased fuel hazard and thepotential for increased establishment and growth of invasive species. This study evaluated the response of herbaceousplants to winter burning of cut western juniper. Vegetation response was compared among 2 burning treatments (burningtrees the first winter after cutting and burning the second winter after cutting), a control (cut-unburned juniper), andthe interspace between cut trees. To minimize fire impacts to herbaceous perennials, cut trees were burned in the win-ter when soils and ground litter were frozen and/or soils were at field capacity. Only felled trees were burned, as fire didnot carry into interspaces or litter mats around western juniper stumps. We hypothesized that winter season burningwould increase herbaceous perennials and would reduce cheatgrass establishment when compared to the cut-unburnedcontrol. After 10 years, total herbaceous and perennial grass cover was 1.5- to 2-fold greater, respectively, in burnedtreatments compared to cut-unburned controls. Perennial grass density was 60% greater in the burned treatments thanin the cut-unburned treatment and the interspace. Cheatgrass cover was twice as great in the control than in the 2 burntreatments and the interspace. We concluded that burning cut western juniper when soils were wet and frozen in winterenhanced community recovery of native perennials compared to leaving cut juniper unburned.

Key words: Bromus tectorum, annual grass, Juniperus occidentalis, litter, fire, Thurber’s needlegrass, squirreltail.

1USDA–Agricultural Research Service and Oregon Agricultural Experiment Station, Eastern Oregon Agricultural Research Center, 67826-A Hwy. 205,Burns, OR 97720.

2E-mail: [email protected]

9

grasses and retaining site nutrient capital( Jacobs and Gatewood 1999, Brockway et al.2002, Pierson et al. 2007). A major disadvantageto leaving cut western juniper in place is theincreased fuel hazard, particularly during thefirst 2–3 years posttreatment when dried leavesremain on downed trees. Another disadvantageto retaining cut trees and debris after mechan-ically treating western juniper is that theseareas create ideal microsites for invasion byundesirable nonnative annual grasses and mayslow recovery of desired native species (Younget al. 1985, Bates et al. 2005, 2007a,). Recently,efforts have shifted to removing cut westernjuniper by burning to reduce potential firehazards.

In 1997 we established a study to evaluatethe response of herbaceous vegetation to burn-ing cut western juniper. Herbaceous dynamicswere compared among 2 winter burning treat -ments (burning trees the first winter after cut-ting and burning the second winter after cut-ting), a control (cut and unburned juniper),and the interspace between cut trees. Burningwas done in the winter or early spring whensoils and ground litter were frozen and/orthoroughly wetted to minimize the impacts offire to perennial herbaceous vegetation. Winterburning was selected over fall burning becauseburning in the fall may produce more severeimpacts to native vegetation. Fall burning offelled juniper produces temperatures suffi-cient to kill most herbaceous species (Gifford1981), especially perennial grasses, and canincrease abundance of exotic species (Haskinsand Gehring 2004). Fall burning of cut westernjuniper in quaking aspen (Populus tremuloidesMichx.) woodlands when soils and ground litterwere dry severely impacted herbaceous vege-tation resulting in the complete removal ofperennial bunchgrasses and a high percentageof perennial forbs (Bates et al. 2006). Earlyspring burning of cut western juniper in aspenwoodlands when soils and ground litters weresaturated and frozen caused no mortality ofperennial herbaceous vegetation (Bates et al.2006). Thus, we hypothesized that burning cutwestern juniper in the winter would cause lit-tle measurable mortality to existing nativeherbaceous perennials, resulting in faster andgreater recovery of these species, and wouldreduce the potential for introduced annualgrasses to establish and increase when com-pared to cut-unburned controls.

METHODS

Study Site

The study site was on Steens Mountain,southeastern Oregon (118°36�E, 42°55�N). Ele -vation at the site is 1575 m and aspect is westfacing (11% slope). The majority of annualprecipitation falls between November and lateMay. Annual precipitation (1 October–30 Sep-tember) at Malheur National Wildlife Refugeweather station, located 30 km northwest(1250 m) of the study site, has averaged 254mm over the past 64 years (Fig. 1). Soils aremainly Typic Vitrixerand with inclusions of aTypic Calcixeroll that are underlain by awelded volcanic ash tuff, which restricts rootpenetration at 40 cm.

The site was dominated by 90-year-oldwestern juniper woodland. Juniper had fullyoccupied this site as indicated by limited lateraland terminal leader growth, crown lift, and lackof further juniper recruitment (Bates et al.2000). Juniper canopy cover averaged 30.5%,and tree density averaged 283 trees ⋅ ha–1.Sagebrush was eliminated from the site byjuniper interference, though previous shruboccupancy was evident by the presence ofnumerous shrub skeletons. The dominant shrubprior to juniper encroachment was basin bigsagebrush (Artemisia tridentata ssp. tridentata[Beetle & Young] Welsh). Herbaceous coveraveraged 5.5%. Bare ground and rock in theinterspace was about 95%. Understory compo-sition was a mix of native grasses and nativeand nonnative forbs. Sandberg’s bluegrass (Poasecunda J. Presl.) and pale alyssum (Alyssumalyssoides L.) were the most common grassand forb species. Other characteristic specieswere bottlebrush squirreltail (Elymus ely-moides [Raf.] Swezey), bluebunch wheatgrass(Pseudo roegeneria spicata [Pursh] A. Löve),Thurber’s needlegrass (Achnatherum thurberi-anum [Piper] Barkworth), and basalt milkvetch(Astragalus filipes Torr.). Cheatgrass (Bromustectorum L.) was present (<0.1% cover) andprimarily grew beneath juniper canopies.

This particular site has been used to evalu-ate long-term vegetation dynamics after me -chanical cutting of western juniper in severalearlier studies (Bates et al. 2005, 2007a).These past studies demonstrated that cuttingof western juniper was effective at increasingunderstory biomass, density, cover, and diver-sity (Bates et al. 2005), and that succession

10 WESTERN NORTH AMERICAN NATURALIST [Volume 69

dynamics were influenced by microsite (Bateset al. 2007a). For example, under cut juniper,cover and density of herbaceous species asso-ciated with interspaces were reduced andspecies characteristic of canopy locations in -creased (Bates et al. 2007a). These studies alsoreported that significant cheatgrass responsewas delayed until 5–6 years after tree cuttingand was confined to areas of litter accumula-tion beneath cut trees and in old canopy littermats. However, by the 12th year after tree cut-ting native perennial bunchgrasses were dom-inant and cheatgrass was only a minor compo-nent of the study sites.

Experimental Design

The experimental design was a randomizedblock with 5 blocks and 3 treatments (Peterson1985). Blocks were 1.5 ha in size (treatmentswithin each block occupied 0.5 ha). Baselinevegetation characteristics were measured inJuly 1997. In September 1997, all trees werecut using chainsaws and left in place. Post-treatment vegetation was measured in mid-May 1998–2001 and 2006. Livestock wereexcluded during the study.

Cut juniper treatments were the following:a control (cut-unburned juniper), cut treesburned the first winter after cutting (1st-yearburn), and cut trees burned the second winterafter cutting (2nd-year burn; Fig. 2). The areasinfluenced by cut trees are equivalent to debris

locations described by Bates et al. (2007a). Cuttree (debris) locations were former interspaceareas that were covered by trees after cuttingand were identical to the interspace in speciescomposition and canopy cover prior to cutting.All treatments’ measurements were collectedon former interspace zones that were coveredwith felled juniper trees. Interspaces not cov-ered by trees were measured across each blockin cut-unburned and cut-burned treatments.

The 1st-year burn treatment was applied on11 March 1998. The 2nd-year burn treatmentwas applied on 20 February 1999. Gravimetricsoil water (0–10 cm) and fuel (herbaceous finefuels and ground litter) moisture were mea-sured on the day of fire application. Fuel mois-ture and soil water were determined by dryingfuel and soil, separately, at 100 °C to a constantweight. Weather data (humidity, wind speed,and temperature) were recorded on the day offire applications. Burn conditions are providedin Table 1. The criteria for burning were that(1) soils and surface litter under cut trees werewet (at field capacity) and preferably frozen,and (2) suspended juniper leaf litter was dry(<20% water content). Downed juniper cov-ered between 25% and 35% of the area, socare had to be taken in selecting days whenthere was minimal wind to reduce the possi-bility of fire spreading into adjacent treatments.Burning was done using drip torches containinga 60:40 mixture of diesel and unleaded gas. In

2009] SUCCESSION AFTER BURNING CUT JUNIPER 11

Year

1996-97 1997-98 1998-99 1999- 2000-01 2001-02 2002-03 2003-04 2004-05 2005-06

Pre

cip

itat

ion (

mm

)

0

100

200

300

400

500

*

*

2000

64 year average (254 mm)

Fig. 1. Annual precipitation (mm) for years 1997–2006 and the long-term average at the Malheur National WildlifeRefuge weather station (1250 m), located 29 km northwest of the study site on Steens Mountain, Oregon.

the burn treatments, fire was confined to areasoccupied by the cut trees and did not burninto interspaces or around the litter mats underformer juniper canopies. Litter moisture of theolder litter mats of former juniper canopies wastoo high for fires to burn and consume the mats.

Fire severity was estimated using an indexof juniper consumption and percentage ofperennial grasses killed. The severity categorieswere light (1%–29% of surface litter consumed,<20% of the perennial grasses killed, and<100% of suspended 1-hour fuels [leaves andtwigs, <6.3 mm diameter] consumed), moder-ate (30%–79% of surface litter consumed,<20%–70% perennial grasses killed, <100%of suspended debris up to 10-hour fuels [<25mm diameter] consumed), and high (>80% of

surface litter consumed, >70% perennialgrasses killed, only trunks and branches [>25mm diameter] of downed juniper remaining).

Understory Sampling

Herbaceous canopy cover and density byspecies were measured using 0.2-m2 frames(0.4 × 0.5 m) in 1997–2001 and 2006. Groundcover was visually estimated for the followingcategories: herbaceous species, bare ground,rock, litter ( juniper and herbaceous), and bioticcrust. Density of herbaceous perennials wasmeasured by counting all individuals withinthe frames. Density of the cheatgrass wasmeasured by counting individuals in a 0.05-m2

nested plot (0.1 × 0.5 m) within each frame.For the cut and cut-and-burn treatments,


Fig. 2. Examples of unburned and burned cut western juniper treatments on Steens Mountain, Oregon. Also shownare location delineations for the cut juniper treatments (debris, interspace, and canopy). The debris locations are formerinterspaces that were covered by trees after cutting. The debris location was the only area that burned in the cut-and-burn treatments. Interspaces are open areas between canopy and debris locations. The canopy location is the litter matbeneath precut western junipers, and herbaceous vegetation in the canopy areas was not measured in this study.

TABLE 1. Weather and water content conditions for 1st-year and 2nd-year burns in cut juniper woodlands, SteensMountain, Oregon. Water content and soil temperatures were measured under cut trees prior to burning.

11 March 1998 burn 20 February 1999 burn

Weather and soil variablesAir temperature (°C) 8–16 3–7Relative humidity (%) 41–55 36–72Wind speed (kph) 10–13 6–16Soil temperature (5 cm, °C) 4.4 –3.2

Water content (%)Soil (0–10 cm) 35 31Surface juniper litter (0–5 cm)a 107 109Suspended juniper needles (>5 cm) 7 8

aWater content of surface juniper litter exceeded 100%, probably because of accumulations of snow or frost in samples.

UnburnedDebris

UnburnedDebris

BurnedDebris

BurnedDebris

Canopy

InterspaceInterspace

Canopy

herbaceous plants were measured under 10randomly selected trees in each treatmentreplicate (Fig. 2). Under the trees, 4 randomlylocated frames were measured for herbaceouscover and density (40 per treatment replicate).Interspace measurements were randomlylocated in remaining open areas between cuttrees (40 frames per treatment replicate).Herbaceous vegetation was initially measuredin the litter mats or canopy locations beneaththe precut western juniper tree in 1997. Insubsequent years we discontinued measure-ments in this area because the fires did notburn in the old canopy mounds. The rationalefor not continuing measurements in the canopyareas was that herbaceous dynamics wouldlikely be similar to what was reported forcanopy locations by Bates et al. (2007a). Bateset al. (2007a) reported that cheatgrass andperennial grasses codominated canopy areas 6years after cutting; however, by the 12th year ofthe study perennial grasses dominated canopyareas as cheatgrass declined in cover and bio-mass. In the current study, observations in 2007

indicated that perennial grasses and annualgrasses were codominating in canopy locationsin the burned and unburned treatments.

Statistical Analysis and Data Organization

Analysis of variance was used to test fortreatment effects on herbaceous cover (speciesand functional group) and density (species andfunctional group). Functional groups were com-posed of Sandberg’s bluegrass, deep-rootedperennial bunchgrasses (e.g., Thurber’s needle-grass, bluebunch wheatgrass, and bottlebrushsquirreltail), cheatgrass, perennial forbs, andannual forbs. Cover and density response vari-ables were analyzed using a repeated-measureANOVA for a randomized complete blockmodel (SAS Institute, Inc. 2002). The modelincluded block (5 blocks, df = 4), year (1997–2001 and 2006, df = 5), treatment (cut-un -burned, interspace, 1st-year burn, 2nd-yearburn; df = 3), and year × treatment interaction(df = 15, error term df = 92). Because of astrong year effect, years were also analyzed


TABLE 2. P-values for herbaceous cover and density comparing burned and unburned treatments and interspace incut juniper woodlands, Steens Mountain, Oregon, 1997–2001 and 2006. Asterisks (*) indicate significant main effect andyear × treatment interactions at P < 0.05.

Cover Density_____________________________________ ____________________________________

Response variable Year Treatment Year × Treatment Year Treatment Year × Treatment

GROUND COVER

Total herbaceous 0.0001* 0.0001* 0.0930* — — —Bare ground/rock 0.0001* 0.0001* 0.0001* — — —Juniper litter 0.0001* 0.0001* 0.0001* — — —Herbaceous litter 0.0001* 0.0001* 0.0001* — — —Moss/crust 0.5402 0.3629 0.4284 — — —

FUNCTIONAL GROUP

Sandberg’s bluegrass 0.0001* 0.0001* 0.0330* 0.0001* 0.0001* 0.7813Perennial grass 0.0001* 0.0001* 0.0833 0.0001* 0.0055* 0.0001Cheatgrass 0.0001* 0.0074* 0.0007* 0.0001* 0.0163* 0.0077*Perennial forb 0.0391* 0.0010* 0.0073* 0.7274 0.0018* 0.3546Annual forb 0.0001* 0.0001* 0.0063* — — —

SPECIES

Bluebunch wheatgrass 0.0001* 0.0001* 0.1550 0.0001* 0.0393* 0.9631Squirreltail 0.0001* 0.0252* 0.2203 0.0001* 0.0009* 0.0032*Thurber’s needlegrass 0.0001* 0.0001* 0.2440 0.0001* 0.0490* 0.0069*Pale agoseris 0.0129* 0.0858 0.2920 0.0004* 0.1773 0.1030Basalt milkvetch 0.0027* 0.3389 0.1095 0.0741 0.1424 0.1084Mariposa lily 0.2924 0.4217 0.2603 0.0012* 0.1828 0.0804Western hawksbeard 0.0001* 0.0058* 0.3850 0.0037* 0.5097 0.0841Donnell’s lomatium 0.0340* 0.0016* 0.3555 0.0212* 0.0007* 0.2149Tailcup lupine 0.1617 0.0067* 0.2947 0.4879 0.0391* 0.5484Other perennial forbs 0.6133 0.0028* 0.7426 0.0160* 0.6982 0.7942Pale alyssum 0.0001* 0.0006* 0.0008* — — —Epilobium 0.0001* 0.0005* 0.0147* — — —Prickly lettuce 0.0001* 0.0001* 0.0001* — — —Microseris 0.0001* 0.0001* 0.0012* — — —Line-leaf phacelia 0.0001* 0.0001* 0.0001* — — —

separately to simplify presentation of the resultsand to assist in explaining interactions. Datawere tested for normality using the SAS uni-variate procedure. Data not normally distrib-uted were arcsine square-root transformed tostabilize variance. Back-transformed meansare reported. Statistical significance was set atP < 0.05 and means were separated usingFisher’s protected LSD.

RESULTS

Ground Cover

Values for ground cover response variablesindicated differences among treatments acrossthe sampling period (Table 2). Herbaceouscover increased in all treatments between1997 and 2006 (Fig. 3A). However, in 2006herbaceous cover was less in the cut-unburnedtreatment than in the other treatments. Juniperlitter was less in the interspace than in all thecut (burned and unburned) juniper treatments(Fig. 3B). Under cut juniper (burned andunburned), juniper litter initially increasedand then declined over time. Juniper litter wasgreatest in the cut-unburned treatment fol-lowed in order by the 2nd-year burn and 1st-year burn treatments. Herbaceous litter in -creased in all the treatments over time (Fig.3C). By 2006, herbaceous litter was greatest inthe cut-unburned and 2nd-year burn treatmentsfollowed in order by the 1st-year burn treat-ments and interspace. In 2006, we observedthat herbaceous litter in the cut-unburnedtreatment was mainly composed of the previ-ous year’s growth of cheatgrass. Herbaceouslitter in the other treatments was mainly com-posed of earlier years’ growth of perennialgrasses. Because tree cutting increased juniperlitter, all the cutting treatments had less bareground than the interspace (Fig. 3D). Bareground declined over time in the interspace aslitter and herbaceous cover increased. Bareground in the cut-unburned treatment was lessthan both cut-and-burn treatments.

Functional Group Cover and Density

Functional group measurements were usefulin detecting plant compositional changes amongthe treatments. Year × treatment interactionswere significant for density and cover of mostfunctional-group response variables (Table 2).Perennial grass cover increased in all treatmentsover time. However, after the 3rd year post-

cutting (2000), perennial grass cover was lessin the cut-unburned treatment than in bothcut-and-burn treatments and the interspace(Fig. 4A). By 2006, perennial grass cover wasnearly twice as great in the 1st-year and 2nd-year burn treatments and the interspace com-pared to the cut-unburned treatment. Sand-berg’s bluegrass cover in the cut-unburnedtreatment declined over time and tended to beless there than in one or more of the othertreatments in most years (Fig. 4B). In 2006,Sandberg’s bluegrass cover was nearly 5 timesgreater in the interspace and 1st-year burntreatments than in the cut-unburned treat-ment. Perennial forb cover in the cut-unburned treatment was less than it was inone or more of the cut-and-burn treatmentsbetween 1998 and 2000 (Fig. 4C). However, inno treatment did perennial forb cover increaseabove pretreatment levels. In 2006, treat-ments did not differ in perennial forb cover.Cheatgrass cover in creased in all treatmentsby 2000 or 2001 (Fig. 4D). Cheatgrassincreased similarly among the cut-unburnedand both cut-and-burn treatments in the first4 growing seasons (1998–2001). However, in2006, cheatgrass cover was nearly 3 timeshigher in the cut-unburned treatment than itwas in the other treatments. Annual forb covervaried among years, and until 2006 no treat-ment differences were measured (Fig. 4E). In2006, annual forb cover was 4.5 and 6 timesgreater in the interspace and 1st-year burntreatment compared to the cut-unburnedtreatment.

Functional-group density response was sim-ilar to treatment relationships, and it trends asdescribed above for functional-group cover.By 2006, Sandberg’s bluegrass density haddeclined in all treatments except the 1st-yearburn (Fig. 5A). Sandberg’s bluegrass density in2000, 2001, and 2006 was less in the cut-unburned treatment than in the interspaceand both cut-and-burn treatments. Perennialgrass density increased in all treatments bythe end of the study (Fig. 5B). By 2006, peren-nial grass density was greater in both cut-and-burn treatments than in the cut-unburned treatment and the interspace.Cheatgrass density increased in 2000 and2001 in all treatments (Fig. 5C). By 2006,cheatgrass density was twice as high in the cut-unburned treatment than in the interspaceand both cut-and-burn treatments. Perennial



A. Herbaceous Canopy CoverC

over

(%

)

10

20

30

40

50

Cut-Unburned

First-Year Burn

Second-Year Burn

Interspace

C. Herbaceous Litter

Cover

(%

)

0

10

20

30

40

B. Juniper Litter

Cover

(%

)

0

20

40

60

80

D. Bareground and Rock

Year

1997 1998 1999 2000 2001 2006

Cover

(%

)

0

20

40

60

80

a a

b, b, b

a a a a

b, b, b

b

b

b b

c

c

c

c, c

d d

d

aa

aa, ab

b

b

b

ab

c

cc

d

d

d

a

a

b

ab

b

b

b

ab

a

bb

c

a

bbb

b

a

c

d

a

bb

c

a

ab

a

Fig. 3. Ground cover (%) response variables for the debris treatments (1st-year burn, 2nd-year burn, cut-unburned)and interspace on Steens Mountain, Oregon: A, herbaceous cover; B, herbaceous litter; C, juniper litter; D, bare ground.Data are means with bars representing one standard error. Significant differences (P < 0.05) among the treatments areindicated by different lowercase letters.

D. Bare ground and Rock


A. Perennial Grass Cover

Can

opy C

over

(%

)

0

5

10

15

20

25

Cut-Unburned

First-Year Burn

Second-Year Burn

Interspace b

bb

a

b

B. Sandberg's bluegrass

Can

op

y C

ov

er (

%)

0

2

4

6

C. Perennial Forbs

Can

op

y C

ov

er (

%)

0

1

2

3

D. Cheatgrass

Can

opy C

over

(%

)

0

10

20

30

a, a

ab

a, a

b

b

ab, ab

b

b

bc

a

c

a

b, b, b

b

ab, ab

a

b

b

b

bab

a

b

a

abab

a

b

E. Annual Forbs

Year

1997 1998 1999 2000 2001 2006

Can

opy C

over

(%

)

0

2

4

6

a

b

b

b

a

a

ab

b

b

a, a, a

b

b

b

ab

a

Fig. 4. Functional group cover (%) for the debris treatments (1st-year burn, 2nd-year burn, cut-unburned) and inter-space on Steens Mountain, Oregon: A, perennial bunchgrasses; B, Sandberg’s bluegrass; C, perennial forbs; D, cheat-grass; E, annual forbs. Data are presented as means with bars representing one standard error. Significant differences (P< 0.05) among treatments for the response variables are indicated by different lowercase letters.


B. Perennial Grass Density

pla

nts

m-2

5

10

15 Cut-Unburned

First-Year Burn

Second-Year Burn

Interspace

a, a, a

b

ba

a

a

A. Sandberg's bluegrasspla

nts

m-2

2

4

6

8

10

12

C. Cheatgrass

pla

nts

m-2

0

200

400

600

a

b

b

abbc

a

c

b

b

a

b

a

ab

ab

b

D . Perennial Forbs

Year

1997 1998 1999 2000 2001 2006

pla

nts

m-2

0

1

2

3

4

5

b, b, b

a a, a

c

b

b

b

ab

a

ab

b

b

a

b

b

b

aa

b

b

a, a, a

b

a

ab

b

ab

b

Fig. 5. Functional group densities (plants ⋅ m–2) for the debris treatments (1st-year burn, 2nd-year burn, cut-unburned) and interspace on Steens Mountain, Oregon: A, Sandberg’s bluegrass; B, perennial grasses; C, cheatgrass; D,perennial forbs. Data are presented as means with bars representing one standard error. Significant differences (P <0.05) among treatments for the response variables are indicated by different lowercase letters.

pla

nts

⋅m

–2

pla

nts

⋅m

–2

pla

nts

⋅m

–2

pla

nts

⋅m

–2

forb densities were variable across time, andno consistent pattern emerged among thetreatments (Fig. 5D).

Species Response

Species composition was influenced bytreatment. Year × treatment interactions weresignificant for densities of bottlebrush squir-reltail, bluebunch wheatgrass, and Thurber’sneedlegrass; and they were also significant forcover of several annual forb species (Table 2).Other species only displayed significant yearand/or treatment effects (cover and/or density).

Squirreltail densities increased in the cut-unburned, 1st-year burn, and 2nd-year burntreatments such that density values in thosetreatments were all greater than density in theinterspace by 2006 (Fig. 6A). At the end of the

study, density of Thurber’s needlegrass wasgreater in the 1st-year burn than it was in allother treatments (Fig. 6B). The interspace hadgreater density of Thurber’s needlegrass thanthe cut-unburned treatment. Bluebunch wheat-grass density increased in all treatments overtime (Table 2). By 2006, bluebunch wheatgrassdensity was greater in the 1st-year burn thanin the cut-unburned treatment. Cover ofThurber’s needlegrass and bluebunch wheat-grass began to differ among treatments by thesecond and third year after cutting, respec-tively (Figs. 7A, 7B). In 2006, Thurber’s needle -grass cover was greater in the interspace and1st-year burn treatment than in the cut-unburned and 2nd-year burn treatment. Blue-bunch wheatgrass cover was lower in thecut-unburned treatment than in all other


Pla

nts

m-2

0

2

4

6

8

Cut-Unburned

First-Year Burn

Second-Year Burn

Interspace

A. Bottlebrush Squirreltail

a

b

bc

c

a

b

b

b

Year

1997 1998 1999 2000 2001 2006

Pla

nts

m-2

0

2

4

6

8

B. Thurbers needlegrass

c

b

a

ab

Fig. 6. Densities (plants ⋅ m–2) of bottlebrush squirreltail (A) and Thurber’s needlegrass (B) for the debris treatments(1st-year burn, 2nd-year burn, cut-unburned) and interspace on Steens Mountain, Oregon. Data are means with barsrepresenting one standard error. Significant differences (P < 0.05) among treatments for each response variable are indi-cated by different lowercase letters.j

pla

nts

⋅m

–2

pla

nts

⋅m

–2

B. Thurber�s needlegrass

treatments by the fourth growing season aftercutting (2001).

Cover of basalt milkvetch (Fig. 7C) andwestern hawksbeard (Crepis occidentalis Nutt.)(Fig. 7D) are illustrative of cover and densitydynamics of many other perennial forb species(Table 2). Treatment differences for perennialforb species developed mainly within the first

3–4 years after cutting. Cover of basalt milk -vetch and hawksbeard were less in the cut-unburned treatment than they were in one ormore of the other treatments between 1998and 2001. By 2006, no differences in coverfor these species were found among thetreatments. Other perennial forb speciesdeveloping similar patterns for cover and/or


Can

opy C

over

(%

)

0

2

4

6

8

10

12Cut-Unburned

First-Year Burn

Second-Year Burn

Interspace

A. Thurber's needlegrass

a

b

aab

bb

a

Can

opy C

over

(%

)

0

1

2

C. Basalt milkvetch

b

a

b

ab

a, a

b

b

ab

Can

opy C

over

(%

)

0

2

4

6

8

10

12

aa

a

a

ab

B. Bluebunch wheatgrass

1997 1998 1999 2000 2001 2006

Can

opy C

over

(%

)

0

1

2

D. Western Hawksbeard

b

a, a

b

ab

a

abab

b

b

ab

a, aa, a

bb, b

aaab

b

c

a

ab

b

c

ab

b

b

Year

Fig. 7. Cover of Thurber’s needlegrass (A), bluebunch wheatgrass (B), basalt milkvetch (C), and western hawksbeard(D) for the debris treatments (1st-year burn, 2nd-year burn, cut-unburned) and interspace on Steens Mountain, Oregon.Data are means with bars representing one standard error. Significant differences (P < 0.05) among treatments for eachresponse variable are indicated by different lowercase letters.

density were pale agoseris (Agoseris glauca[Pursh] Raf.), Donnell’s lomatium (Lomatiumdonnelli Coult. & Rose), tailcup lupine (Lupi-nus caudatus Kell.), and mariposa lily (Calo-chortus macrocarpus Dougl).

Annual forb species developed specificresponses to the treatments (Table 2). Cover ofpale alyssum, a nonnative species, was lowerin the cut-unburned treatment compared tothe other treatments, particularly in 2006 (Fig.8A). Epilobium (Epilobium paniculatum Nutt.)and Microseris (Microseris gracilis Hook.) hadgreater cover in the interspace in the final 4years of measurement compared to the cut-unburned and cut-and-burn treatments (Table2). Treatment differences of other annual forbswere transient and were mainly detectedwithin the first 4 years post-cutting. Cover ofprickly lettuce (Latuca serriola L.; Fig. 8B)and tansy-mustard (Descurainia pinnata[Walt.] Britt.; Table 2) were greater in thecut-unburned treatment than in the othertreatments in the third year after cutting(2000). Two forbs, line-leaf phacelia (Phacelialinearis [Pursh] Holz; Fig 8C) and sinuate gilia(Gilia intermedia Dougl.; Table 2) had greatercover in the 1st-year burn treatment than inthe other treatments in 1999 and 2000.

Burn Severity

Weather conditions were different betweenthe 2 burn treatments. Air and soil tempera-tures tended to be lower and humidity andwind speed greater and more variable in the2nd-year burn compared to the 1st-year burn(Table 1). Soil and litter water content, how-ever, did not differ appreciably between the 2burns, though soils and ground litter in the2nd-year burn were frozen. By our criteria,burning in winter was effective at removingaboveground light fuels ( juniper leaves up to1-hour fuels); however, heavier fuels (>10-hour fuels) and much of the juniper leaf littercontacting the ground were not consumed byfire. Winter burning reduced but did not elim-inate western juniper fuels. As a result it wasdifficult to apply a severity rating to eitherburn treatment. Consumption of surface litterwas light in both burn treatments. In the 1st-year burn treatment, cover of juniper litterwas reduced by 30.3% +– 4.7% after fire and inthe 2nd-year burn, litter cover was reduced by12% +– 2.3% after fire (Fig. 3B). Abovegroundconsumption of cut juniper removed all mate-

rial up to the 10-hour fuels category, whichwould indicate a moderate severity rating. Thefire severity index did not prove useful forpredicting mortality of perennial bunchgrasses.Neither burn treatment resulted in a declinein Sandberg’s bluegrass density after fire (Fig.5A). Perennial grass densities decreased by41% in the 1st-year burn between 1997 and1998 (Fig. 5B). Yet, over the same time period,perennial grass densities in the cut-unburnedand 2nd-year burn (not yet burned) treatmentsdeclined by greater than 50%. In the 2nd-yearburn, perennial grass densities increased 2-fold between 1998 and 1999 (first growingseason after fire).

DISCUSSION

Effects of Burning

The burning of cut western juniper trees inthe winter had a long-term, positive effect onthe recovery of native herbaceous perennialswhen compared to the effect of leaving cuttrees unburned and in place. Though differ-ences among the 3 cutting (unburned andburned) treatments were detected in early suc-cession (1998–2001), the results highlight theimportance of evaluating disturbance impactsover long time periods. Differences among thetreatments for perennial bunchgrasses, Sand-berg’s bluegrass, and cheatgrass response vari-ables, as well as under story composition,became much more obvious by 2006, tenyears after treatments were applied. Herba-ceous cover, perennial grass cover, and peren-nial grass densities were 1.5–2 times greater inthe 1st-year and 2nd-year burn treatments thanin the cut-unburned treatment. Burning cuttrees did not limit the establishment of cheat-grass in early succession (1999– 2001) but didlimit the duration of high cheatgrass cover anddensity. The increases in peren nial grass den-sity and cover in the burned treatments sug-gest that perennial suppression of cheatgrasswas a primary factor for cheatgrass decline by2006. In contrast, the cut-unburned treat-ment retained high levels of cheatgrassthrough the end of the study. Thus, weaccepted our hypotheses that winter burningof cut western juniper would result in morerapid and larger increases in perennial herba-ceous cover/density and lower cheatgrasscover/density compared to the results of leav-ing cut trees unburned.


Our results contrast with herbaceous re -sponse reported by Haskins and Gehring (2004)following burning of pinyon-juniper slash inArizona. They measured 4-fold higher abun-dance of invasive annuals in burned slash areascompared to unburned slash and suggested thatslash burning may be detrimental to nativeperennial plant recovery in treated pinyon-juniper woodlands. Haskins and Gehring (2004)did not provide burning conditions or time ofburn in their article. The timing or conditions

under which cut trees or slash/debris areburned appears to be an important determinantof post-fire herbaceous response. For example,burning of cut western juniper and trees inquaking aspen stands in the fall (dry soils andfuels) severely impacted the understory, shift-ing understory composition from dominanceby herbaceous perennials (largely eliminatedby fire) to native and nonnative invasive annu-als and biennials (Bates et al. 2006). In aspenstands where cut juniper was burned in spring


A. Pale alyssum

Cover

(%

)

0

2

4

6

8

Cut-Unburned

First-Year Burn

Second-Year Burn

Interspace

B. Prickly lettuce

Cover

(%

)

0

1

2

3

4

C. Line-leaf phacelia

Year

Cover

(%

)

0

1

a

b

ab

a

a

a

a

b

b

a, a, a a, a, a

b

a

ab

b

b

1997 1998 1999 2000 20062001

Fig. 8. Cover of pale alyssum (A), prickly lettuce (B), and line-leaf phacelia (C) for the debris treatments (1st-yearburn, 2nd-year burn, cut-unburned) and interspace on Steens Mountain, Oregon. Data are means with bars represent-ing one standard error. Significant differences (P < 0.05) among treatments for each response variable are indicated bydifferent lowercase letters.

(wet/frozen soils, higher fuel moisture), therewas no mortality of herbaceous perennials, andafter 3 years, perennial herbaceous cover andrichness were nearly 2-fold greater in treat-ments burned in spring than those burned infall.

We had hypothesized that burning cutjuniper would not result in a reduction ofperennial bunchgrass density compared to cut-unburned trees. There was a decrease inperennial bunchgrass density in the 1st-yearburn, though the reductions were not differ-ent than those for unburned trees (cut-unburned and 2nd-year burn) the first grow-ing season (1998) after treatment (Fig. 5B).This suggests that smothering by unburnedcut juniper had about the same effect onperennial grasses as did burning the first yearafter cutting. The 2nd-year burn removedsmaller amounts of juniper leaf litter (Fig. 3B)and perennial grass density tripled (Fig. 5B) inthe year of the burn (1999). We suspect thatsaturated, frozen surface litter and soils lim-ited heat penetration into the soil, which likelylimited mortality of grass seeds or seedlings.This would explain why perennial grass den-sity increased in the 2nd-year burn in 1999.

An interesting aside was the populationdynamics of cheatgrass. Despite drought con-ditions, cheatgrass increased between 1999 and2001. The subsequent decline of cheatgrass in2006 occurred in an above-average precipitationyear. The decline in cheatgrass may haveresulted from (1) less favorable seedbed condi-tions as litter was incorporated into the soiland exposure increased; (2) reduced soil nutri -ent availability; and (3) competition fromperennial grasses increased. The continuedincrease of perennial grass cover and densityduring the drought years suggests that peren-nial suppres sion of cheatgrass was a primaryfactor for cheatgrass decline.

Effects of Cut Juniper

Several studies of other species of juniperindicate that establishment and productivity ofherbaceous plants are negatively correlated tojuniper litter depth (Jameson 1966, Schott andPieper 1985, Horman and Anderson 2003) andpotential allelopathic compounds contained inlitter (Jameson 1970). Horman and Anderson(2003) concluded that allelopathic compoundsare species specific but rarely reach toxic lev-els in natural settings because they decom-

pose quickly. We did not test for allelopathicor litter depth effects on herbaceous plants;however, it appears that juniper leaf litter thatfell from cut trees influenced plant composi-tion and dynamics among the treatments.Felled trees and subsequent leaf drop createdlitter layers ranging from a shallow surfacecovering to a depth of several inches.

The higher amounts of juniper litter cover(Fig. 3B) in the cut-unburned treatment wereassociated with the following plant communitychanges: (1) perennial bunchgrass compositionshifted from primarily Thurber’s needlegrassand bluebunch wheatgrass to bottlebrush squir-rel tail; (2) there was limited establishment orreduced presence of several species, includingThurber’s needlegrass, Sandberg’s bluegrass,and most perennial forbs; and (3) since thefourth growing season after juniper cutting,cheatgrass dominated the herbaceous com -ponent. Horman and Anderson (2003) indicatedthat emergence of cheatgrass and native grassspecies was suppressed by increasing juniperlitter depth. However, in our study and others(Young et al. 1985, Miller et al. 2005, Bates etal. 2007a), cheatgrass has demonstrated theability to increase in areas of juniper litterdeposition about 4–5 years after westernjuniper control. Squirreltail density also in -creased in cut-unburned treatment in spite ofcheatgrass dominance. Others have reportedincreases in squirreltail in annual grass–domi-nated areas (Hironaka and Tisdale 1963, Bateset al. 2006). Cover of squirreltail in the cut-unburned treatment was less than in the othertreatments, which may indicate that squirreltailgrowth was being suppressed by the higherjuniper litter cover and/or continued domi-nance of cheatgrass. Other perennial grassspecies and native forbs were reduced or didnot increase under cut-unburned juniper, whichsuggested suppression by juniper litter and/orinterference by cheatgrass.

However, a certain level of juniper littercover appears to have been beneficial to estab-lishment of perennial bunchgrasses, as theirdensities were 60%–80% greater in the 2burned treatments than they were in the inter-space (lowest litter cover) and the cut-unburnedtreatment (high amounts of litter; Fig. 5B).Thurber’s needlegrass established most success-fully in the 1st-year burn treatment where littercover was lower than in the other cut treat-ments but greater than in the interspace.


As a group, perennial and annual forbs pre-ferred interspace and cut-and-burn treat-ments. On this site the potential for forbs to bea significant component of the understory islow, as perennial grasses dominate the under-story following juniper control (Bates et al.2005). On sites where native forbs are a greatercomponent of the understory, there may begreater potential for these species to increaseafter burning cut juniper (Miller et al. 2005).

Implications for Restoration and Management

Mechanical control of western juniper andother pinyon-juniper species is an effectivemethod to restore herbaceous productivity anddiversity in rangelands of the western UnitedStates (Miller et al. 2005). In this study, we didnot compare the juniper removal treatmentsto untreated woodland. However, based onearlier comparisons between cut and uncutwoodlands on this site (Bates et al. 2000, 2005,2007a), juniper cutting and cutting and burn-ing were effective at increasing understorycover and density. Without shrubs or juniperpresent, herbaceous cover potential on this sitefell between 25%–30% and potential perennialgrass densities ranged between 8 and 10plants ⋅ m–2 (Bates et al. 2005). From the earlierstudies, recovery of herbaceous cover requires2–4 years. All treatments except interspacewere at po tential cover levels by 2001, thefourth year after cutting. In 2006 the cut-unburned treatment was the only treatmentbelow 25% herbaceous cover. Perennial grassdensities reached site potential by the fourthyear after treatment in the 1st-year burn treat-ment. Perennial grass density in the othertreatments reached potential between 2002and 2006.

As mechanical treatment of invasive westernjuniper woodlands has become increasinglycommon, so have concerns about large amountsof western juniper debris and slash generatedby control operations. After cutting on juniper-dominated sites, 20% to nearly 100% of anarea may be covered with cut juniper and slash(Miller et al. 2005). Although there is thepotential for western juniper to be used as abiofuel, for many areas distance and accessi-bility limit economic viability. Thus, for manyrestoration treatments, mechanically treatedjuniper will need to be managed on site. Insome cases it may be best to leave cut juniper

in place to retain nutrients stored in litter(Miller et al. 2005) and to reduce soil erosion(Hastings et al. 2003). Seeded perennial grassestablishment may be enhanced under cuttrees and slash, particularly on more-arid sites(Eddleman 2002). However, as demonstratedin our study and others (Bates et al. 1998,2007a, Brockway et al. 2002) cut juniper andjuniper debris accumulations alter plant com-position and may slow native perennial recov-ery. Cut juniper and slash promote dominanceby undesirable annual grasses (Young et al.1985, Bates et al. 2005). Brockway et al. (2002)found that the smothering effects on herbaceousdiversity by debris could be mitigated by scat-tering one-seed juniper ( Juniperus mono -sperma) slash. The scattering of western juniperslash often more than triples treatment costsand may not be practical for large-scale appli-cations in these woodlands.

Cheatgrass presence remains a concern inwestern juniper woodlands, as it has shownpotential to increase rapidly and dominate theunderstory following western juniper control(Quinsey 1984, Evans and Young 1985). Theplant community in our study is an ecotonewhere annual grass infestation may or may notpose a threat following disturbance (Miller etal. 2000). Eliminating cheatgrass from thesecommunities is not practical, but our resultsindicate that designing treatments that alter ortake away preferred establishment sites havethe potential to reduce annual grass influence.Svejcar (2003) and Sheley and Krueger-Man-gold (2003) recommended that managers con-sider site availability, species availability, andspecies performance when developing strate-gies to reduce weed infestations. Targetingestablishment sites for weeds may redirectsuccessional trajectories so that desired out-comes are achieved (Svejcar 2003). Becauseareas of juniper litter accumulation providefavorable sites for establishment and develop -ment of cheatgrass, an obvious solution toreduce cheatgrass establishment is to removecut juniper trees or debris. Brockway et al.(2002) demonstrated that mechanical removalof cut pinyon-juniper increased plant diversityon New Mexico rangelands. Where mechani-cal removal may not be practical, our resultssuggest that winter burning of cut junipertrees and debris has the potential for enhancingnative plant community recovery. The speedof response may depend on site potential. In


aspen woodlands, early spring burning of cutjuniper resulted in no measurable mortality ofnative plants, and by the third year after treat -ment herbaceous cover had increased by 300%and species richness by 50%. Recovery ofnative plant vegetation in the present studydid not proceed as rapidly. However, ourresults demonstrated that burning cut westernjuniper in winter has the potential to speedrecovery of the native plant community andlimit the duration of high cheatgrass presence.

Winter burning under the conditions de -scribed in the study (Table 1) is limited to thefirst 2–3 years after tree cutting. From obser-vation, by the third year after cutting, most ofthe juniper leaves have fallen to the groundand burning in winter is not practical unlesssurface litter on the ground is dry. However, asdiscussed previously, burning cut westernjuniper when surface litters are dry severelyimpacts native herbaceous vegetation andpotentially may permit cheatgrass or otherweeds to dominate the site (Bates et al. 2006,2007b) and subsequently open these areas upto dominance by nonnative weeds. Given thehigh potential for damaging native plants andstimulating weed response by burning whenlitter and other fuels are dry, the best optionmight be to leave cut western juniper in place.In the other long-term studies on this sitethere was a period of cheatgrass dominancebeneath cut trees; however, with patience nativeperennials largely replaced cheatgrass afterabout 13 years (Bates et al 2005, 2007a). Inour study area, if current vegetation trendscontinue, cheatgrass will likely continue todecline in the unburned treatment as nativeperennial grasses increase.

ACKNOWLEDGMENTS

We are especially grateful to Fred Otley ofOtley Brothers, Inc., for the use of their prop-erty for the study. Completion of field mea-surements was accomplished with the assis-tance of a host of undergraduate summer rangetechnicians. Thanks are due to Jane Mangold,Kirk Davies, and 2 anonymous reviewers fortheir editorial comments and suggestions onearlier drafts of the manuscript.

LITERATURE CITED

BATES, J.D., R.F. MILLER, AND K.W. DAVIES. 2006. Restora-tion of quaking aspen woodlands invaded by western

juniper. Rangeland Ecology and Management 59:88–97.

BATES, J.D., R.F. MILLER, AND T.J. SVEJCAR. 1998. Under-story patterns in cut western juniper ( Juniperus occi-dentalis spp. occidentalis Hook.) woodlands. GreatBasin Naturalist 58:363–374.

______. 2000. Understory dynamics in cut and uncutwestern juniper woodlands. Journal of Range Man-agement 53:119–126.

______. 2005. Long-term successional trends followingwestern juniper cutting. Rangeland Ecology andManagement 58:533–541.

______. 2007a. Long-term spatial dynamics in a cut westernjuniper woodland. Western North American Natu-ralist 67:549–561.

BATES, J.D., R. MILLER, T. SVEJCAR, K. DAVIES, F. PIERSON,S. HARDEGREE. 2007b. Western juniper controlstudies: EOARC research report [online]. EasternOregon Agricultural Research Center, Station Report,Burns. 57 pp. Available from: http://oregonstate.edu/dept/EOARC/abouthome/scientists/jon.htm

BATES J., T. SVEJCAR, AND R.F. MILLER. 2002. Effects ofjuniper cutting on nitrogen mineralization. Journalof Aridland Environments 51:221–234.

BROCKWAY, D.G., R.G. GATEWOOD, AND R.B. PARIS. 2002.Restoring grassland savannas from degraded pinyon-juniper woodlands: effects of mechanical overstoryreduction and slash treatment alternatives. Journalof Environmental Management 64:179–197.

BUCKHOUSE, J.C., AND J.L. MATTISON. 1980. Potential soilerosion of selected habitat types in the high desertregion of central Oregon. Journal of Range Manage-ment 33:282–285.

BURKHARDT, J.W., AND E.W. TISDALE. 1969. Nature andsuccessional status of western juniper vegetation inIdaho. Journal of Range Management 22:264–270.

EDDLEMAN, L. 2002. Establishment and development ofbroadcast seeded grasses under western juniperslash. Pages 36–42 in Range Field Day ProgressReport. Range Science Series Report #5, Departmentof Rangeland Resources, Oregon State Universityand Eastern Oregon Agricultural Research Center,Corvallis.

EVANS, R.A., AND J.A. YOUNG. 1985. Plant succession fol-lowing control of western juniper ( Juniperus occi-dentalis) with Picloram. Weed Science 33:63–68.

GIFFORD, G.F. 1981. Impact of burning pinyon-juniperdebris on select soil properties. Journal of RangeManagement 34:357–359.

HASKINS, K.E., AND C.A. GEHRING. 2004. Long-term effectsof burning slash on plant communities and arbuscularmycorrhizae in a semi-arid woodland. Journal ofApplied Ecology 41:379–388.

HASTINGS, B.K., F.M. SMITH, AND B.F. JACOBS. 2003. Rapidlyeroding pinyon-juniper woodlands in New Mexico:response to slash treatments. Journal of Environ-mental Quality 32:1290–1298.

HIRONAKA, M., AND E.W. TISDALE. 1963. Secondary suc-cession in annual vegetation in southern Idaho.Ecology 44:810–812.

HORMAN, C.S., AND V.J. ANDERSON. 2003. Understoryspecies response to Utah juniper litter. Journal ofRange Management 56:68–71.

JACOBS, B.F., AND R.G. GATEWOOD. 1999. Restorationstudies in degraded pinyon-juniper woodlands ofnorth-central New Mexico. Pages 294–298 in S.B.Monsen and R. Stevens, compilers, Proceedings:


ecology and management of pinyon-juniper commu-nities within the Interior West, Sept 15–18, 1997,Provo, Utah. RMRS-P-9, Rocky Mountain ResearchStation, USDA Forest Service, Ogden, UT. 411 pp.

JAMESON, D.A. 1966. Pinyon-juniper litter reduces growthof blue grama. Journal of Range Management 19:214–217.

______. 1970. Degradation and accumulation of inhibitorysubstances from Juniperus osteosperma. Plant andSoil 33:213–224.

JOSIATIS, R.M. 1990. The effects of western juniper occu-pancy on changes in soil characteristics in relation toshrub and grass establishment in Owyhee County,Idaho. Master’s thesis, University of Idaho, Moscow.84 pp.

MILLER, R.F., J.D. BATES, T.J. SVEJCAR, F.B. PIERSON, AND

L.E. EDDLEMAN. 2005. Biology, ecology, and man-agement of western juniper. Technical Bulletin 152,Oregon State University Agricultural ExperimentStation, Corvallis. 77 pp.

MILLER, R.F., AND J.R. ROSE. 1995. Historic expansion ofJuniperus occidentalis (western juniper) in south-eastern Oregon. Great Basin Naturalist 55:37–45.

MILLER, R.F., T.J. SVEJCAR, AND J.R. ROSE. 2000. Impactsof western juniper on plant community compositionand structure. Journal of Range Management 53:574–585.

MILLER, R.F., M. WILLIS, J. ROSE, D. REINKENSMEYER,AND B. ANTHONY. 1999. Effects of community struc-ture on avian populations in juniper woodlands andshrub steppe. Pages 106–111 in Range Field Day1999 Progress Report. Special Report 1002, Depart -ment of Rangeland Resources, Oregon State Univer-sity and Eastern Oregon Agricultural Research Cen-ter, Corvallis.

NOSON, A.C., R.A. SCHMITZ, AND R.F. MILLER. 2006.Influence of fire and juniper encroachment on birdsin high-elevation sagebrush steppe. Western NorthAmerican Naturalist 66:343–353.

PETERSON, R.G. 1985. Design and analysis of experiments.Marcel Dekker, Inc., New York. 429 pp.

PIERSON, F.B., J.D. BATES, T.S. SVEJCAR, AND S. HARDGREE.2007. Long-term changes in runoff and erosion aftercutting western juniper. Range Ecology and Man-agement 60:285–292.

QUINSEY, S.D. 1984. Fire and grazing effects in westernjuniper woodlands of central Oregon. Master’s thesis,University of Washington, Seattle. 125 pp.

ROBERTS, C., AND J.A. JONES. 2000. Soil patchiness injuniper-sagebrush-grass communities of centralOregon. Plant and Soil 223:45–61.

SAS INSTITUTE, INC. 2002. User’s guide. Release 8.03.SAS Institute, Inc., Cary, NC.

SCHOTT, M.R., AND R.D. PIEPER. 1985. Influence of canopycharacteristics of one-seed juniper on understorygrasses. Journal of Range Management 38:328–331.

SHELEY, R., AND J. KRUEGER-MANGOLD. 2003. Principles ofrestoring invasive plant-infested rangelands. WeedScience 51:260–265.

SVEJCAR, T. 2003. Applying ecological principles to wild-land weed management. Weed Science 51:266–270.

VAITKUS, M., AND L.E. EDDLEMAN. 1987. Compositionand productivity of a western juniper understoryand its response to canopy removal. Pages 456–460in R.L. Everett, editor, Proceedings: Pinyon-JuniperConference. General Technical Report INT-215,USDA Forest Service, Intermountain Research Sta-tion, Ogden, UT. 581 pp.

WALL, T., R.F. MILLER, AND T.S. SVEJCAR. 2001. Juniperencroachment into aspen in the northwest GreatBasin. Journal of Range Management 54:691–698.

YOUNG, J.A., R.A. EVANS, AND C. RIMBY. 1985. Weed controland revegetation following western juniper control.Weed Science 33:513–517.

Received 13 December 2007Accepted 9 June 2008


Documents

Herbaceous Succession After Burning of Cut Western Juniper Trees