Effects of Eradication and Restoration Treatments On Italian Thistle (Carduus pycnocephalus)

2011·5.4 MB·English

by  Thomas Mcginnis

I JUL 1 1 2012
Madrono, Vol. 58, No. 4, pp. 207 213, 2011
EFFECTS OF ERADICATION AND RESTORATION TREATMENTS ON ITALIAN
THISTLE (CARDUUS PYCNOCEPHALUS)
Thomas McGinnis and Jon Keeley
CA
United States Geological Survey, Three Rivers, 93271-9651
[email protected]
Abstract
Low elevation grasslands in California long have been dominated by Mediterranean grasses, but
many areas still have large native forb populations. Alien forbs invade these grasslands, displacing
both native and other alien species. Italian thistle is a noxious alien herb that has recently invaded
these grasslands, including ungrazed blue oak {Querciis douglassii) and interior live oak {Qucrcus
wis/izenii)standsin Sequoia National Park. Here, Italian thistletendstodominate underoaksand has
the potential to substantially alter the foothill ecosystem by displacing native plants and acting as a
ladderfuelthatcancarryfiresinto theoakcanopy. Wetested theeffectsofselectivelyreducing Italian
thistle populations alone and in combination with restoration of native species. Two thistle
eradication techniques (clipping and the application of clopyralid herbicide) and two restoration
techniques (addition of native forb seeds or planting native grass plugs) were used. After two
consecutive years of treatment we found: a) clipping was not effective at reducing Italian thistle
populations (clipping reduced Italian thistle density in some areas, but not vegetative cover), b)
herbicide reduced both Italian thistle density and vegetative cover for the first two growing seasons
after application, but cover rebounded in the third growing season, c) native forb cover and species
richness were not significantly affected by clipping or spot-treating with herbicide, d) the grass and
forb addition treatments by themselves were not effective at reducing Italian thistle during the course
of this study and e) sowing annual forb seeds after cHpping resulted in greater forb cover and
moderately reduced Italian thistle vegetative cover in the short term.
KeyWords: Alienthistle,annualgrasslands,blueoak,foothills,Italianthistle,nativeplantrestoration.
Approximately 20% ofCalifornia's 6550 plant Lichti 2000). They also potentially affect wildlife
species are naturalized non-natives (http://ucjeps. movement due to deterrence from sharp spiny
berkeley.edu/interchange) and 200 of these are leaves and stems (Parsons 1973). Moreover,
considered to be invasive, having the ability to Italian thistle can potentially threaten oak trees,
displace native species and disrupt ecosystem because, they can grow to 2 m and generate
processes (Bossard et al. 2000). Low elevation ladder fuels. Ladder fuels connecting surface
grasslands and savannas are among the most litter to oak canopies during wildfires are a major
highly invaded ecosystems in the state, and today contributor to blue oak canopy scorch and top-
they aredominated bya fewspecies ofnon-native kill (Horney et al. 2002). The potential for Italian
Mediterranean grasses and a mixture of native thistle to shade out native forbs, alter grazing
and non-native forbs (Heady 1988; Bartolome patterns and convert surface fires into crown fires
et al. 2007). These grasslands are vulnerable to presents significant management concerns in the
additional invasions by ahen forbs. Some of the blue oak woodlands throughout the state.
more noxious species are the spiny Asteraceae When caught during their early phase of
such as thistles in the tribe Cynareae (Bossard colonization invasive plant eradication may be
et al. 2006). feasible, but once established, they present
Italian thistle {Carduus pycnocephalus L.), formidable challenges for resource managers
which is native to Europe and Asia, is a (DiTomaso et al. 2007). Methods aimed at
widespread noxious thistle with an annual or eradication ofinvasive species may produce short
biennial life cycle in California annual grasslands term reductions in cover and density, but
and oak woodlands. It is particularly robust populations typically return once direct control
under blue oak canopies, suggesting it has high ceases. One reason is that eradication methods
nitrogen and/or moisture requirements (Holm can disrupt ecosystem processes and create
et al. 1997; Perakis and Kellogg 2007). Such disturbance sites for future colonization of
invasions may threaten native understory species invasive species (DAntonio and Meyerson
on these sites, which could be a particular 2002). Also, even with precise treatments, erad-
concern if rare species occupy the same habitat. ication methods affect potential native competi-
Dense Italian thistle populations with overlap- tors as well as other invasive species (DiTomaso
pingrosette leaves arecapable ofexcluding native et al. 2007). Non-native species thrive in Califor-
plant species by monopolizing light (Bossard and nia annual grasslands partially because these
MADRONO
208 [Vol. 58
Fig. 1. Italian thistle study sites, Sequoia NP, California.
grasslands are highly disrupted ecosystems where the Sierra Nevada foothills in Sequoia National
the competitive balance between native species Park (Tulare County, Cahfornia; Fig. 1). The
has diminished (Heady 1988). vegetation was blue oak and interior live oak
In the Sierra Nevada foothills of Sequoia woodland and savanna; herecoolwetwintersand
National Park, Itahan thistle has been a persis- warm dry summers characterize its Mediterra-
tent invasive and the target of eradication nean climate. Snow is uncommon in the foothills
attempts since 2002. Typically resource managers ofthe southern Sierra Nevada and thereforemost
have utilized two eradication treatments: spray- precipitation is from rain. Except for the winter
ing with the herbicide clopyralid and cutting or of 2009-2010, precipitation was at or below the
pulling (http://www.nps.gov/seki/naturescience/ 57 year average during this study (Fig. 2).
badcapy.htm). The present study was conducted Using both National Park Service (NPS) maps
to compare clopyralid and clipping and to couple and on-the-ground surveys of Italian thistles we
these treatments with two restoration treatments established study sites throughout the 20 largest
(sowing native forb seeds or planting native thistle populations in the park: 10 populations in
perennial grass plugs) meant to increase potential the middle fork watershed and 10 populations in
native competitors. We hypothesized that sup- the north fork watershed. Within each ofthe two
pression ofItalian thistle would be best achieved watersheds, 10 "canopy" sites were selected
by couplingeradication with restoration ofnative beneath the drip-line of blue oak, interior live
competitors. Results were determined over mul- oak or California buckeye {Aesculus californica
tiple years so that treatment longevity could be [Spach] Nutt.) trees, at least 1 m from a tree bole.
determined. An additional 10 sites were selected outside the
drip-line and were considered "open" sites. The
Methods amount of shading and solar intensity in both
canopy and open plots varied widely due to
Study Areas and Treatments aspect, canopy height and proximity to trees,
shrubs and boulders. Each population consisted
Our study was conducted between May 2006 of several thistle patches located both beneath
and June 2010 in the Kaweah River watershed in trees and in the open. Because most Italian thistle
2011] McGINNIS AND KEELEY: MANIPULATION OF ITALIAN THISTLE 209
Fig. 2. Precipitation at Ash Mountain, Middle Fork Kaweah River, at the elevation of the study sites. Data
collected by NPS and compiled by California Dept. of Water Resources (http://cdec.water.ca.gov/) and Desert
Research Institute (http://www.wrcc.dri.edu).
patches were too small to fit all the treatment Althoughherbicidetreatmentsareoftenapplied
plots, few study siteswere in a singlepatch; rather with broadcast spraying, our study treatment
a site was defined as any number of nearby consisted of spot-treating Italian thistle plants
patches with similar slope position and substrate with a 0.08% concentration ofclopyralid and care
(i.e., a series of open or canopy patches along a was taken to target just thistles. This resulted in
single ridge or slope). Because all ofthe blue oaky concentrations of herbicide commensurate with
interior live oak savannas in the park were easily the density and size of the thistles during the
accessed by roads and trails, there was no need to two consecutive years that herbicide was applied.
limit our selection based on access. Clopyralid has both pre and post-emergent
Nine2 X 2mplotswerecentered overthemost qualities. Theherbicideisaplantgrowth regulator
denselypopulatedthistle patches ateachofthe20 and is used to kill many forbs while not kiUing
sites. There were three eradication treatments grasses and trees (DOW Agro-Sciences 1998). It
(clipping, herbicide and no clipping or herbicide) is effective on species of Asteraceae, Apiaceae,
fully crossed with three restoration treatments Caryophyllaceae, Fabaceae, Polygonaceae and
(native grass planting, native forb seeding and no Solanaceae families (Rice and Toney 1996; Tyser
plantingorseeding)(3x3 = 9;Table 1). Abuffer et al. 1998; Reever-Morghan et al. 2003).
zone was created around each plot for walking Clippingand herbicide treatments wereapplied
between plots without disturbing the plots. early in the growing season during bolting to
The clipping treatment was applied with a gas- early rosette stage before peak fiowering and
powered line trimmer and was evenly applied before thistle seeds had dispersed. Clipping was
near ground level to all non-woody plants in an applied in early May 2007 and 2008. Herbicide
assigned plot and its buffer zone. Clipped was applied in late April-early May in 2007 and
biomass was not removed. 2008 (Table 2).
Table 1. Italian Thistle Eradication Treatments (Rows) and Site Restoration Treatments
(Columns) in Sequoia National Park. These treatments were replicated beneath tree canopy and in open
grassland sites. ''Forb seed treatment: 176 Amsinckiamenziesiiand 486 Phacelia cicutaria seeds per m-. Ml Melica
califomica and 40 Elymus trachycaulus plants were one year old when they were planted 25 cm apart; some
plantings were skipped due to rocks. The two species were planted on separate sides ofthe plot; each species was
planted over a 1 X 2 m area. ''Due to the nature ofthis clipping treatment, all grasses and forbs were clipped near
ground leveland biomassleft. '^Thistleplantswereindividuallycoated with0.08% clopyralid (1.9 mLTransline®/L
H2O, or 1/4 oz/gal).
Thistle eradication Site restoration treatment
treatment Control Forb seed" Planted grass*"
Control No treatment Seeded only Planted grass only
Clipped'' Clipped only Clipped and seeded Clipped and planted grass
Herbicide'^ Herbicide only Herbicide and seeded Herb, and planted grass
MADRONO
210 [Vol. 58
Table 2. Timelineof Treatments (Listed Above) and Plot Surveys.
JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASOND
2006 2007 2008 2009 2010
Start Collectdata.
Collectgrassand Seed Plant Collectdata,herbicideand
grass herbicide Collectdata Collectdata
forbseeds forbs grass clip(secondtreatment)
plugs andclip
Seeds used for restoration treatments were ed maximum likeHhood calculations. Denomina-
collected separately within each watershed and tor degrees of freedom were computed using the
m
within +/— 500 elevation. Seed collection was Satterthwaite approximation (Littell et al. 1996;
during spring and summer 2006 (Table 2). SAS Institute 2007). Square-root and natural
Grass plugs were propagated (shaded, misted log transformations were used, when necessary,
and weeded) for one year in potting soil at the Ash to correct normality and homoscedasticity.
Mountain greenhouses in Sequoia National Park Tukey-Kramer tests were used to detect differenc-
(Table 2). Thegrassplantingtreatmentwasapplied es between treatment pairs.
in late January through mid-February 2008 (Ta- Native forbcoverincanopy sites, Italian thistle
ble 2). Pottingsoilwasnot removedwhileplanting. density, native grass cover and non-native grass
Dibbletoolswere usedtopunch4X15 cmholes in cover could not be made normal and homosce-
theground tomatchthesizeofthegrassplugs. The dastic through transformation; therefore, for
two grass species used in the restoration treatment these variables the Kruskel-Wallis test was used
were Melica californica Scribn. (41 one-year-old instead of the mixed model and Nemenyi tests
plugs spaced 25 cm apart) and Elynms trachycau- were used to detect differences between treatment
lus (Link) Shinners (40 one-year-old plugs spaced pairs (Kruskel-Wallis tests were performed with
2th5ecpmloatpsaarnta).lyTsihsowuagshfpolratnhteedenotnirsee2paxra2teplsoitd,esnootf cSaYlcSuTlaAtTed i1n1assopfrtewaadrseheaetn,dasNepmerenZyair 1t9e9st0s).wPer<e
for the individual grass species planted. 0.05 was considered significant for all analyses.
The native forb seed treatment was applied
in mid-November 2007. Based on average seed Results
weights, forb seeds comprised 180 Amsinckia
nienziesii (Lehm.) A. Nelson & J. F. Macbr. and Italian Thistle Cover and Density
500 Phacelia cicutaria Greene seeds per square
meter, which were mixed and scattered for a In the short term Italian thistle density (Fig. 3)
density of680 seeds/m^ in the assigned plots and and vegetative cover (Fig. 4) were significantly
buffer zone. reduced by herbicide treatment both under tree
canopy and in the open (2009; two years after the
Data Collection and Analysis initial treatment and one year after the follow-up
treatment, see Table 2 for fimeline). Coupling the
Percentage vegetative cover of individual spe- herbicide treatment with a restoration treatment
cies within each plot was assessed in April-May (seeding or planting native species) did not change
2009, using 1%, 5% and 10-100% in 10% the results for thistle density and cover in the short
increments (Table 2). Italian thistle cover was term. In the longerterm(2010; threeyears afterthe
measured in April June 2007-2010 and its initial herbicide treatment and two years after the
density was measured in April-June 2009. Native follow-up herbicide treatment) the herbicide treat-
grass density was assessed in April-June 2009. ment alone did not significantly affect thistle
Native forb cover in open sites, Italian thistle vegetative cover, but the combined treatment of
cover and native and alien species richness were herbicideplusseedingresultedinlowerthistlecover
analyzed with mixed effects models (SAS Insti- under tree canopy but not in the open (Fig. 5).
tute 2007). Treatment and river drainage were Clipping did not significantly affect thistle
fixed effects and sitewas the random effect. These density or cover in the short term or long term
modelscompared treatment effects across all sites when canopy and open sites were analyzed
independently for canopy and open sites. For a independently (Figs. 4 and 5) or combined
multiple year comparison of Italian thistle cover (Fig. 6). Clipping combined with seeding was
(2007 2010), canopy and open sites were com- effectiveatreducingthistlecoverintheshortterm
bined. All mixed models were fitted using restrict- only and only in canopy sites (Figs. 4 and 5).
2011] McGINNIS AND KEELEY: MANIPULATION OF ITALIAN THISTLE 211
120
ab 50 a ac
100 45
80 40 ac ad
=c 1re: 60 ad 70) oo> 3350 ab
- ^ 40 bed bed beTd beTd .f*«^? TCD3 2250 bed
20 fl ce ce = P 1150 b 1^ bd
0 5 Q H Q
oo 0
Treatmentundertreecanopy
Treatmentundertreecanopy
70 ab
O)
« ce 60 icS 40
=OZ*iS-S -rJQc52. 435000 aed aed .!+*a0io0S•EJ)f)2> Tiu>oC3- 2233150055 ab ab ac ab
«^ -0o) 10 de de cCD o 10 be
TO 5 - c Q
oo> 0
o
CM
Treatmentin openareas
Fig. 3. 2009 Italian thistle density in canopy sites (A) Treatmentin openareas
awtIrfsotierinogasendtnltathiiotiefarolmrinalpecstnae.inottnothsfniTdssthtiii2elnetf0ceef0she27nrse0i-e(ar0qnB2ma7c)u0deee0isw8aacilt)a(ntestdtaheiteioe=nrdrnrieenps0aSg(ep.becea0aolcq5tnv.tiuepdedoptiipoainlbngaacnrNots2aami0ntnb0igidi8non)donainhtaccaeailerntobedniisPnscaistrnidhotkoeefe cihstfFinoehiotoergmetcb.bhriaiwie4clnin.silnoatsdtp.toeeyri2rato0Ttrns0hieois9oteafentosImft2tsea(e0Inalac0tit)mha7aslen-naii2iantqn0nhldu0iee2tst8sh0tto)ile0ps(er7tesilvenneaeaegnbdeseSiodritentavaeqgdretsiueiacopv(nabeiebtdaaa)icrtposoewNlnvdiaaet(nitirhcntnildipoiir2nenpce0grapas0ciltpo8eneen)gnsccPtteaaaannnrgtiddokeon
significant difference at a = 0.05.
Native Grasses and Forbs
Native grass plantings were not very successful
due to extreme mortality in the first year, but sites in 2009; clipping plus planting native grass
native grass cover began to increase after the significantly reduced alien grass cover in open
initial die-off. Only one third of plots that were sites in 2009. The five most abundant alien grass
planted had survivors after 1.5 years (June 2009), species (by cover) in these plots were Bromus
but one year later (June 2010) all of these plots diandrus Roth (most abundant), Avena barbata
still contained live native grasses. Pott ex Link and Bromus hordeaceus L. (both
Native forbs occurred in all but one of the species were equally abundant and together
study plots before the treatments and in all plots covered 15% less ground surface area than B.
after treatments. In canopy sites in 2009, native diandrus), Avena fatua L. and Bromus arenarius
forb cover was not significantly different in Labil. (both equally abundant and comprised
seeded versus control plots. In open sites, native 40% less ground surface area than A. barbata and
forb cover was not significantly affected by any B. hordeaceus). Bromus diandrus was the domi-
of the treatments. Native forb cover was not nant alien grass species in canopy sites and
assessed in 2010. Bromushordeaceouswas the dominant alien grass
species in open sites. Cover ofthe other abundant
Ahen Grasses grass species was not significantly different
between open and canopy sites. In general,
Alien grass cover was only significantly affect- clipping reduced alien grass cover while herbicide
ed by one of the treatments, and only in open increased it.
MADRONO
212 [Vol. 58
70 50
60
40
ab
0? 50
ab
I 40 R 30
20
2 20
CD 10
Pre-treatment 2008 2009 2010
Year
Treatmentundertreecanopy Fig. 6. Italian thistle percentage vegetative cover in
control (white bars), clipped (striped bars) and herbi-
cide-treated (dotted) plots from 2007 (pre-treatment)
through2010. Canopyandopenplotsarecombinedfor
this analysis. The same letter above bars indicates no
significant difference at a = 0.05.
Italian thistles in blue oak and interior live oak
savanna plots in the short-term; however, it was
not effective at controlHng thistles two years after
treatmentswere stopped. Due to clopyralid's pre-
emergent effect (DiTomaso et al. 2007) treat-
ments applied in the spring one year controlled
Italian thistle density and cover the next year, yet
the Carduus seed bank can remain viable for 8-
Treatmentinopenareas 10 years (i.e., Burnside et al. 1996; Sindel 1997)
Fig. 5. 2010Italian thistlevegetativecoverpercentage and in order to be effective in the longer term,
in canopy sites (A) and open sites (B) with respect to herbicide would need to be reapplied to draw
combinationsofItalianthistleeradicationItalianthistle down the seed bank. Unlike the herbicide
eradication (clipping and herbicide treatments in 2007 treatment, clipping was not an effective eradica-
and repeated in 2008) and site restoration techniques tion treatment even in the short-term.
(seeding and planting once in the winter of2007-2008) Because Italian thistle seeds could eventually
ainboSveequboairas Niantdiiocnataels Pnaorksifgonotihfiilclasn.t Tdhifefesreanmcee laetttear enter from outside sources and the fact that
= 0.05. healthy native plant populations can resist aliens
(Young et al. 2009), it has been proposed that
Species Richness eradication methods should be coupled with
restoration treatments for the longest-lasting
Native species richness the first year after all effect(D'AntonioandMeyerson2002; DiTomaso
treatments were complete (2009) was not signif- et al. 2007). We hypothesized that suppression of
icantly different between treatments and controls Italian thistle would be best achieved by coupling
in canopy or open sites. The controls had a mean eradication with restoration of native competi-
of4-5 species per 4 m^ plot, while the treatments tors, but we did not find sufficient evidence to
had 4-7 species per plot. Treatments with the support this hypotheses. Although clipping fol-
greatest mean native species increases (2 more lowedbyseedingnativeforbsmodestlycontrolled
nativeforbspeciesperplotthanthecontrol)were: Italian thistle density and cover in canopy sites in
clipped plus seeded plots in canopy sites and the short term, it did not have an effect two years
seeded, clipped plus seeded, clipped plus planted after treatment.
and herbicide plus seeded plots in open sites. The other restoration treatment, planting
Compared to the controls in both the canopy and native grasses, was also ineffective at reducing
open, aliengrassandforbspeciesrichnesswasnot Italian thistle cover, at least in the few years of
significantly affected by any ofthe treatments. this study. Despite the low percentage of native
grass cover and high mortality the first year, the
Discussion fact that planting increased native grass fre-
quency indicates that foothill sites currently
Spot-treating Italian thistle with the selective devoid of native grasses could support these
herbicide clopyralid was effective at controUing species, once established. Also, because peren-
2011] McGINNIS AND KEELEY: MANIPULATION OF ITALIAN THISTLE 213
nial grasses grow so slowly in this environment D'Antonio, C. and L. a. Meyerson. 2002. Exotic
longer term monitoring of this treatment might plant species as problems and solutions in ecolog-
be useful. In these sites Italian thistle density ical restoration: a synthesis. Restoration Ecology
and cover were stable over four years indicating 10:703-713.
that this species had saturated sites before our DiToMASo, J. M., S. F. Enloe, and M. J. Pitcairn.
experiment began. Without intervention these 2007. Exotic plant management in California
annual grasslands. Pp. 281 296 in M. R. Strom-
populations are likely to remain as seed sources berg, J. Corbin, and C. D'Antonio, (eds.), Califor-
for further dispersal and will present formidable nia grasslands: ecology and management. Univer-
challenges for resource managers in the Sierra sity ofCalifornia Press, Berkeley, CA.
Nevada foothills. We recommend more research Dow Agro-Sciences LLC. 1998, Clopyralid a North
in the area of eradication combined with American technical profile. Indianapolis, IN.
restoration so that an effective combination of Heady, H. F. 1988. Valley grassland. Pp. 491-514 in
treatments can be found. We learned that spot- M. G. Barbour and J. Major, (eds.). Terrestrial
treating with clopyralid was an effective short- vegetation of California. California Native Plant
term treatment, but we did not come up with a Society, Sacramento, CA.
longer-term solution. Broadcast application of Holm, L., J. Doll, E. Holm, J. Pancho, and J.
clopyralid would be easier to apply and would aHnedrbdeirsgtreirbu.ti1o9n9.7.JoWhonrlWdilweeeyds&. NSaotnusr,alInhci,stoNreiews
protect larger areas from reinvasion, but could York, NY.
potentially harm certain native plants; therefore, HoRNEY, M., R. B. Standiford, D. McCreary, J.
a study on the effects ofthis herbicide on native Tecklin, R. Richards. 2002. Effects of wildfire
flora would be warranted. on blue oak in the northern Sacramento Valley.
Pp. 261^267 in R. B. Standiford, D. McCreary,
Acknowledgments AND K. L. Purcell (tech. eds.). Proceedings of the
(ReTsheaanrkcsh tEocoJluolgiiestY)eeand(StaTtiosdtidciaEn)s,quMeat(tReBsreoaorkcsh fCiafltihforsnyima'psoscihuamngionng olaankdscwaopoed.laUndSsD:AoaFkosresitn
Ecologist) from the USGS Western Ecological Re- Service General Technical Report PSW-GTR-184.
search Centerfortheir helpful comments and advice on USDA, Forest Service, Pacific Southwest Research
this manuscript. We thank Melanie Baer-Keeley, Amy Station, Albany, CA.
Brown and John Nelson for their many hours ofhard Littell, R. C, G. a. Milliken, W. W. Stroup, and
work throughout the project. Athena Demetry and R. D. WOLFINGER. 1996. SAS system for mixed
Anne Pfaff assisted in the planning and logistics. The models. SAS Institute, Inc. Cary, NC.
following people helped with the field work: Corie Maron, J. L. AND R. L. Jefferies. 2001. Restoring
Cann, Pam Collins, Linda Franson, Lori Gant, Lee enriched grasslands: effects of mowing on species
Goldstein, Marcia Goldstein, Sarah Graber, Jonathan richness, productivity and nitrogen retention.
Humphrey, Audrey Lucterhand, Kelly McGinnis, Kate Ecological Applications 11:1088-1100.
McLaren, Liz Nanney, Anne Pfaff, Shelley Quaid, Parsons, W. T. 1973. Noxious weeds of Victoria.
Dylan Schwilk, Christine Shook and Rich Thiel. We Inkata Press, Melbourne, Australia.
also appreciate the following SEKI divisions that Perakis, S. S. and C. H. Kellogg. 2007. Imprint of
provided help: Interpretation, Resources Management oaksonnitrogenavailabilityand 5''^N inCalifornia
& Science, and Maintenance. Fundingwas from USGS grassland-savanna: a case of enhanced N inputs?
Park-Oriented Biological Support. Any use of trade, Plant Ecology 191:209-220.
product, or firm names in this publication is for Reever-Morghan, K. J., E. A. Leger, and K. J.
descriptive purposes only and does not imply endorse- Rice. 2003. Clopyralid effects on yellow starthistle
ment by the U.S. Government. (Centaurea solstitalis) and nontarget species. Weed
Science 51:596-600.
Literature Cited Rice, P. M. and J. C. Toney. 1996. Plant population
BartHMoO.lPoKRmI.eN,SStOJrN.o.mWb.2e0,r07gW.,.VJaJ..llCBeoayrrbgrirynas,,slTaa.nnddG.rCPi.pg.gDs3',7A1na-t3on9nd7ioPi,.n SASrspeIosntpstotenidstekustneat.opwbe2re0o0da7d.ccoansSttrAoSlh.erObDnioclwiidnneeEDaaporpclt®ihca59t1.i2:o.1n4s-S1Af9o.Sr
(eds.), California grasslands: ecology and manage- Institute, Cary, NC.
ment. UniversityofCalifornia Press, Berkeley, CA. SiNDEL, B. M. 1997. The persistence and management of
BossARD, C, M. Brooks, J DiTomaso, J. Randall, thistlesinAustralianpastures. ProceedingsoftheNew
C. RoYE, J. SiGG, A. Stanton, and P. Warner Zealand Plant Protection Conference 50:453^56.
(eds.). 2006. California invasive plant inventory. Tyser, R. W., J. M. AsEBROOK, R. W. Potter, and
California Invasive Plant Council, Berkeley, CA. L.L. KuRTH. 1998. RoadsiderevegetationinGlacier
AND R. LiCHTl. 2000. Carduuspycnocephalus L. National Park, USA: effects of herbicide and
Pp. 86-90 in C. C. Bossard, J. M. Randall, and seedlingtreatments. Restoration Ecology6:197-206.
M. C. Hoshovsky, (eds.). Invasive plants of Young, S. L., J. N. Barney, G. B. Kyser, T. S.
California's wildlands. University of California Jones, and J. M. Ditomaso. 2009. Functionally
Press, Los Angeles, CA. similarspeciesconfergreaterresistancetoinvasion:
BuRNSiDE, O. C, R. G. Wilson, S. Weisberg, and implications for grassland restoration. Restoration
K. G. Hubbard. 1996. Seed longevity of41 weed Ecology 17:884-892.
species buried 17 years in eastern and Western Zar, J. H. 1990. Biostatistical analysis. Prentice-Hall,
Nebraska. Weed Science 44:74-86. Englewood Cliffs, NJ.