Ecologicalnicheandspeciestraits:keydrivers ofregional plant invader assemblages
Wilfried Thuiller • Nu´riaGasso´ • Joan Pino•
MontserratVila`
Abstract Linkingspeciestraitstonichepropertiesis fundamentaltounderstand thespatialstructureof invasivespeciesassemblagesandtheinvasionprocess itself.Usinginformation on74invasivespeciesin Spain, theaimsofthispaperareto(1)testwhether invasiveplantspeciesassemblages followanested patternattheregionalscale,(2)inspecttherelationship betweenrangesizeandnicheproperties(positionand breadth)ofinvasivespeciesto testwhetherthe nested patternisaproductofspeciesnicheoverlap;andfinally (3)examinehowspeciestraitsofinvadersarerelatedto theirnicheproperties.Weshowthatregionalinvasive plantspeciesassemblagesinSpainareorganizedin nestedsubsets.Invasivespecieswithrestrictedrange
W.Thuiller()
Laboratoired’EcologieAlpine,CNRS-UMR5553, Universite´JosephFourier,BP53,38041
GrenobleCedex9,France
e-mail:
N.Gasso´ J.Pino
CREAF(CentreforEcologicalResearchandForestry
Applications),UniversitatAuto`nomadeBarcelona,
08193Bellaterra,Catalonia,Spain e-mail:
J.Pino
e-mail:
M.Vila`
Estacio´nBiolo´gicadeDon˜ana-CentroSuperiorde
InvestigacionesCient´ıficas(EBD-CSIC),Avda.Ame´rico
Vespucios/n,IsladelaCartuja,41092Sevilla,Spain e-mail:
occurinareasinvadedbywidespreadinvaders.By relatingnestednessmetricstospecies’nicheproperties frommultivariateanalyses,wefound thatthese restrictedinvaders arelesstolerantofbroadclimatic andlandscapeconditionsthanwidespreadinvaders. Finally, regarding theassociationbetween niche properties andspeciestraits,wefoundthatspecies withlargeenvironmental nichebreadthcommonly exhibitnonN-fixing strategy,short-lifespan,and clonalgrowth, whilethosewithnichepositionin anthropogeniccoastalareaswereperennialandclonal speciesofunintentionalandagriculturalorigin.Using anintegrativeapproach linkingtheregionalspatial structureof invasiveplantassemblages,speciesniche properties andspeciestraits,wewereabletounder- standthepotentialcausesofinvasivespeciesdistribu- tioninSpain.Theapproachdevelopedinthisresearch couldbeeasilyappliedtootherareastodisentanglethe mechanismsdrivinginvasivespeciesdistributions.
KeywordsAlienplants Clonalgrowth Disturbance Invasiveness Inference-basedmodel Life-historytraits Nestedness Nicheposition Nichebreadth Rangesize
Introduction
Lesseningtheecologicalandsocio-economicimpactsof invasivespeciesrequiresacomprehensiveunderstanding
ofthe factorsdrivinginvasions(Mack1996;Pysˇek etal.2010).Invasionecologyhasmainlyfocusedon thefactorsthatmakeeitheraspecies aninvader (Hamiltonetal.2005;Rejma´nek1995)orahabitat pronetoinvasion(Pauchardetal.2004;Pysˇek2004; Stohlgren etal.2005).However, thesetwoapproaches havebeenkeptseparateuntilrecently,when such species havebeguntobeconsidered ashavingasetof intrinsic and extrinsic traits that result invasiveness (i.e.potentialtoinvade) incertainenvironmental conditions (Jiangetal.2010;SaxandBrown2000; SheaandChesson 2002;Thuilleretal.2010b).This integrated approachisfundamentalto understanding thecomposition ofinvasivespeciesassemblages, becausebothspeciesproperties andrecipienthabitat characteristicsareknown todeterminethelevelof invasioninhabitatsandregions.
Fromabiogeographic perspective,onemight wonderwhether certaininvadershaveverygeneral requirements explainingtheirlargedistributional rangesandoverlaps, orwhethertheyneedspecific environmentalconditions.Thisquestionhasseldom beenexplored.Inageneralcontext,afirststepwould betoinvestigateifspeciesassemblages arenestedat theregional scale.Inotherwords, arespecies present inspecies-poor sitesonlysubsetsofthespeciesfrom species-richsites?Inthiscase,assemblageswithfew species tendtobemadeupofthemostfrequently occurringspecies,whiletheleastfrequentlyoccurring speciesarenotrestrictedtospecificsitesbutoccurin species richareas.Thispatternhasbeenmostly investigatedfornativeanimalcommunities(Wright etal.1998),andwhether such anestedpatternisalso foundininvasiveplantspeciesassemblages is unknown.
In the caseofinvasiveplants,speciesdistribu- tion could be the outcome of a combination of factors such as minimum residence time (Ahern etal.2010; Gasso´etal.2010; Phillips etal.2010; Wilsonetal.2007),propagulepressure(Lockwood et al. 2005; Pysˇek et al. 2010) and species ecologicalrequirements asrepresentedbytheir realizedecological niche(Thuilleret al. 2005; Va´zquez2006;Jime´nez-Valverde etal.2011). Althoughseveralstudieshaveappliednichetheory toinvasion biology(Fitzpatrick andWeltzin2005; Gallagheret al. 2010;Roura-Pascualet al. 2009), little effort has been made to establish a link between the physiological, morphological and
reproductive attributesofspeciesandtheirniche propertieslikenicheposition(e.g.averageposition on a given gradient) and niche breadth (e.g. standarddeviationaroundthe averagepositionon a given gradient) (but see Ku¨hn et al. 2006; Niinemets 2001;Thuilleretal.2010a).Theinves- tigationofthistopicwillenlargeourunderstanding ontherelationshipbetweeninvaderrangesize(i.e. thesizeofthedistribution area)andspeciestraits (Gasso´etal.2009;Lloretetal.2005).
Weargueherethattherangesizeofaninvader is relatedtoitsniche(Brown1995;Gaston andBlack- burn2000).Abroaderniche,forinstance, might enablethespeciestobecomemoresuccessful andto reachlargerangesizes(Va´zquez2006). Ourhypoth- esisisthatanested patternofspecies assemblages couldbetheresultofanicheoverlap among rareand widespreadinvasivespecies(GastonandBlackburn
2000),anddifferencesamongnichesmightbe,inturn, aconsequenceofdifferencesbetweenspeciestraits (Albertetal. 2010;Thuiller et al.2004;Va´zquez
2006).
Weexaminedthesizeofinvader plantspecies’ rangesacrossmainlandSpain, theirtraitsandthe environmental characteristicsofinvadedareasin orderto(1)testwhetherinvasive plantspecies assemblages attheregionalscalefollowanested pattern,(2)inspecttherelationshipbetweenrangesize andnichepositionandbreadthtotestwhether the nestedpatternisaproduct ofspeciesnicheoverlap; andfinally(3)examineifcertainspeciestraitsare relatedtonicheposition ornichebreadth,and therefore,arerelatedtoaspecies’ potentialof becomingwidespread.
Methods
Studyregion
SpainislocatedintheSWofEuropeandoccupies
493,486km2.Theregionhasalonghistoryofplant speciesintroductions,enoughformanyspeciestobe
distributed acrossawiderangeofenvironmental conditions(Sanz-Elorzaetal.2004). Duetoits geographic,topographic,climaticandgeologicheter- ogeneity,Spainhasahighlandscape andhabitat diversity (Ninyerola et al. 2000; Valladareset al.
2005),includingarid,Mediterranean,temperateand alpineecosystems,providingabroadenvironmental rangeforplantinvasions(Gasso´etal.2009).
Table1 Intrinsic and extrinsic species traits taken into accounttoexplaininvasive plantspeciesnicheposition and breadthinSpain
Dataset
Intrinsictraits
CategoryCode
Invasivespeciesandtraits
Dataonthedistribution ofinvasiveneophytes(i.e. established aliensintroducedafter1500)were extractedfromtheAtlasofInvasive PlantSpecies in Spain(Sanz-Elorzaetal.2004).Invasive speciesare definedfollowingPysˇeketal.(2004):alienspecies withself-sustainablepopulations withoutdirect humaninterventionthatproducesoffspringatconsid- erabledistancesfromtheparentplants,and thushave thepotentialtospread overlargeareas.TheAtlas containspresence recordsforover100species ata resolution of 10km UTM (Universal Transverse Mercator)grid,butwerestrictedtheanalysesto74 speciesfoundatleastin10UTMgridcellsinmainland Spain.Thiswasmadetoavoidsporadicspecies with relativelyunknowndistributionsand tomakesure we hadenoughinformation relatedtoenvironmental descriptors. Duetothepotentialheterogeneityof samplingeffortacrossthecountryandtheconsequent difficulty ofdistinguishingbetweentheabsenceof invasive speciesandmissingdata,theanalyseswere alsolimitedtoUTM cellswithatleastoneinvasive speciesrecorded. Inordertohavecomparablepixels withsimilarlandsurface, wealsoexcludedcoastal UTMcellswithalandproportionlowerthan60 %. Thefinal numberofUTMcellsincludedinthe analyseswas2401.
Foreachspecieswecalculateditsrangesizeas the numberofoccupiedUTMcells(Gasso´etal.2010)and exploredalistofintrinsic (i.e.seedsize,clonal growth,longevity,life-form,dispersalandpollination syndromes,andnitrogenfixingcapacity)andextrinsic traits(i.e.regionoforiginandpathway ofintroduc- tion)thatarecommonly associatedtoinvasiveness (PysˇekandRichardson 2007;Thuilleretal.2006) (Table1).
Residencetimewasestimatedastheyearoffirst bibliographicorherbariumrecordinSpain.Allthe aboveinformationwastakenfromCastroviejo etal. (1986–2000),Casasayas(1990),Bolo´setal.(1993) andSanz-Elorzaetal.(2004).SeeTable6inAppendix forthetraitvaluesofall74species.
SeedsizeVerylarge([10mm)5
Large(5–10mm)4
Medium(2–5mm)3
Small(1–2mm)2
Verysmall(\1mm)1
N-fixingYesYes
NoNo
ClonalgrowthYesYes
NoNo
LongevityAnnualAnnual BiennialBiennial PerennialPerennial
Life-formChamaephyte Ch GeophyteGe HemicryptophyteHe MacrophanerophyteMc Mesophanerophyte Ms Nanophanerophyte Nn Therophyte Th Vine Vi
DispersalsyndromeAnemochoryWind ZoochoryAnimal HydrochoryWater
PollinationtypeAnemophilyWind
ZoophilyAnimal
Extrinsictraits
RegionoforiginAmericaAM SouthAmericaSM NorthAmericaNM AsiaA EuropeE AfricaSF Oceania O
Pathwayofintroduction Agriculture A GardeningG SilvicultureS WeedW UnintentionalU
Thedifferentcategoriesforeachtraitandthecodeusedin
Fig.4andTable6inAppendixareshown
Environmentaldata
Modelsforinvaderspecies’realizednicheswerebased oninformation fromasetofnineteenenvironmental predictors groupedunderland cover,topographicand climaticvariables(Table2)gatheredfromdifferent sources andaveraged at10kmUTMgridresolution. Theproportion ofmainlandcovercategories(i.e. built-up areas,agriculturalareas,forests, scruband herbaceous vegetation,openspaces,wetlandsand waterbodies)werecalculatedfromtheCORINELand CoverMapofSpain( coverdiversityperUTMcellwasestimatedbycalcu- latingtheShannonindexoftheabove-mentionedland covercategories(Shannon1948).
Roadlength(m)andrailway length(m)were calculated from maps obtained from the official serveroftheSpanishMinistry( Weconsidered thesehumansettlementvariablesasa measureofhumanfootprint(Sandersonetal.2002) andassurrogates forpropagulepressure (Pysˇek etal.
2002).Mean altitude(m)andaltituderange(m)were calculatedfromthenationalDigitalElevationsModel (DEM) at 100m resolution ( uab.es). Meandistancetothecoastline(m)wascal- culatedusingMiraMon distance algorithms.
Finally,sixclimaticvariableswerecalculatedfrom theDigitalClimaticAtlasofSpain(Ninyerolaetal.
2005; agridsetof200m9200mpixelsizegeneratedby modellingfromtheSpanishnetwork ofmeteorologi- calstations.Wecalculatedthemeanvaluespersiteof annualtemperature(°C), minimumwintertempera- ture(°C),temperaturerange(i.e.difference between maximum temperatureinJulyandminimaltempera- tureinJanuary),annualrainfall(mm),summerrainfall (sumoftherainfallinJune,JulyandAugust) and annualsolarradiation(KJm-2 day-1 lm-1).
Nestednessinspeciesassemblages
Weusedanestednessindextoestimatethedegreeof associationamonginvasivespecies’occurrences in UTMcells.Thisprocedure aimedatevaluating whetherspeciescomposition inagivensitewas random(lownestedness)orspecies-poorsiteswere madeupbythemostfrequent invadersfromthe regionalpool(highnestedness). Theminimum requirementfornestednessisthatsitesdifferintheir
suitability, andspeciesdifferintheirabilitiesto colonizeandpersistinthem(RytiandGilpin1987).
Tovalidatetheexistenceofanested structure amongspeciesandsites,weestimatedthenestedness index (N)as N=(100-T)/100,where T is the so-called ‘Temperature’index,whichmeasures the site-occurrence matrixdisorderwithvaluesranging from0°to100°(AtmarandPatterson1993).Because wewereinterestedinemphasizingorderednestedness ratherthandisordered, wecalculatedthelevelofN, withvaluesrangingfrom0(minimumN) to1 (maximumN).Nwascalculatedusingthe Nestedness CalculatorSoftware(AtmarandPatterson 1993; Wrightetal.1998),whichreorders therowsand columns ofthematrixtominimizetheunexpected occurrences,andpackthematrixintoastateofmax- imalnestedness (i.e.sitesindescending orderof species richness and species in decreasing order ofrangeextension).Then,Tiscalculatedasthesum ofsquareddeviations fromtheisoclineofperfect nestedness ofunexpectedpresencesandabsences, dividedbythemaximumvaluepossibleforthematrix, multipliedby100.ToassessthesignificanceofTwe comparedtheobservedvalueofTwithabenchmarkof
500randommatricesprovided byanullmodelin whicheachcellinthematrix hasthesameprobability ofbeingoccupied(GastonandBlackburn2000).This isanefficient waytotestiftheobservednestedness differedfromarandompattern(Wright etal.1998).
Species’nicheproperties
Toseparate invaderspeciesnichesbasedonspecies distribution andtheenvironmentalvariablesselected (Table2),weusedtheoutlying meanindex analysis (OMI,hereafter),amethodproposedbyDole´decetal. (2000) andimplementedintheade4library(Chessel etal.2004)inR(RDevelopment CoreTeam2012). TheOMImakesnoassumption abouttheshapeof curvesofspeciesresponsetotheenvironment,and givesequalweighttospecies-rich andspecies-poor sites.Thismethod determinesnichepositionasthe meanlocationofthespecies intheenvironmental space.Therefore,itmeasuresthepropensity ofthe speciestoselectamarginalenvironment (asopposed toaverageorcommonenvironment wherethemean position iscloseto0).Nichebreadthiscalculatedas thevariability(SD)oftheenvironment usedbyeach species along the main axes calculated from the
Table2 Environmental
VariablesDatasource
predictorsanddatasources
usedtomodelinvasive
plantspeciesnicheposition andbreadthinSpain
Landscape
Percentageofbuilt-upareasCORINELandCoverMapofSpain
Percentageofagriculturalareas
Percentageofforests
Percentageofscrubandherbaceousassociations
Percentageofopenspaces Percentageofwetlands Percentageofwaterbodies
Landcoverdiversity(Shannonindex)
(
Roadslength(m)OfficialserveroftheSpanishMinistry
Railwaylength(m)
Topography
(
Meanaltitude(m)DigitalElevationModel(DEM)
Altituderange(maximum-minimum;m) Meandistancetothecoastline(m) Climate
(
Meanannualtemperature(°C)DigitalClimaticAtlasofSpain
Temperaturerange(maxJuly–minJanuary;°C) Minimumwintertemperature(°C)
Annualrainfall(mm) Summerrainfall(mm)
Meanannualradiation(KJ/m2 daylm)
(
index.htm)
environmentalvariablesincludedintheOMI(Thuiller etal.2004).
Species’nichepropertiesasapredictorofrangesize
The nicheconceptstatesathree dimensionalrelation- shipbetweennicheposition,nichebreadthandspecies abundanceorrangesizeinanassemblage(Shugart andPatten1972).Wemodelledrangesize(i.e.number ofUTM where eachspecies ispresent,logtrans- formed) asafunctionofnichepositionandniche breadthonthetwomaingradients selectedbythe OMI.
Becausespeciesarelinkedbyphylogeny (Harvey andPagel1991),usingspeciesasindependentdata pointsmayinflatethedegreesoffreedom(Blomberg etal.2003;Felsenstein1985)andincreasetheType-I error.Wethusbuiltaphylogenetic treeforour74 species.Thetopologyofthesupertreewasconstructed withphylomatic ( phylomatic/phylomatic.html, Webb et al. 2008).
Branchlengthsweresetbyassigning anagetostem familynodesbasedonWikstro¨metal.(2001),and interpolatingtheremainingbranchlengthsoftheother nodesusingthebranchlengthadjustment (bladj) algorithminPhylocom(Webbetal.2008).
Totesthowspecies rangemightbeexplainedby nicheproperties andminimumresidencetime,we usedgeneralizedleastsquaresmodels (GLS) control- lingforthepotentialcorrelationbetween species associated withphylogenetichistory. Thevariance explainedbytheGLSmodel wasestimatedby extractedtheR2 valuefromtheregressionbetween observedspeciesrangeandpredictedonesfromthe GLSmodel.Weused theglsfunctionwithinpackage nlmeinR(RDevelopmentCoreTeam2012).
Speciestraitsexplainingspeciesnicheproperties
Werelatedspecies’nicheproperties (position and breadth) onthetwomainnicheaxestospeciestraits withthesameapproachthanabove;i.e.generalized
leastsquares model(GLS) controlledbyphylogeny. Tomeasuretheactualpowerofeachspeciestraitover nichepositionandnichebreadth, weusedamulti- modelinferenceapproach(BurnhamandAnderson
2002;LinkandBarker 2006)onselectingall-subsets oftheGLSusingtheAICc(Akaike information criterion,Akaike1974)measure. Theweightof evidence(wpi)ofeachspecies traitasapredictorof nicheposition andnichebreadth canbesimply estimatedasthesumofthemodelAICweights over allmodelsinwhichtheselectedpredictorappeared (BrookandBradshaw 2006;Carbonietal.2010; Thuilleretal.2007).Thepredictorwiththehighestwpi (theclosestto1)getsthehighest weightofevidence (i.e.hasthehighestrelativeimportance)toexplainthe response variable.Thisapproachbasedonasetof multiplemodels isfarmorerobustthaninferring variableimportance basedonasinglestepwise selectedmodel(BurnhamandAnderson2002;Link andBarker2006).Weusedthedredgefunctionwithin thepackage MuMIn inRtorunthisoverallanalysis.
Results
Nestednessinspeciesassemblages
Site-occurrencematrixtemperaturewas4.85°,which gavealevelofnestedness(N) of0.952.After comparingthisvaluewithabenchmarkof500random matrices,withanaveragedTof29.9°±0.24(i.e.a meannestednessof 0.70),we foundthattheprobabil- ityofarandomreplicatebeingequallyormorenested thanourstudymatrixwaslessthan0.0001.Therewas thusahighlevelofnestedness inthesite-occurrence matrix,indicatingthatthespeciespresentinsiteswith lowinvasiveplantrichness wereasubsetofspecies alsooccurringinrichersites.
Speciesnichesproperties
ThefirsttwoaxesoftheOMIsuccessfullyexplained
45and32%respectivelyofthe totalvariability when separatinginvasivespeciesniches(Fig. 1).Thefirst axis(anthropizationhereafter)summarisedagradient ofhumantransformation. Thisgradientopposed highlybuilt-upareasinflatlands (i.e.dense coverage ofroadsandrailways) tohighaltitudeareasfarfrom thecoastandwitharelativelylowurbanpressure.The
second axis(climate-landscapehereafter)opposed forested tocroplandareas.Forestedareaswere associated withcoldandhumidmountainranges withinrelativelydiverselandscapes, whilecroplands wereconcentratedinlowlands withwarmanddry climaticconditions. Speciesnichepositionsofinva- sivespecieswerewidelydistributed ontheenviron- mental ordinationdiagram(Fig.1), mostlyon the lowland areaswithahighdegreeofanthropization.
Thespecies withahigherlevelofspecialization (i.e.nichepositioninmarginalenvironmentsinthe studyareaandnarrownichebreadth)had,ingeneral,a nestednicheinside theniche ofmoregeneralist species(seeFig.2foranexample).
Speciesnichepropertiesasapredictorofrange size
Thegeneralized leastsquare model ofrange sizeasa function ofnichepositionand breadthwhile account- ingforphylogeneticrelationships explainedalmost half of the variation in range size (R2=0.43, p\0.0001)(Table3).Inthismodel,rangesizewas negativelyrelatedtopositionandbreadth onaxis1. Thespecieswithnichepositions closetotheaverage environmentalconditionsshowingsmalltomedium nichebreadth,suchasSorghumhalepense,Robinia pseudoacacia orAmaranthusretroflexus,werealso thespecieswiththe largestrangesize(Fig.3,Table 5 inAppendix).Thosespeciesdonotneedtohavea largenichebreadthtohavelargerangesizegiventhey occupythemostwidespreadenvironmental condi- tions. Specieswithrestricteddistributionrangeswere foundinhighlyhuman-disturbed placeswitharela- tivelywidetoleranceovertheanthropizationgradient (Fig.3).Incontrast, rangesizewaspositivelyrelated withnichebreadth onaxis2:largerangesizespecies were,onaverage, moretolerantofbroadclimateand landscapeconditionsrepresentedbyaxis2(Table3). Interestingly, minimum residence timehadnoinflu- enceonspeciesrangeinrespecttonicheproperties (Table3).
Speciestraitsexplainingspeciesnicheproperties
Modelsaccountingforbothnichepositionandbreadth onaxis1explainedslightlymorevariancethanmodels foraxis2ontheOMIordination(Table4).
Fig.1 Representationof
theexplicativevariablesand nichepositionof74invasive plantspeciesinSpaininthe firsttwoaxesoftheOMI ordinationanalysis.Thefirst axisdenotes agradient of anthropization.Thesecond axisis agradientofclimatic andlandscapegradientfrom mountainousforestedareas
RANGE.TEMP
AGRICULTUR
MEAN.TEMP
RADIATION
MIN.TEMP
tolowland croplands.See Table2foracomplete descriptionofvariables and Table5inAppendixfor species acronyms
DIST.COAST
SCRUB.HERB
OPEN.SPWACATEEWRE.BTOLADNIDS
ROAD.LENGT RAILWAY.LE
BUILD.UP
Axis1
MEAN.ALTIT
RANGE.ALTI
LAND.DIVER
FOREST
ANNUAL.RAI
SUMMER.RAI
Hel_curas
Zyg_fabag
Opu_dille
Bid_pilos
NOixca__gplaruecsS.col_bonar
Aca_cyano
Euc_camal Opu_ficus
Ipo_sagit
Dat_Aimnnao_xvirid
Chl_gayan
Sor_halepAga_aAmuesr_iAsumbau_lmuric
Sop_japon
Ela_angus
Gle_trAAiabcu_theop
IpLLoi_pi_nfdilifc
Gom_frutiEle_indic
Bid_aurea
Ara_sericArc_calen
Ama_bAlistot_squam
Con_bonCaoMrnPi_ar_ssuj__ampplaapstpa
Bid_CCsuabr_aelduli
Axis1
Ama_albusXan_strum
Bro_willd
Ach_Xfailinp_spino
Ail_altisIpo_purpu
Cot_coron
Dat_stram
Ama_hybri
Ama_retro
Con_canad
Ace_negun
Bid_frond
Lon_japon
Pas_dilat
Pas_vagin
Ste_secun
Hel_tuAbret_rverlo
Fal_balds
Ama_powel
Oen_glazi
Sebn__pinsaeeuqd
Tra_flumi
Cor_sello
Spa_paten
Oen_bienn
Aca_dealb
Euc_globu
Sen_mikan
Bud_david
Aca_melan
Bac_halim
Tro_majus
Rey_japon
a
bc
Axis1
d
e
Fig.2 Graphicalrepresentationofthenicheoffourinvasive plantspecieswith restricteddistributionranges:(b)Eucalyptus camaldulensis, range size=75 UTM 10km; (c) Opuntia dillenii17UTM;(d)Fallopia baldschuanica, 86UTMand (e)Acaciamelanoxylon,71UTMnestedinsidethenicheof (a)Conyza canadensis,505UTM,aspecieswithawide distributionrange.Thefirst axisdenotesagradientof anthropization.Thesecondaxisisagradientofclimaticand landscapegradientfrommountainousforestedareastolowland croplands.Nichepositionis the centreof eachspeciesellipsoid andnichebreathisthewidthoftheellipsoidasafunctionofthe maximaldeviationofsiteconditionsfromnicheposition
Table3 Phylogeneticgeneralized linear modelsaccounting forvariationinthedistribution rangesizeofinvasiveplant speciesinSpain,asafunction oftheirnicheposition and breadthonthemaintwoaxesoftheOMIordination and minimumresidencetime
(Table4). Perennial, N-fixingand clonal invasive speciesoccur generallymoreinhighly human- disturbedplacesclosetothecoast(Fig.4).
Nichebreadth overaxis1wasmainlyrelatedto pathway, N-fixing and longevity traits (Table4). Specieswithwidernichebreadth overthenaturalto human-disturbed gradientwerepredominantlyintro- ducedforgardeningandagriculturalpurposes,arenon N-fixingandbiennialspecies.Reversely,smallniche breadth species overaxis1werepredominantly speciesintroducedforsilviculture(Fig.4).
Nichepositionoveraxis2representing agradient fromforestedtocroplandareaswasmainlyrelatedto N-fixing speciesandlongevityintrinsictraitsand pathwayof introduction (Table4). Unintentionally introducedspecies,agriculturalweeds,andcropplants werethose presentinthewarmest,driest,agricultural areas.Speciesoccurring insuchhabitats wereperen- nial,nonN-fixingspecies(Fig.4).
Finally, speciesnichebreadth overaxis2was mainlyaffectedbylongevity, clonalgrowthand pathwayofintroduction(Table4).Short-lived clonal species(i.e.annualsandbiennials)introducedthrough agriculture(i.e.cropsandweeds) hadgenerallya widertolerancetodifferent climaticandlandscape conditionsthantheotherspecies(Fig.4).
Discussion
Nested patternandspeciesniches
Attheregionalscale,invasiveplantspeciesassem- blages in Spain are organized in nested subsets:
Value / SE / tvalue / pvaluespecieswithrestricteddistributionrangesoccurinsiteswithhighinvasiveplantrichness,whereassites
\0.0001withfewspeciesaregenerallypopulatedbywide-
\0.0001spreadinvaders.Oneofourinitialhypotheseswasthat
Intercept / 5.66 / 0.47 / 11.95
Axis1position / -0.60 / 0.11 / -5.68
Axis1breadth / -0.30 / 0.10 / -3.12 / \0.01 / such a nested pattern could reflect species niche
Axis2position / 0.03 / 0.09 / 0.32 / 0.75 / overlap.Hereweshowthatrangesizeisindeedrelated
Axis2breadth / 0.24 / 0.08 / 3.13 / \0.01 / tonichepositionandbreadthofinvaders,suggestinga
Residencetime / 0.00 / 0.00 / 1.02 / 0.31 / nicheoverlapamonginvasiveplantspecies.However,
Themodelexplainedalmosthalfofthevariationinrangesize
(R2=0.43,p\0.0001)
Ingeneral,species nicheposition overaxis1 representing agradientofanthropizationwasmainly conditionedbylongevity,N-fixingandclonalgrowth
there are several alternative explanations for this
nestedpattern.Nestedness mightbegeneratedby factorssuchaspassivesamplingeffect,patchareaand isolation,andhabitatdistribution(Wrightetal.1998), ormaysimply beanartefactofparticularspecies’ distributions (GastonandBlackburn2000).Inour database,therecouldbeapassivesamplingeffect
01234
NichepositiononOMIaxis1
234567
NichebreadthonOMIaxis1
0123456
NichebreadthonOMIaxis2
Fig.3 Relationships between range size of invasive plant speciesinSpainandnicheproperties.Onlythevariableshaving asignificanteffectonrangesizeareplotted(nichepositionon OMI axis 1,nichebreadthon OMI axes 1and2).Thefirstaxis denotesagradientofhuman-pressure.Thesecondaxisisa
gradientofclimaticandlandscape gradientfrommountainous forestedareastolowlandcroplands. Thestraightlinecorre- spondstotheestimatedresponse ofrangesizetoniche propertiesbytheGLSmodels
Table4Relativeimportanceofthenine selectedtraitsfor explaining invasiveplantspeciesnicheproperties overaxes1 and2oftheOMIordination
withthehighestcolonizationabilitywillbeableto reachthemostremotesites(Conroy etal.1999; Kadmon1995).Finally,thenestedpattern ofinvasive
reportedthatthefloraofurbanizedareasandtheir
R2 0.20 0.27 0.13 0.23
Thethreetraitswiththehighestrelativeimportanceoneach nicheproperty areshown inbold.Thelastrowshowsthe varianceexplainedbythemodels(allsignificant,p\0.0001)
becausethecartographicinformationoftheAtlasof
InvasivePlantSpeciesinSpain(Sanz-Elorzaetal.
2004)havebeengeneratedthrough5 yearsofwork compilingbibliographicandherbariumrecords.How- ever, weminimizedthiseffectbyanalysing onlythe cellswithatleastoneinvasivespeciesandbyselecting onlyspeciesrecorded atleastin10UTMcells. Although webasedtheanalysisinagridcellof constantsize(i.e.UTM10910km),habitatisola- tioncouldalso causenestednessbecause onlyspecies
aggregatedinhighlyanthropogenic areas,while widespreadspecieshavetheirnichepositionscloser totheaverageprevailingenvironmental conditions. Coastal areas in Spain have a mild climate and intensehumanpressures, concentratingtourism, tradingandtransport centresand,inturn,they accumulatemostofthefirst recordsofalienspecies (Gasso´etal.2009).
Thelackofassociation between rangesizeand nicheposition ontheclimate-landscapegradient(2nd OMIaxis)reinforcestheideathatinvasive speciesin Spainarenotparticularlyconstrainedbyenvironmen- talconditions.Nevertheless,rangesizewaspositively relatedtonichebreadth alongthisgradient, which impliesthatbroader niches inclimateandlandscape enable a species to become more widespread, as
Longevity
N-fixing
Clonality
AnnualBiennial Perennial
NoYes
NoYes
Pathway
N-fixing
Longevity
AGSUW
NoYes
AnnualBiennial Perennial
N-fixing
Longevity
Pathway
NoYes
Longevity
AnnualBiennial Perennial
Pathway
AGSUW
Clonality
AnnualBiennial Perennial
AGSUW
NoYes
Fig.4 PredictedmeannichevaluefromtheGLSmodelsfor invasiveplantspeciestraits.Onlythethreemostimportanttraits explainingnichepropertiesareplotted(byorderofimportance
fromlefttoright).SeeTable1fortraitcodeandTable4fortrait importance(boldnumbers)
suggestedbyotherstudies(Brown1995;Brownetal.
1995;Va´zquez2006).Interestingly, widespreadspe- ciesarenotnecessarilymoretoleranttoawiderange ofanthropogenic conditions thanrestrictedspecies, andtheirmeannichepositionsarelocatedinsiteswith moderatehuman-dominated landscapes.Thispro- vides evidence that the intermediate disturbance
hypothesisalso holds for invader species richness
(HobbsandHuenneke 1992;McKinney2002). Speciesnichesandtraits
An overview of the relationship between species niches and traits revealed that some intrinsic
(longevity,N-fixing capacity,clonalgrowth)and extrinsic(pathway)traitsarerelatedtonicheposition andnichebreadth.Nitrogen-fixingcapacityiscom- monlyrelatedtoinvasion successbecauseofthe competitiveadvantagethatitrepresentsinnitrogen poorsoils(PysˇekandRichardson2007;Thuiller et al.
2006).However,theabilitytofix Nisusuallyatthe costofbeingspecialisedtoNpoorsoils(Fitter1996). Thistrade-offexplainswhyN-fixing speciesmostly occurincoastal areas withhighhuman pressure, whereN-rich habitats(i.e. croplands)arescarcerthan inland, and exhibit relativelysmaller nichebreadthin comparisontononN-fixingspeciesthatcanoccur moreorlesseverywhere.
Short-livedinvasivespeciesoccurredinthemost widespreadenvironmental conditions andthey showedthehighest tolerancetoclimate-landscape variation.Short-livedspecieshavebeensuggested to bebetterinvadersthanlong-livedtypes(Cadotte and Lovett-Doust2001;Pysˇek etal.1995;Pysˇek and Richardson2007)duetotheiryearlyreproduction or theirsurvival asdormant seedsinadverseconditions (Crawley 1997).Inourdataset,clonalspeciesare indeedfoundinmanydifferenttypes ofecosystems (i.e.coastal,human-disturbed,hotanddryplaces)and havelargenichebreadthonthesecondenvironmental gradientsrevealedbytheOMI.
Extrinsictraitshaverecentlybeenconsidered as predictors ofinvasionsuccess(PysˇekandRichardson
2007;Thuilleretal.2006). Tooursurprise,the pathway ofintroduction,acomponent directlyasso- ciatedtohumanactivityisrelatedtonichepositionon theclimate-landscapegradientratherthantoniche
climate-landscapegradientthanthoseintroduced by othermeans.
Conclusions
Wedemonstrate herethatlinkingnichepositionand breadthwithnestednessofspeciesdistributionsand invaderrangesizecanbringinsightfullessons.Our analysis revealsthatattheregionalscaleinvasive specieswithrestricteddistributiondonothavespecific topo-climaticrequirements,butratheroccurinhuman disturbed areasalsoinvadedbygeneralistandwide- spreadinvasivespecies.Theinvaderswitharestricted distribution rangearelesstolerantofbroadclimate- landscapeconditionsthanwidespread invaders.With regardtowidespreadinvaders,thesearethespecies invadingtheaverageenvironmental conditionsin Spainbutnotclimaticallyextremesorheavilyhuman- disturbedsites.Thoseinvasionpatternsare,inturn, modulatedbywithspecific life-historytraitssuchas longevity,N-fixingcapabilityandclonalgrowth.
Acknowledgments WethankI.Bartomeus,B.Sa´nchez,D.Sol, M.Vall-lloseraandtwoanonymousrefereesforcommentson earlierversionsofthemanuscriptandC.Roquetfortheconstruction ofthephylogeny.Thisstudyhasbeenpartiallyfinanced bythe ALARM(GOCE-CT-2003-506675)andDAISIE(SSPI-CT-2003-
511202)projectswithinthe6thFrameworkProgramme ofthe EuropeanCommission;theMinisteriodeCienciaeInnovacio´n projectsRIXFUTUR(CGL2009-07515) andMONTES (CSD2008-00040). WTacknowledgessupportfromtheANR SCIONproject(ANR-08-PEXT-03).Thispapercontributestothe forthcoming VirtualLabonBiologicalInvasions withinthe LIFEWATCHframework.
positionontheanthropizationgradient.Speciesbothintroduced byagriculture(i.e.cropsandweeds)and unintentionallyaremostlyfoundinwarm,dryagri- culturalareas; andexhibitlargernichebreadthonthe
Table5 Listof74plantinvaderspeciesmodelled
FamilySpeciesSpecies acronym / Appendix1
SeeTable5.
Rangesize
(no.ofUTM) / Axis1 position / Axis1 breadth / Axis2 position / Axis2 breadth
Malvaceae / Abutilontheophrasti / Abu_theop / 111 / 0.76 / 2.56 / 0.73 / 2.58
Fabaceae / Acaciacyanophylla / Aca_cyano / 21 / 2.79 / 3.65 / 1.85 / 1.27
Fabaceae / Acaciadealbata / Aca_dealb / 210 / 1.50 / 2.63 / -1.85 / 1.61
Fabaceae / Acaciamelanoxylon / Aca_melan / 71 / 2.60 / 1.45 / -2.30 / 0.50
Table5continued
Family / Species / Species / Rangesize / Axis1 / Axis1 / Axis2 / Axis2
acronym / (no.ofUTM) / position / breadth / position / breadth
Aceraceae / Acernegundo / Ace_negun / 66 / 1.29 / 6.54 / -0.39 / 2.27
Asteraceae / Achilleafilipendulina / Ach_filip / 18 / -0.21 / 7.00 / -0.15 / 2.58
Agavaceae / Agaveamericana / Aga_ameri / 231 / 0.93 / 3.23 / 1.15 / 1.80
Simaroubaceae / Ailanthusaltissima / Ail_altis / 230 / 0.61 / 4.12 / -0.09 / 3.01
Amaranthaceae / Amaranthusalbus / Ama_albus / 467 / 0.14 / 3.43 / 0.02 / 2.50
Amaranthaceae / Amaranthusblitoides / Ama_blito / 483 / 0.55 / 2.91 / 0.34 / 2.51
Amaranthaceae / Amaranthushybridus / Ama_hybri / 402 / 0.49 / 3.53 / -0.46 / 4.24
Amaranthaceae / Amaranthusmuricatus / Ama_muric / 207 / 1.49 / 3.06 / 1.14 / 1.84
Amaranthaceae / Amaranthuspowellii / Ama_powel / 90 / 0.19 / 4.16 / -1.16 / 5.56
Amaranthaceae / A.retroflexus / Ama_retro / 608 / 0.08 / 3.19 / -0.59 / 4.20
Amaranthaceae / Amaranthusviridis / Ama_virid / 116 / 1.58 / 3.41 / 1.27 / 1.76
Asclepiadaceae / Araujiasericifera / Ara_seric / 93 / 2.15 / 2.99 / 0.39 / 1.67
Asteraceae / Arctothecacalendula / Arc_calen / 24 / 3.26 / 2.65 / 0.41 / 6.15
Asteraceae / Artemisiaverlotiorum / Art_verlo / 167 / 1.11 / 4.02 / -0.87 / 4.24
Asteraceae / Astersquamatus / Ast_squam / 350 / 0.92 / 2.87 / 0.29 / 2.60
Cactaceae / Austrocylindropuntiasubulata / Aus_subul / 51 / 1.22 / 1.59 / 1.12 / 1.31
Asteraceae / Baccharishalimifolia / Bac_halim / 14 / 3.98 / 1.73 / -2.32 / 0.32
Asteraceae / Bidens aurea / Bid_aurea / 75 / 1.33 / 5.50 / 0.46 / 2.89
Asteraceae / Bidens frondosa / Bid_frond / 89 / 1.13 / 2.90 / -0.72 / 4.04
Asteraceae / Bidens pilosa / Bid_pilos / 25 / 1.97 / 2.89 / 1.75 / 0.93
Asteraceae / Bidens subalternans / Bid_subal / 134 / 1.79 / 2.60 / 0.22 / 1.82
Poaceae / Bromuswilldenowii / Bro_willd / 155 / 1.42 / 4.94 / -0.01 / 4.90
Buddlejaceae / Buddlejadavidii / Bud_david / 102 / 1.04 / 5.09 / -2.31 / 3.98
Aizoaceae / Carpobrotusedulis / Car_eduli / 105 / 1.98 / 3.46 / 0.27 / 2.64
Poaceae / Chlorisgayana / Chl_gayan / 24 / 2.59 / 2.84 / 1.45 / 0.75
Asteraceae / Conyzabonariensis / Con_bonar / 389 / 0.66 / 3.44 / 0.22 / 2.79
Asteraceae / Conyzacanadensis / Con_canad / 505 / 0.15 / 3.60 / -0.75 / 4.04
Asteraceae / Conyzasumatrensis / Con_sumat / 289 / 1.08 / 3.00 / 0.17 / 2.58
Poaceae / Cortaderia selloana / Cor_sello / 62 / 3.06 / 4.42 / -1.37 / 2.45
Asteraceae / Cotulacoronopifolia / Cot_coron / 43 / 2.81 / 2.92 / -0.26 / 5.93
Solanaceae / Daturainnoxia / Dat_innox / 71 / 1.41 / 3.19 / 1.25 / 1.63
Solanaceae / Daturastramonium / Dat_stram / 421 / 0.35 / 3.89 / -0.34 / 4.57
Elaeagnaceae / Elaeagnusangustifolia / Ela_angus / 63 / 0.46 / 5.45 / 0.82 / 1.35
Poaceae / Eleusineindica / Ele_indic / 43 / 2.50 / 4.25 / 0.66 / 3.60
Myrtaceae / Eucalyptuscamaldulensis / Euc_camal / 75 / 0.05 / 1.85 / 1.32 / 1.60
Myrtaceae / Eucalyptusglobulus / Euc_globu / 209 / 2.12 / 2.60 / -2.03 / 1.42
Polygonaceae / Fallopiabaldschuanica / Fal_balds / 86 / -0.76 / 2.47 / -1.23 / 3.31
Caesalpiniaceae / Gleditsiatriacanthos / Gle_triac / 54 / 0.42 / 4.53 / 0.76 / 1.05
Asclepiadaceae / Gomphocarpusfruticosus / Gom_fruti / 55 / 1.96 / 2.00 / 0.63 / 2.75
Boraginaceae / Heliotropiumcurassavicum / Hel_curas / 42 / 2.84 / 2.61 / 2.21 / 0.50
Asteraceae / Helianthustuberosus / Hel_tuber / 184 / 0.82 / 4.26 / -0.92 / 4.55
Convolvulaceae / Ipomoeaindica / Ipo_indic / 150 / 1.71 / 3.17 / 0.86 / 1.44
Convolvulaceae / Ipomoeapurpurea / Ipo_purpu / 129 / 1.01 / 4.45 / -0.08 / 3.53
Convolvulaceae / Ipomoeasagittata / Ipo_sagit / 38 / 1.84 / 2.14 / 1.48 / 0.87
Table5continued
Family / Species / Species / Rangesize / Axis1 / Axis1 / Axis2 / Axis2
acronym / (no.ofUTM) / position / breadth / position / breadth
Verbenaceae / Lippiafiliformis / Lip_filif / 50 / 1.76 / 3.89 / 0.86 / 3.60
Caprifoliaceae / Lonicerajaponica / Lon_japon / 78 / 2.09 / 5.02 / -0.38 / 3.55
Nyctaginaceae / Mirabilisjalapa / Mir_jalap / 242 / 1.12 / 3.32 / 0.18 / 2.36
Solanaceae / Nicotianaglauca / Nic_glauc / 127 / 1.72 / 2.50 / 1.64 / 1.30
Onagraceae / Oenotherabiennis / Oen_bienn / 112 / 0.80 / 4.87 / -1.67 / 4.04
Onagraceae / Oenotheraglazioviana / Oen_glazi / 54 / 1.04 / 4.19 / -1.26 / 2.52
Cactaceae / Opuntiadillenii / Opu_dille / 17 / 1.81 / 1.88 / 1.81 / 0.60
Cactaceae / Opuntiaficus-indica / Opu_ficus / 434 / 0.66 / 2.15 / 1.32 / 2.42
Oxalidaceae / Oxalispes-caprae / Oxa_pes.c / 213 / 1.78 / 2.55 / 1.67 / 1.36
Poaceae / Paspalumdilatatum / Pas_dilat / 190 / 2.14 / 2.89 / -0.77 / 3.67
Poaceae / Paspalumpaspalodes / Pas_paspa / 231 / 1.22 / 3.49 / 0.18 / 2.77
Poaceae / Paspalumvaginatum / Pas_vagin / 40 / 2.97 / 3.91 / -0.32 / 5.24
Polygonaceae / Reynoutriajaponica / Rey_japon / 14 / 2.71 / 4.00 / -3.46 / 2.14
Fabaceae / R.pseudoacacia / Rob_pseud / 594 / 0.94 / 3.79 / -1.33 / 3.97
Asteraceae / Senecioinaequidens / Sen_inaeq / 14 / 0.95 / 5.20 / -1.37 / 1.99
Asteraceae / Seneciomikanioides / Sen_mikan / 20 / 3.19 / 3.29 / -1.73 / 1.11
Solanaceae / Solanumbonariense / Sol_bonar / 48 / 2.21 / 3.97 / 1.66 / 2.00
Fabaceae / Sophorajaponica / Sop_japon / 11 / -0.51 / 4.35 / 0.60 / 1.43
Poaceae / S.halepense / Sor_halep / 743 / 0.46 / 2.47 / 1.19 / 2.63
Poaceae / Spartinapatens / Spa_paten / 32 / 2.91 / 1.29 / -1.41 / 3.97
Poaceae / Stenotaphrumsecundatum / Ste_secun / 28 / 3.51 / 2.32 / -0.59 / 4.23
Commelinaceae / Tradescantiafluminensis / Tra_flumi / 19 / 2.78 / 5.90 / -1.21 / 4.39
Tropaeolaceae / Tropaeolummajus / Tro_majus / 9 / 2.58 / 3.44 / -3.35 / 0.12
Asteraceae / Xanthiumspinosum / Xan_spino / 411 / 0.02 / 3.66 / -0.09 / 3.28
Asteraceae / Xanthiumstrumarium / Xan_strum / 275 / 0.72 / 3.76 / 0.01 / 3.25
Zygophyllaceae / Zygophyllumfabago / Zyg_fabag / 64 / 1.12 / 2.55 / 1.83 / 1.29
DistributionrangesizewasmeasuredasthenumberofUTM10910kmwherethespecieswasobserved,andnichepositionand breadthwereextractedfromthe2mainaxesoftheOMIordination
Appendix2
SeeTable6.
Table6 Extrinsicandintrinsicspeciestraitsfor74plantinvaderspeciesinSpain
Seed size / Longevity / Clonality / Dispersal / Pollination / N-fixing / Llifeform / Pathway / Origin / Origin codeAbutilontheophrasti / 3 / 1 / 0 / Animal / Animal / 0 / Therophyte / Agriculture / Europe / E
Acaciacyanophylla / 4 / 3 / 1 / Animal / Animal / 1 / Phanerophyte / Gardening / Oceania / O
Acaciadealbata / 3 / 3 / 0 / Animal / Animal / 1 / Phanerophyte / Gardening / Oceania / O
Acaciamelanoxylon / 3 / 3 / 1 / Animal / Animal / 1 / Phanerophyte / Silviculture / Oceania / O
Acernegundo / 5 / 3 / 0 / Wind / Wind / 0 / Phanerophyte / Gardening / NAmerica / NM
Achilleafilipendulina / 2 / 3 / 1 / Wind / Animal / 0 / Geocryptophyte / Gardening / Europe / E
Agaveamericana / 3 / 3 / 1 / Animal / Animal / 0 / Chamaephyte / Gardening / NAmerica / NM
Ailanthusaltissima / 5 / 3 / 1 / Wind / Animal / 0 / Phanerophyte / Gardening / Asia / A
Amaranthusalbus / 2 / 1 / 0 / Wind / Wind / 0 / Therophyte / Weed / NAmerica / NM
Amaranthusblitoides / 2 / 1 / 0 / Wind / Wind / 0 / Therophyte / Weed / NAmerica / NM
Amaranthushybridus / 2 / 1 / 0 / Wind / Wind / 0 / Therophyte / Weed / America / AM
Amaranthusmuricatus / 2 / 3 / 0 / Wind / Wind / 0 / Hemicryptophyte / Unintentionally / SAmerica / SM
Amaranthuspowellii / 2 / 1 / 0 / Wind / Wind / 0 / Therophyte / Weed / NAmerica / NM
A.retroflexus / 2 / 1 / 0 / Wind / Wind / 0 / Therophyte / Weed / NAmerica / NM
Amaranthusviridis / 2 / 1 / 0 / Wind / Wind / 0 / Therophyte / Weed / Africa / A
Araujiasericifera / 4 / 3 / 1 / Wind / Animal / 1 / Vine / Gardening / SAmerica / SM
Arctothecacalendula / 3 / 1 / 1 / Wind / Animal / 0 / Therophyte / Unintentionally / Africa / AF
Artemisiaverlotiorum / 1 / 3 / 1 / Wind / Wind / 0 / Geocryptophyte / Unintentionally / Asia / A
Astersquamatus / 2 / 2 / 0 / Wind / Animal / 0 / Therophyte / Unintentionally / America / AM
Austrocylindropuntiasubulata / 3 / 3 / 1 / Animal / Animal / 0 / Phanerophyte / Gardening / SAmerica / SM
Baccharishalimifolia / 1 / 3 / 1 / Wind / Wind / 0 / Phanerophyte / Gardening / NAmerica / NM
Bidens aurea / 2 / 3 / 1 / Animal / Animal / 0 / Hemicryptophyte / Unintentionally / America / AM
Bidens frondosa / 4 / 1 / 0 / Animal / Animal / 0 / Therophyte / Unintentionally / NAmerica / NM
Bidens pilosa / 2 / 1 / 0 / Animal / Animal / 0 / Therophyte / Unintentionally / America / AM
Bidens subalternans / 2 / 1 / 0 / Animal / Animal / 0 / Therophyte / Unintentionally / America / AM
Bromuswilldenowii / 3 / 2 / 0 / Animal / Wind / 0 / Hemicryptophyte / Weed / SAmerica / SM
Buddlejadavidii / 1 / 3 / 1 / Wind / Animal / 0 / Phanerophyte / Gardening / Asia / A
Carpobrotusedulis / 2 / 3 / 1 / Animal / Animal / 0 / Chamaephyte / Gardening / Africa / AF
Chlorisgayana / 3 / 3 / 1 / Wind / Wind / 0 / Hemicryptophyte / Gardening / Africa / AF
Conyzabonariensis / 2 / 1 / 0 / Wind / Animal / 0 / Therophyte / Unintentionally / America / AM
Conyzacanadensis / 1 / 1 / 0 / Wind / Animal / 0 / Therophyte / Unintentionally / NAmerica / NM
Table6continued
Seed size / Longevity / Clonality / Dispersal / Pollination / N-fixing / Llifeform / Pathway / Origin / Origin code
Conyzasumatrensis / 2 / 1 / 0 / Wind / Animal / 0 / Therophyte / Unintentionally / SAmerica / SM
Cortaderia selloana / 3 / 3 / 0 / Wind / Wind / 0 / Chamaephyte / Gardening / SAmerica / SM
Cotulacoronopifolia / 2 / 1 / 0 / Wind / Animal / 0 / Therophyte / Weed / Africa / AF
Daturainnoxia / 3 / 1 / 0 / Animal / Animal / 0 / Therophyte / Unintentionally / SAmerica / SM
Daturastramonium / 3 / 1 / 0 / Animal / Animal / 0 / Therophyte / Unintentionally / America / AM
Elaeagnusangustifolia / 3 / 3 / 1 / Animal / Animal / 0 / Phanerophyte / Gardening / Africa / AF
Eleusineindica / 3 / 1 / 0 / Wind / Wind / 0 / Therophyte / Weed / Asia / A
Eucalyptuscamaldulensis / 2 / 3 / 0 / Wind / Animal / 0 / Phanerophyte / Silviculture / Oceania / O
Eucalyptusglobulus / 3 / 3 / 0 / Wind / Animal / 0 / Phanerophyte / Silviculture / Oceania / O
Fallopiabaldschuanica / 1 / 3 / 1 / Animal / Animal / 0 / Vine / Gardening / EAsia / A
Gleditsiatriacanthos / 4 / 3 / 1 / Animal / Wind / 1 / Phanerophyte / Gardening / NAmerica / NM
Gomphocarpusfruticosus / 4 / 3 / 0 / Wind / Animal / 1 / Phanerophyte / Gardening / Africa / AF
Heliotropiumcurassavicum / 3 / 3 / 1 / Wind / Animal / 0 / Geocryptophyte / Unintentionally / America / AM
Helianthustuberosus / 4 / 3 / 1 / Animal / Animal / 0 / Geocryptophyte / Agriculture / NAmerica / NM
Ipomoeaindica / 3 / 3 / 1 / Animal / Animal / 0 / Vine / Gardening / America / AM
Ipomoeapurpurea / 4 / 1 / 1 / Animal / Animal / 0 / Vine / Gardening / America / AM
Ipomoeasagittata / 4 / 3 / 1 / Animal / Animal / 0 / Vine / Unintentionally / America / AM
Lippiafiliformis / 1 / 3 / 1 / Water / Animal / 0 / Chamaephyte / Gardening / SAmerica / SM
Lonicerajaponica / 3 / 3 / 1 / Animal / Animal / 0 / Vine / Gardening / Asia / A
Mirabilisjalapa / 4 / 3 / 0 / Wind / Animal / 0 / Hemicryptophyte / Gardening / America / AM
Nicotianaglauca / 5 / 3 / 1 / Wind / Animal / 0 / Phanerophyte / Gardening / SAmerica / SM
Oenotherabiennis / 2 / 2 / 0 / Animal / Animal / 0 / Hemicryptophyte / Gardening / NAmerica / NM
Oenotheraglazioviana / 2 / 2 / 0 / Animal / Animal / 0 / Hemicryptophyte / Gardening / Europe / E
Opuntiadillenii / 3 / 3 / 1 / Animal / Animal / 0 / Phanerophyte / Gardening / America / AM
Opuntiaficus-indica / 3 / 3 / 1 / Animal / Animal / 0 / Phanerophyte / Gardening / America / AM
Oxalispes-caprae / 1 / 3 / 1 / Animal / Animal / 0 / Geocryptophyte / Weed / Africa / AF
Paspalumdilatatum / 2 / 3 / 1 / Wind / Wind / 0 / Hemicryptophyte / Weed / America / AM
Paspalumpaspalodes / 2 / 3 / 1 / Animal / Wind / 0 / Hemicryptophyte / Unintentionally / America / AM
Paspalumvaginatum / 2 / 3 / 1 / Animal / Wind / 0 / Hemicryptophyte / Unintentionally / America / AM
Reynoutriajaponica / 3 / 3 / 1 / Animal / Animal / 0 / Geocryptophyte / Gardening / Asia / A
R.pseudoacacia / 3 / 3 / 1 / Animal / Animal / 1 / Phanerophyte / Gardening / NAmerica / NM