Effects of Fire and Fireline Disturbance on the Plant Community in a Southern California Ecological Reserve

2013·5.5 MB·English

by  Jolene R Moroney

Madrono, Vol. 60, No. 3, pp. 173-178, 2013
EFFECTS OF FIRE AND FIRELINE DISTURBANCE ON THE PLANT
COMMUNITY IN A SOUTHERN CALIFORNIA ECOLOGICAL RESERVE
JOLENE R. MORONEY AND PHILIP W. RUNDEL
Department of Ecology and Evolutionary Biology, University of California, Box 951606,
CA
Los Angeles, 90095-1606
[email protected]
Abstract
Native plants in most California ecosystems are adapted to fire, but altered fire regimes and
disturbance from firefighting activity, such as the construction offirebreaks or firelines, can change
plantcommunitycompositionandtheratioofnativetononnativespecies. InOctober2007, awildfire
burned 710 acres through a chaparral/grassland mosaic on an ecological reserve, providing an
opportunity to quantify fire and fireline disturbance impacts on native and nonnative species under
differing disturbance conditions. In the spring of 2012 we sampled the plant community in three
adjacent sites, focusing on Centaurea melitensis, which is a common nonnative invader after fire in
California chaparral. The first site was burned and bulldozed, the second site was burned but not
bulldozed, and the third sitewasnot burned orbulldozed. The first sitehad also been sampled in the
springof2008. Afterfouryearswithin theburnedfirelinesite, themeanrelativecoverofC. melitensis
decreased from 72% to 28%, but its density increased, and there were increases in the covers of
nonnative annualgrasses, litter, and native plants. Amongthe three sites in 2012, both oftheburned
sites hadhigherdensity andcoverofC. melitensisand lower relativecoverofannual grassesthan the
unburnedsite. Theonlysitewithnotablenativeperennialpresencewastheburnedfireline. Theresults
ofourstudy suggestthat therecruitment ofC. melitensisand some native speciesis promotedbyfire.
Intheabsenceofadditionaldisturbancebyfirelines,persistenceofthesetaxaislimitedbycompetition
from nonnative annual grasses.
Key Words: California chaparral, Centaurea melitensis disturbance, fire, firelines, invasive plants,
,
Mediterranean ecosystems, nonnative annual grasses.
Fire has been a presenceforthousands ofyears part of the University of California Natural
in southern California, from prehistoric infre- Reserve System. The land was used primarily for
quent lightning ignitions (Keeley and Fothering- cattle grazing from the early 19th century until
ham 2001) to Native American burning (Keeley the site became a reserve in 1997. The Reserve
2002) to the relatively frequent anthropogenic supports coastal sage scrub, chaparral, native
fires of the present (Keeley and Fotheringham grassland, valley oak savanna, and other native
2001). Although most native plants in California vegetation communities but, perhaps due to the
ecosystems are adapted to fire, fire frequency in heavy disturbance associated with cattle grazing,
many areas has increased beyond the extent of these native communities have been invaded in
the natural regime, facilitating the recruitment of some places by nonnative species. Wildfire had
nonnative invaders (Hobbs and Huenneke 1992), been excluded from the Reserve for at least
supporting their persistence (Haidinger and 100 years when, in October 2007, a wildfire
Keeley 1993), and engendering the exclusion of burned 710 acres through a chaparral/grassland
native species (Keeley and Brennan 2012). mosaic on the Reserve. Protective firelines had
Because of the threats to property, safety, and been established at Sedgwick many years before.
native biodiversity, California has an active The Sedgwick Fire provided an opportunity to
approach to fire management, with practices quantify fire and fireline disturbance impacts
such as the use of firebreaks or firelines that on native and invasive species under differing
promote disturbances to plant communities. The disturbance conditions.
disturbance effects of the construction and Centaureamelitensis L. (Asteraceae) is a weedy
maintenanceoffirebreakscanpromotenonnative annual forb that is native to the Mediterranean
plant colonization (D’Antonio et al. 1999; Basin and an aggressive invader in California
Merriam et al. 2006). The interaction of the shrublands and grasslands (Moroney and Rundel
effects ofdisturbance from firebreaks or firelines 2013). It is commonly observed after fire in
with impacts of frequent fire would be expected California chaparral (Keeley et al. 2005). It can
to additionally promote the colonization of be dominant in disturbed areas, and has been
nonnative annual species (Merriam et al. 2006). found to out-compete native species (Moroney
The Sedgwick Reserve in the Santa Ynez et al. 2011). In a comparative study of the
Valley of Santa Barbara County, California, is demographics of C. melitensis in its native and
MADRONO
174 [Vol. 60
invasive ranges, a dense population of C. County Fire Department, personal communica-
melitensis was found and surveyed in a burned tion). Prior to the establishment of the Natural
fireline on the Sedgwick Reserve in the spring Reserve on the site the land was used for cattle
following the Sedgwick Fire (Moroney and grazing. We sampled three sites located on the
Rundel 2013). We returned to the site four years Paso Robles Formation with a Shedd silty clay
later to quantify the changes in cover and density loam soil.
of C. melitensis and the associated plant commu- The first of the three sites (Burned, Fireline)
nity in the area disturbed by fire and the fireline. was located along a ridge that was both burned
We also compared the cover and density of C. and bulldozed during the 2007 Sedgwick Fire. A
melitensis and the associated community compo- bulldozercleareda one-bladewidth(12ft)fireline
sition in this site to an adjacent burned site with on the ridgeline after the fire burned through the
no fireline, and to an undisturbed (unburned, area to create access to the rest of the fire. This
ungraded) site. Of particular interest was the fireline was constructed several years before the
comparative behavior of C. melitensis with Sedgwick Fire and had been intermittently
nonnative annual grasses, which appear to
maintained (S. Alderete, Santa Barbara County
competitively displace C. melitensis in California
Fire Department, personal communication). In
sitTeshe(MoobrjoecnteiyveasnodfRtuhnisdesltu2d0y13w)e.re to examine tthhee svperignegtaftoilolnowionng tthheefirBeur(2n0e0d8,), Fwiereslaimnepleidn
the relationship ofdisturbance events to patterns
association with a previous study that compared
ofnative and nonnative dominance by asking the
dqluaoensdntoicnoonnmas:tmiuv(n1ei)tspDyeocreieescs,otvhCe.er minenaltiaitvebenusricnshaeipdnarfpriaarreltl/iigcnreu,alsaosr,r- nt20ha1et3id)v.eenaWsinetdysiaannvmadpslidevodemtrihaninsagnescistee(aoMgfoarCio.nnmieenyltiahteenndsspiRrsuiinnngdeiotlfs
persist and dominate?, and; (2) Does the post-fire 2012. The second site (Burned, No Fireline),
community differ in sites with and without sampled in the spring of2012, was located along
firelines, and do those sites differ from an the same ridge immediately adjacent to the
undisturbed site? This was an opportunistic study Burned, Fireline. This site burned in the Sedg-
with no replication of sites, so the results should wick Fire but was not disturbed by bulldozing.
be interpreted with this limitation in mind. The third site (No Burned, No Fireline), also
m
sampledin the springof2012, was 300 southof
Methods the Burned, Fireline on the adjacent and parallel
ridge, with a similar elevation, slope, and aspect.
Study Species This site was not burned or disturbed in the
Sedgwick Fire.
Centaurea melitensis is an annual thistle that is Thmemtwo survmeymyears, 2008 and 2012, had
native to the western Mediterranean Basin, but 739 and 577 ofprecipitation, respectively,
has been dispersed by humans and is invasive during the rainy season (Lisque weather station,
globally. It is a problem pest in California Sedgwick Reserve, 34.72449N, -120.0635W).
because it threatens the health of livestock and
the persistence of native plants and animals
(DiTomaso and Gerlach 2000; Moroney et al. Sampling
2011). It has been in California sinceat least 1797
(Hendry 1931). Centaurea melitensis is one ofthe In June 2008, sampling was conducted in the
most common nonnative plants in the first five Burned, Fireline site. In June 2012, the same site
years after fire in chaparral and coastal sage was resampled, and the two additional sites were
scrub of southern California, with an average also sampled usingtwo-stage systematic sampling
density of >16,000 individuals ha-1 in chaparral (Elzinga et al. 1998). Ten transects were placed at
m
sites and >285,000 individuals ha~‘ in sage scrub randomly selected points within each 10
m
sites (Keeley et al. 2005). increment of a 100 baseline that followed the
ridgeline. A series of 1 m X 0.2 m plots were
Study Sites placed at regular intervals along the transects
starting at a randomly selected point. To
The Sedgwick Reserve, University of Califor- determine population density of C. melitensis
We,
nia Natural Reserve System (34 42'47.7"N, individuals-were counted within each plot.
120°02'00.7"W) contains a mix of vegetation estimated the percent cover of each of the
types including chaparral, coastal sage scrub, following groups: C. melitensis nonnative annual
,
native and nonnative grasslands, and valley oak grasses, all nonnative species (including C.
savanna. The recorded fire history for the melitensis and annual grasses), all native species,
Reserve begins in 1912, and there have been no litter, bare ground, and rock. Plot totals were
fires recorded on the Reserve until the 2007 averaged per transect. All taxa present in the
Sedgwick Fire (S. Alderete, Santa Barbara plots were recorded.
2013] MORONEY AND RUNDEL: FIRE AND FIRELINE DISTURBANCE IN A RESERVE 175
TSiatbe.leAF1. =TAhnenuParlesfeornbc,eSo=f TSharxuab,inGth=eGSeoapmhpyltee,PlPoGts=RPeeproenrntiealdgarsasts,heAPGer=ceAnntnuaaglegorfassP.lots in Each
Percentage ofplots
2008 2012 2012 2012
Life- Burned, Burned, Burned, No Burned,
Species form Fireline Fireline No Fireline No Fireline
Native
Adenostomafasciculatum Hook. & Arn. S 20 5 0 0
Amsinckia sp. AF 0 0 5 15
Artemisia californica Less. S 10 20 0 0
G
Calochortus sp. 25 0 0 0
Ceanothus sp. S 35 5 0 0
Daucuspusillus Michx. AF 0 0 10 0
Dichelostemma capitatum (Benth.) Alph. Wood G 10 0 0 0
AF
Galium sp. 5 0 0 0
Hazardia squarrosa (Hook. & Arn.) Greene S 0 0 0 20
Deinandrafasciculata (DC.) Greene AF 0 0 25 0
Lupinus sp. AF 0 0 30 0
Navarretia sp. AF 0 0 30 0
Plantago erecta E. Morris AF 0 0 10 0
Salvia mellifera Greene S 0 5 0 0
Sisyrinchium bellum S. Watson G 0 0 5 0
Stipa sp. PG 15 40 5 15
Native species richness 7 5 8 3
Nonnative
Anagallis arvensis L. AF 70 35 25 0
AG
Avena sp. 0 15 100 100
AG
Bromus sp. 5 95 85 90
Carduuspycnocephalus L. AF 0 5 0 10
Centaurea melitensis L. AF 95 85 85 30
Erodium sp. AF 35 45 60 30
Hordeum murinum L. AG 0 25 25 30
Lactuca serriola L. AF 0 0 5 25
Nonnative species richness 4 7 7 7
Total species richness 11 12 15 10
Statistical Analyses regressions were followed bypost hoc testing using
multiplecomparisons with bonferroni corrections
To compare differences in the Burned, Fireline to test for the differences in predicted means
site between the samplingyears 2008 and 2012, we betweenpairs ofsites. All statistical analyseswere
performedamultivariateanalysisusingHotelling’s done using Stata statistical software (Stata,
T2 test on the followingvariables: relative cover of version 12.1, Statacorp, College Station, TX).
C. melitensis, relative cover of annual grass,
relative cover of native plants, relative native Results
species richness, percent cover of rock, percent
cover oflitter, and percent cover ofbare ground. A total of24 species were recorded in the plots
The multivariate test was followed by separate over all sites and years collectively, including 16
linearregressionsforeachvariable. Relativecover native species and eight nonnative species. Native
was calculated as the percent cover of the target life forms included seven annual forb species, five
group divided by the total vegetative cover ofall shrub species, three geophytes, and one perennial
species in a plot. These variables were arcsine- grass. Therewereno native annual grasses. All ofthe
square root transformed to stabilize the variance nonnative plants were annuals, with five annual forb
and then back-transformed to proportions for species and three annual grass species (Table 1).
interpretation. The difference in C. melitensis Centaurea melitensis was the most common
density between years was analyzed with a species in all of the sites collectively (74% of the
Poisson regression with robust standard errors. plots).
To compare differences in the three sitesin 2012
(Burned, Fireline, Burned, No Fireline, and No Fireline Between Years
Burned, No Fireline), we used the same analytical
MANOVA
approach as above, substituting for There was a significant difference in the
Hotelling’sT2inthemultivariateanalysis.Thelinear relative cover composition of the community
MADRONO
176 [Vol. 60
Table 2. Two Poisson Regression Models with
Robust Standard Errors Comparing Centaurea
MELITENSIS DENSITY WITHIN THE BURNED, FlRELINE
BetweenYearsand BetweenYears Among Sites.
2008 is the reference year in the first model, and the
Burned, Fireline is the reference site in the second
model. *P < 0.001.
Source Coefficient X2 P
Burned, Fireline between years
Year 0.677 3.881 0.049
Constant 2.5 107.537 *
Sites within year
Site 2 0.347 0.846 0.356
Site 3 -2.736 15.761 *
Constant 3.174 170.825 *
2wi0t0h8in(2t-hgeroBuuprnHeodt,elFliirnegl’isneTi2n, 2F01=2 c1o4m.4p9a5r,ePd t=o fFoigu.r s1i.tesDaetnstihteySbeodxgwplioctks RoefsCe.rvmee.liTtehnesitswmoeabsouxreesdoinn
0.0001). This difference was due to a decrease in the left represent sites within a bulldozed fireline, with
the relative cover of C. melitensis, an increase in datameasured in 2008 and 2012. The two boxes on the
the relativecoverofannual grasses, an increasein right represent two sitesadjacent to thefireline, butnot
the relativecoverofnative plants, and an increase bulldozed, one burned site and one unburned site, with
in the percent cover of litter (Table 2). The data measured in 2012.
density of C. melitensis increased despite the
decrease in relative cover (Fig. 1). There was no All ofthe native species in the Burned, Fireline
change in native species richness. were perennials, with four shrub species and one
The number of plots with native shrubs, perennial grass species. The native species in the
geophytes, and annual forbs decreased in four Burned, No Fireline site were mostly annual
years, but the number of plots with native forbs, with one perennial grass, one geophyte,
perennial grasses increased (Table 1). In 2008, and no shrubs. In the No Burned, No Fireline
the most common natives were shrub seedlings, site, the natives included one annual forb species,
but the most common group in 2012 was one shrub species, and one perennial grass
perennial grasses. All of the shrub species (Table 1).
recorded in 2008 were still present in 2012 and
the number of shrub species found within the
Discussion
Burned, Fireline increased from three to four.
Overall, native species recovery was positive with Four years after the disturbances of fire and
an increase in native species relative cover and no bulldozing, the mean relative cover of Centaurea
significant change in native species richness. melitensis decreased significantly, from 72% to
28%, within the Burned, Fireline. In contrast, the
Sites Within Year cover of annual grasses increased in four years
from almost zero to a mean cover of 6%. Litter
The three sites (Burned, Fireline, Burned, No also increased significantly, from a mean cover of
Fireline, and No Burned, No Fireline) differed 10% in 2008 to 29% in 2012. Despite the decrease
significantly in community relative cover compo- in relative cover of C. melitensis the density
sition (Pillai’s trace; F2,27 = 44; P < 0.0001). The remained the same. The same, number of
No Burned, No Fireline site had lower relative individuals germinated and survived, but they
cover and density of C. melitensis and higher were smaller in size. This could have been due to
relative cover of annual grasses than both the the reduction in water availability in 2012
Burned, Fireline site and the Burned, No Fireline compared to 2008, or to competition with annual
site (Table 3, Fig. 1). The Burned, Fireline had grasses that were not present in 2008. This
higher percent cover of both rock and bare suggests that even after severe disturbances such
ground than either of the other two sites. The as fire and bulldozing that reduce the cover of
percent cover of litter was significantly higher in annual grasses, these grasses can quickly regain
the Burned, No Fireline site than the Burned, dominance and displace annual forbs.
Fireline site, but not different from the No The differences in the three sites in 2012 were
Burned, No Fireline site (Table 3). There was most dramatic in the relativecoverand frequency
no difference among the sites in relative cover of of C. melitensis. While both of the Burned sites
natives or in native species richness. had more than 25% cover and 85% frequency of
2013] MORONEY AND RUNDEL: FIRE AND FIRELINE DISTURBANCE IN A RESERVE 177
Table 3. Seven Linear Regression Models for melitensis. This site also had the lowest nonnative
the Effectof Year Withinthe Burned, Fireline relative cover and the highest native relative
on Each of the Following Variables: C. cover. This may be related to the depth of the
MMEELLIITTEENNSSIISS, DAENNNSIUTAYL, GTRHAESSERSE,LANTOINVNEATCIOVVEERPLOAFNTSC,. seed banks of annual grasses and C. melitensis.
Native Plants, Relative Native Richness, and Smaller seeds are generally shallower in the soil
the Percent Cover of Litter, Rock, and Bare than heavier seeded species, and thus more
Ground. Thecoefficientsarepredictedvalues foreach vulnerable to mortality from fire (Bond et al.
year. Theconstantistheintercept. The referenceyearis 1999). If heavier, more compact C. melitensis
2008. *P <0.001. seeds are buried deeper in the soil, while lighter
grass seeds stay nearer the soil surface, then both
Source Coefficient t P the fire intensity and the depth of the bulldozer
Relative cover of C. melitensis blade might have been factors in the reduction of
Year -0.547 -5.86 * annual grasses and the persistence of C. meliten-
Constant 1.066 16.13 * sis. High intensity, warm-season fires can kill
Relative cover ofannual grass annual grass seeds on the surface of the soil and
Year 0.5 8.35 * increase the cover of native species (Meyer and
Constant 0.017 0.41 0.689 Schiffman 1999). Furthermore, the bulldozer
might have cleared surface seeds, exposing the
Relative cover ofnative plants deeper C. melitensis seeds to the surface. Alter-
Year 0.347 2.84 0.011 natively, clearing the litter may have been the
Constant 0.228 2.65 0.016 more important effect of bulldozing and fire.
Relative native richness With barriers to germination removed, C. meli-
Year -0.018 -0.14 0.889 tensis and native seeds in the seed bank would
Constant 0.463 5.13 * have had an opportunity to recruit. The distur-
Percent cover ofrock bances of fire and bulldozing might reduce
Year -0.082 —2 0.061 annual grasses and recover forbs and shrubs in
Constant 0.185 6.38 * the short term as long as the seed bank is deep
enough and remains intact.
PerYceeanrt cover oflitter 0.235 3.59 0.002 Woody plant canopy closure (i.e., native
Constant 0.318 6.87 * shrubs) has been shown to be the most important
direct factor in explaining alien plant dominance
Percent cover ofbare ground in southern California chaparral and sage scrub
Year -0.05 -0.51 0.618 sites within five years after fire (Keeley et al.
Constant 0.667 9.54 * 2005). The only sitewith substantial woody plant
recruits four years after fire was the Burned,
Fireline, with no shrubs in the Burned, No
C. melitensis, the 2012 No Burned, No Fireline Fireline site, and the No Burned, No Fireline site
site had only 5% cover and 30% frequency. This supporting only a few small individuals of one
suggests that the disturbance caused by fire, shrubspecies, Hazardiasquarrosa(Hook. &Arn.)
regardless ofthe additional clearing by bulldozer, Greene. In 2008, the shrubs recorded in the
opens colonization sites sufficiently for C. meli- Burned, Firelinewere seedlings, asthesoil surface
tensis to establish. The increased frequency and hadbeengraded, removingallmature shrubs that
cover of annual grasses in the Burned, Fireline might havebeen present. Fouryears afterthe fire,
siteafterfouryears suggeststhat propagulesfrom shrubs growing in the Burned, Fireline were still
nearby unburned patches of annual grass colo- relatively small, and annual plant coverwashigh,
nize cleared sites overtime. The lowercover ofC. suggesting that none ofthe shrubs in the Burned,
melitensisin theNo Burned, No Fireline site may Fireline were large enough to close the canopy
be linked to the increase in annual grass cover sufficiently to shade out annual plants.
and associated litter over time, as their relative Fire frequency is an important determinant of
covers seem to have a somewhat inverse relation- the relative success of native versus nonnative
ship. Once seeds are present, annual grasses species in chaparral. In sites that burn at
germinate and grow tall earlier in the season intermediate fire frequencies, total species diversi-
than C. melitensis, possiblyblockingout light and ty is typically highest in the first few years
preempting germination potential. Litter accu- following fire (Keeley and Fotheringham 2003).
mulation may also suppress germination by In sites that have burned at high frequency,
limiting light and changing the temperature and nonnative annuals dominate after fire, but in sites
moisture availability on the soil surface (Carson that have not burned for several decades, native
and Peterson 1990). annualsdominateafterfire(HaidingerandKeeley
Ofthe three sites sampled in 2012, the Burned, 1993). At Sedgwick Reserve fire has been absent
Fireline had the lowest relative cover of annual for atleast 100 yr. However, in the first fouryears
grasses and the highest relative cover of C. afterafire, nonnativeannualsdominated. Perhaps
MADRONO
178 [Yol. 60
the disturbance from past grazing has been a Elzinga, C. L., D. W. Salzer, and J. W. Wil-
factor in determining the present composition of loughby. 1998. Measuring and monitoring plant
the community and the high relative cover of populations. U.S. Department of the Interior,
nonnative species. Due to sampling design limita- BureauofLand Management, TechnicalReference
tions, ourresultsmustbeinterpretedwithcaution. 1730-1, Denver, CO.
However, our data can be useful to management Haidinger, T. L. and J. E. Keeley. 1993. Role of
of disturbed areas. The results indicate that in chhiagpharrfailre. Mfardeqruoennocy40:i1n41d-e1s4t7r.uction of mixed
these sites, fire promoted native species diversity, Hendry, G. W. 1931. The adobe brick as a historical
and the Burned, Fireline had the most native source: reporting further studies in adobe brick
perennials. Multi-year monitoring ofthe commu- analysis. Agricultural History 5:110-127.
nity is important to assess the fate of early Hobbs, R. J. andL. F. Huenneke. 1992. Disturbance,
colonizers after such disturbances. The results of diversity, and invasion: implications for conserva-
our study suggest that the recruitment of C. tion. Conservation Biology 6:324—337.
melitensis, along with some native species, is Keeley, J. E. 2002. Native American impacts on fire
promoted by fire. Over time, its abundance is regimes oftheCaliforniacoastalranges. Journalof
limited by competition, not from woody native Biogeography 29:303-320.
species, but from another group of nonnative and T. J. Brennan. 2012. Fire-driven alien
invaders, Mediterranean annual grasses. invasion in a fire-adapted ecosystem. Oecologia
169:1043-1052.
ACKNOWLEDGMENTS and C. J. Fotheringham. 2001. Historic fire
regime in Southern California shrublands. Conser-
WethankRodneyJ. Masonforfieldassistance, Kate vation Biology 15:1536-1548.
McCurdy ofthe Sedgwick Reserve, and Scot Alderete and 2003. Species-area relationships in
.
ofthe Santa Barbara County Fire Department. Mediterranean-climate plant communities. Journal
ofBiogeography 30:1629-1657.
Literature Cited , M. Baer-Keeley, andC. J. Fotheringham.
Bond, W. J., M. Honig, and K. E. Maze. 1999. Seed 2005. Alien plant dynamics following fire in
size and seedling emergence: an allometric rela- Mediterranean-climateCaliforniashrublands. Eco-
tionship and some ecological implications. Oeco- logical Applications 15:2109-2125.
logia 120:132-136. Merriam, K. E., J. E. Keeley, and J. L. Beyers.
Carson, W. P. and C. J. Peterson. 1990. The role of 2006. Fuel breaks affect nonnative species abun-
litter in an old-field community: impact of litter danceinCalifornianplantcommunities. Ecological
quantity in different seasons on plant species Applications 16:515-527.
richness and abundance. Oecologia 85:8-13. Meyer, M. D. and P. M. Schiffman. 1999. Fire
D’Antonio, C. M., T. L. Dudley, and M. MACk. season and mulch reduction in a California
1999. Disturbance and biological invasions: Direct grassland: a comparison of restoration strategies.
effects andfeedbacks. Pp. 413^452 in L. R. Walker Madrono 46:25-37.
(ed.), Ecosystemsoftheworld,Vol. 16. Ecosystems Moroney,J. R.andP.W. Rundel. 2013.Abundance
of disturbed ground. Elsevier Science B.V., Am- and dispersion of the invasive Mediterranean
sterdam. annual, Centaurea melitensis in its native and
Ditomaso, J. M. and J. D. Gerlach, Jr. 2000. non-native ranges. Biological Invasions 15:495-
Centaurea solstitialis L. Pp. 101-106 in C. C. 507.
Bossard, J. M. Randall, and M. C. Hoshovsky, , P. M. Schiffman, andC. A. Brigham. 2011.
(eds.). Invasive plants of California’s wildlands. Invasive European annual plants impact a rare
University ofCalifornia Press, Berkeley, CA. endemic sunflower. Madrono 58:69-77.