Biogeography can be briefly defined as the science that attempts to describe and interpret the geographic dis- tributions of organisms
13 mar 2006 · Biogeography can be defined as the study of the geographical distribution of the organisms This simple definition is accurate but it hides
2005), and herbivory by deer defines the metapopulation ecology of plants on the Florida Keys (Barrett Stiling 2006) Invertebrate distribution can be
It could further be reckoned as a discipline that investigates regarding the frequencies of organisms living on earth, its expansion, its interior and exterior
Biogeography means different things to different researchers, and thus it has many 'schools' or disciplines Some examples are Conservation Biogeography
A definition of areas of endemism can be derived from the vicari- ance model, because the term has been used mainly within the vicariance biogeography
Each patch could be defined as a significant region that could be studied within an ecological biogeographic framework We could even draw random
Ecological biogeography studies the factors that define the spatial distribution of species in the present of individuals per species can be graphically described
PDF document for free
- PDF document for free
31526_751_4_648.pdf
648SYSTEMATICBIOLOGYVOL.51
Syst.Biol.51(4):648-652,2002
DOI:10.1080/10635150290102320
UnitsinBiogeography
BERNHARDHAUSDORF
ZoologischesInstitutundZoologischesMuseumderUniversit¨atHamburg,Martin-Luther-King-Platz3, D-20146Hamburg,Germany;E-mail:hausdorf@zoologie.uni-hamburg.de
Inthepasttwodecades,vicariancebio-
geographyhasbecomeamajorsubdiscipline ofbiogeography.Thesubjectofvicariance biogeographyisthestudyofarearela- tionships(e.g.,NelsonandPlatnick,1981;
MorroneandCrisci,1995;Humphriesand
Parenti,1999).Animportantstepinthein-
vestigationofarearelationshipsisthetrans- formationofataxoncladogramintoan areacladogram.Differentdelimitationsof "unitareas"canresultindifferentoutcomes concerningarearelationships(Henderson,
1991).Therefore,areadelimitationisacru-
cialissue.Mostauthorsagreethatareasofen- demismshouldbetreatedasunits.However, thede?nitionandthedelimitationofareasof endemismarecontroversial.Whereasmost authorsconsideredanextensivesympatryof atleasttwospeciesasafundamentalrequire- ment(NelsonandPlatnick,1981;Platnick,
1991;Morrone,1994;Linder,2001),Harold
andMooi(1994)didnotrequiresympa- trybutratherusedcongruenceamongarea cladogramsasarecognitioncriterion.
AREASOFENDEMISM:DEFINITION
Althoughthereisgeneralagreementthat
areasofendemismaretheunitsinbiogeogra- phy,hardlyanyclearde?nitionofthatterm canbefoundintheliterature(Henderson,
1991;butseeHaroldandMooi,1994;Linder,
2001).Beforerecognitioncriteriaandthede-
limitationofareasofendemismcanbedis- cussed,thetermareaofendemismshould bede?nedclearly.Ade?nitionofareasof endemismcanbederivedfromthevicari- ancemodel,becausethetermhasbeenused mainlywithinthevicariancebiogeography framework.
Accordingtothevicariancemodel,an
ancestralbiotawasfragmentedbytheap- pearanceofabarrier.Thebarrierlimited orobviatedthegene?owbetweenpopula- tionsseparatedbythebarrier.Thisvicari- anceeventresultedinallopatricspeciationof manyofthespeciesformerlyconstitutingthe ancestralbiota.Inthisway,twonewbiotas emerged,whichareseparatedbythebarrier.
Thecausallinkbetweentheappearanceof
thebarrierandtheformationofnewbiotas isessentialinthevicariancemodel.Byrepeti- tionsofthisprocess,smallareaswithdistinct biotas,thatis,withmanyspeciesrestricted toindividualareas,emerge.Theseareareas ofendemism.Thus,areasofendemismcan bede?nedasareasdelimitedbybarriers,the appearanceofwhichentailstheformationof speciesrestrictedbythesebarriers.
Usually,vicarancecannotbeobserveddi-
rectlybecausethisprocessgenerallytakesge- ologictimeperiods.Therefore,itisimportant toderivepredictionsaboutobservablepat- ternsfromthemodel.
Onepredictionthatcanbederivedfrom
thevicariancemodelconcernsthespatial similaritiesofdistributionareasoriginating byvicariance.Afteravicarianceevent,there shouldbeagroupofspeciesrestrictedtoone sideofabarrierandagrouprestrictedto theothersideofthebarrier.Onaverage,the rangeofaspecieswillbemoresimilartothe rangesofotherspecieslivingonthesame sideofthebarrier,thatis,inthesameareaof endemism,thantotherangesofspeciesliv- ingontheothersideofthebarrier.Thus,two groupsoftaxa,eachwithsimilarranges,orig- inatefromavicarianceevent.Suchgroupsof taxahavebeencalledfaunalor?oristicele- ments.Therangesoftwospecieslivingonthe samesideofthebarriermaynotnecessarily overlap,especiallywhenthesespeciesdiffer ecologicallyandthusarerestrictedtodiffer- entportionswithintheareaofendemism.
Thede?nitionofareaofendemismin
thesenseofNelsonandPlatnick(1981:468),
Platnick(1991),Morrone(1994),andLinder
(2001)isbasedonspatialsimilaritiesofdistri- butionareas,thatis,relativelyextensivesym-
patryofatleasttwospecies(Platnick,1991).Downloaded from https://academic.oup.com/sysbio/article/51/4/648/1698781 by guest on 15 August 2023
2002POINTSOFVIEW649
FIGURE1.Hypotheticaldistributionsforthediscussionofthede?nitionofareaofendemism.(a)Twogroups oftwospecieseachwithrelativelyextensivesympatrydescribetwoareasofendemism.(b)Thedistributionarea ofoneofthreegroupsoftwospecieseachwithrelativelyextensivesympatryoverlapsthedistributionareaofthe othergroups.Thus,thesethreegroupsofspeciescannotdescribethreeareasofendemism.
Linder(2001:893)presentedanexplicitedef-
initionofareaofendemismaccordingto whichareasofendemismare"areasdelim- itedbythecongruentdistributionofatleast twospeciesofrestrictedrange."Icallthisthe standardde?nitionofareaofendemism.The hypotheticalexampledepictedinFigure1a demonstratestheprinciple.Therangesof fourspeciesareshown.Thedistributionsof species1and2,and3and4,respectively, aremoreorlesscoincident.Thus,theyde- ?netwoareasofendemism.However,the exampledepictedinFigure1bshowsthat thestandardde?nitionisinsuf?cienttode- cidewhichareasshouldbeconsideredareas ofendemism.Therearethreegroupsoftwo specieswithlargelycongruentranges.The rangesofspecies5and6overlapwiththose ofallotherspecies.Accordingtothede?ni- tionabove,therewouldbethreepartlyover- lappingareasofendemism.Thisinterpre- tationisnotcompatiblewiththevicariance model,whichstatesthatareasofendemism arenonoverlapping.
Asecondpatternthatcanbepredicted
basedonthevicariancemodelconcernsthe phylogeneticrelationshipswithinthetaxain- habitingtheareasofendemism.Becauseof thecausallinkbetweentheappearanceof barriersandtheoriginofnewspecies,area cladogramsbasedontaxonomiccladograms ofindividualtaxashouldbecongruentand shouldre?ecttheuniquehistoryofthear- eas(withtheexceptionofsomespeciesnot respondingtosomeofthevicarianceevents; seePlatnickandNelson,1978;Nelsonand
Platnick,1981).Thispatternisthebasisof
thede?nitiongivenbyHaroldandMooi (1994:262):anareaofendemismis"ageo- graphicregioncomprisingthedistribution oftwoormoretaxathatexhibitaphylo- geneticanddistributionalcongruenceand havingtheirrespectiverelativesoccurringin othersuch-de?nedregions."Contrarytothe suggestionsofNelsonandPlatnick(1981),
Platnick(1991),Morrone(1994),andLinder
(2001),thisde?nitiondoesnotrequirethat allspeciesoccurinalargeportionofthe areaofendemism;therangesofsomespecies maynotoverlapwithinanareaofendemism (e.g.,becausetheyarerestrictedtodifferent habitats).
WhereasHaroldandMooi"s(1994)de?ni-
tionandmyde?nitionofareasofendemism donotnecessarilyrequiretherelativeexten- sivesympatrydemandedbyPlatnick(1991), therequirementthattheareasofendemism bedelimitedbybarriers,theappearanceof whichentailedtheformationofspeciesre- strictedbythebarriers,ismorestringentthan
Platnick"s(1991)requirementofcongruent
distributionallimitsoftwoormorespecies.
Thatareasaredelimitedbybarriers,theap-
pearanceofwhichentailedtheformationof speciesrestrictedbythebarriers,isahypoth- esisthatmustbetested.HaroldandMooi (1994)proposedtotestthepredictionthatthe areacladogramsofdifferenttaxaarecongru- ent.Theproblemwiththisapproachisthat theunitsofthebiogeographicanalysismust bede?nedbeforethetaxoncladogramscan beconvertedintoareacladograms.Thepro- posedtestcanbeusedtodeterminewhether theprede?nedunitareasareareasofen- demism.Obviously,theunitsmustbede- ?nedbeforeareasofendemismcanberec- ognized.Thus,areasofendemismasde?ned hereorbyHaroldandMooi(1994)cannotbe theunitsofbiogeographicanalyses.
Areasofendemismarenotusedasbasic
unitsintheprocedureproposedbyHarold andMooi(1994).Intheirprocedure,areasof occurrencearethebasicunits.Axelius(1991) showedthattheuseofdistributionareasas basicunitsfortheconstructionofareaclado-
gramscanresultinparadoxicalconclusionsDownloaded from https://academic.oup.com/sysbio/article/51/4/648/1698781 by guest on 15 August 2023
650SYSTEMATICBIOLOGYVOL.51
(e.g.,anareaismorerelatedtoadifferent areathantosomepartofitself)whendis- tributionareasoverlap.HaroldandMooi (1994)avoidedthisproblembycombiningall (partly)overlappingareas.Thus,theywere unabletorecognizeareasofendemismwhen dispersalhadoccurred.
AREASOFENDEMISM:DELIMITATION
Althoughthestandardde?nitionofareaof
endemismisbasedondistributionpatterns ofspecies(moreorlesscongruentdistribu- tionalboundariesofatleasttwospecies),ge- omorphologicalorpalaeogeographicalunits (e.g.,islands,continents,mountainranges) areconsideredareasofendemisminmost empiricalstudieswithoutanexplicitanaly- sisofdistributionpatterns.Thisassumption isprobablymadebecauseofproblemsinde- limitingareasofendemism.
Thedelimitationofareasofendemismis
notproblematicwhenspeciesoriginateby vicarianceandthereisnodispersal.Under theseconditions,therearegroupsofspecies havingseparateranges(asinFig.1a).How- ever,thedelimitationofareasofendemism becomesproblematicwhendispersaloc- curs.Theoccurrenceofdispersalhasnever beendoubtedbyvicariancebiogeographers (PlatnickandNelson,1978).Withoutdisper- sal,therewouldbecontinuitywithinand discontinuitybetweenareasofendemism.
Thisconditionhasbeencitedasarecognition
criterionofbioticprovinces(Peters,1955).
However,changesinspeciescompositionare
gradualshiftsoverlargezonesratherthan sharpbreaksbetweenneighboringhomoge- neousareas(Peters,1955;Kaiseretal.,1972).
Considerthedistributionareasofthe
10speciesdepictedinFigure2.Assumethat
thesituationistheresultofavicarianceevent separatingeasternandwesternspeciesand subsequentdispersal.Howlargewerethear- easofendemism?Weretheyaslargeasthe distributionareasofspecies3and9or2and
7,orweretheyevensmallerthanthedis-
tributionareasofspecies5and6?Thebio- geographicaldataaloneareinsuf?cientfor delimitingtheareasofendemism.Dispersal decouplesthehistoryofspeciesfromthehis- toryofareas.Thebest-knownrealexample ofthegeneralproblemdepictedinFigure2 isthedelimitationoftheOrientalandthe
Australianregions.Althoughmanyauthors
haveaddressedthissubject,itisnotpossible
FIGURE2.Hypotheticaldistributionsof10species
demonstratingthedif?cultyinvolvedindelimiting areasofendemismbasedonbiogeographicaldataalone whendispersalhasoccurred. todeterminethelimitsofthesetworegions onthebasisofbiogeographicaldata(Mayr,
1944;HollowayandJardine,1968;Simpson,
1977;Vane-Wright,1991).Mayr(1944)pro-
poseddrawingtheborderlinebetweenthe twozoogeographicregionsintheareawhere thefaunalelementsofthetworegionsinter- minglesuchthatthefaunalelementofone regionprevailsononesideoftheborderline andthefaunalelementoftheotherregion prevailsontheotherside.However,Mayr admittedthatthislinewouldbeanarbitrary separationofacontinuousseriesofvalues atthehalfwaypointbetweentheextremes andthatthelinecouldbedifferentfordif- ferenttaxa.Ifitisnotpossibletoestablish theborderbetweentworegionsseparatedby severalhundredkilometersofoceanonthe basisofbiogeographicaldata,itwillhardly bepossibletodelimitareasofendemismsep- aratedonlytemporarilybyclimaticbarriers onacontinent,forexample.
InsituationssuchasthatinFigure2,the
operationalmethodsproposedforidentify- inganddelimitingareasofendemismby
Morrone(1994)andLinder(2001;seealso
LinderandMann,1998)cannotbeused
to?ndthelimitsofareasofendemism.
Moreover,thereasonsforusingparsimony
algorithmsdesignedtoreconstructadi- chotomoussplitsequencetodelimitareasof endemismareunclear,becausethedistribu- tionoftaxaingridquadratsisnottheresultof suchasplitsequencebetweenthequadrats.
Existingmethodsalsodonottestwhethera
distributionpatternisnonrandom.
Hovenkamp(1997)triedtoevadethe
problemofdelimitingareasofendemism byfocusingonsupposedvicarianceevents
insteadofareasofendemism.However,inDownloaded from https://academic.oup.com/sysbio/article/51/4/648/1698781 by guest on 15 August 2023
2002POINTSOFVIEW651
situationssuchasthatshowninFigure2, thisapproachisnosolutionbecausethepo- sitionofthebarriercausingthevicariance eventcannotbelocatedbecauseofsubse- quentdispersal(atleastnotwithbiogeo- graphicaldata).
BIOTICELEMENTSASBIOGEOGRAPHIC
UNITS
ReconsiderthesituationdepictedinFig-
ure2.Itisnotpossibletodeterminebyan analysisofdistributionareaswhichspecies dispersedhowfarortodelimittheareas ofendemismwiththeavailableinformation.
However,bycomparingthedistributionar-
easofthevariousspecies,twogroupsof specieswithsimilarrangescaneasilyberec- ognized.Suchgroupshavebeencalledfau- nalor?oristicelementsorgroups(e.g.,Rebel,
1931;deLattin,1957;Mayr,1965;Holloway
andJardine,1968;Udvardy,1969;Jardine,
1972;Birks,1987;Frey,1992;Dennisetal.,
1998).Ide?nebioticelementasagroupof
taxawhoserangesaresigni?cantlymoresim- ilartoeachotherthantothoseoftaxaofother suchgroups.Acomputationalmethodfor theanalysisofbioticelementswasdescribed byHausdorfandHennig(unpubl.).
Theexistenceofbioticelementsispre-
dictedbythevicariancemodel(seeabove), andincontrasttoareasofendemism,biotic elementscanbedeterminedbyusingdistri- butiondataalone.Hence,bioticelementsare suitableforuseasbiogeographicunits.
Areasofendemismaresuitableforuseas
biogeographicunitsonlywhenspeciesorig- inatebyvicarianceandthereisnodisper- sal.However,thesepresuppositionsarenot testedinvicariancebiogeographicstudies.
Vicariancebiogeographyregardsdispersal
asasecondaryprocessthatcreatesnoisein thedata.However,bioticelementsdonot presupposeaspeci?cspeciationmodeand aresuitableunits,evenwhenextensiverange expansionsorrangeshiftshaveoccurred, suchasinresponsetoclimaticchanges(Frey,
1992).Theconceptofbioticelementstriesto
summarizeallgeneralitiesofthegeographi- caldistributionsoforganisms.
Thedelimitationofbioticelementsisonly
the?rststepinabiogeographicanalysis.
Whether(partsof)bioticelementsarehis-
toricalunitsandhowfartheyarein?uenced byecologicalfactorsmustbeinvestigatedby furtheranalyses.
Ifbiotasevolveaccordingtothevicariance
modelwithoutdispersal,eachbiotawould consistofasinglebioticelementandthe bioticelementswouldnotoverlap.Justas sympatryofsistergroups,sympatryofbiotic elementsisevidencefordispersal.Ifabiota consistsofasinglebioticelement,thereis noevidenceforvicarianceeventswithinthat biota.Onlywhendifferentbioticelements canbedistinguishedisitmeaningfultoin- vestigatehowtheyhavebeenin?uencedby vicarianceanddispersal.
Bioticelementsarenotnecessarilygen-
eratedbyvicarianceevents.Bioticelements canalsooriginatewhenanareaiscolonized fromdifferentsourceareasacrosspreexist- ingbarriersbychancedispersalandwhen thepopulationsintheareaunderconsider- ationevolveintonewspecies.Anotherpos- siblescenarioisthatspeciesthatoriginated inseparateareascolonizethesamelargerre- gion,eitheraftertheremovalofbarriersor bychancedispersal.
Insuchcases,bioticelementsaregeo-
graphicalbutnothistoricalunits.Whether bioticelements(orpartsofthem)arehis- toricalunitscanbetested,suchasbya comparisonofthecladogramsoftherespec- tivegroups.Thetaxoncladogramscanbe convertedintoelementcladogramsbyre- placingthetaxonnameswiththerespec- tivebioticelement(analogoustothearea cladogramapproach).Theelementclado- gramsofthosegroupsthatformhistorical unitsshouldmatch.Thepartsofthebiotic elementsthatformhistoricalunitscanbe termedvicarianceelements.
Someproblemswiththeareacladogram
approachcanbesolvedbyusingbioticel- ementsasunitswiththeelementclado- gramapproach.Oneimportantissuecon- cernsgroupswithdifferenthistoriesinthe samearea.Forexample,theterrestrialmol- luscfaunaofnortheasternAfricaisdom- inatedbyPalearcticgroups,whereasthe freshwatermolluscfaunaisdominatedby
Ethiopiangroupsthatcolonizednortheast-
ernAfricathroughtheNile(Pallary,1909).
Areacladogramsbasedonterrestrialsnails
implythatnortheasternAfricaisrelatedto otherMediterraneancountries,whereasarea cladogramsbasedonfreshwatermolluscsin- dicatethatnortheasternAfricaisrelatedto partsofsubsaharanAfrica.Aconsensustree ofsuchcladogramswouldbeunresolved.
TheareacladogramapproachdoesnotrevealDownloaded from https://academic.oup.com/sysbio/article/51/4/648/1698781 by guest on 15 August 2023
652SYSTEMATICBIOLOGYVOL.51
informationonthehistoryofnortheastern
Africabecausethefaunaiscomposedofdif-
ferentelementswithdifferenthistoriesthat cannotbeshowninasingleareacladogram.
Ananalysisofthebioticelementsisamore
appropriateŽrststeptowardunderstanding thegenesisofcompositebiotas.
REFERENCES
AXELIUS,B.1991.Areasofdistributionandareasof
endemism.Cladistics7:197-199.
BIRKS,H.J.B.1987.Recentmethodologicaldevelop-
mentsinquantitativedescriptivebiogeography.Ann.
Zool.Fenn.24:165-177.
DELATTIN,G.1957.DieAusbreitungszentrender
holarktischenLandtierwelt.Zool.Anz.Suppl.20:
380-410.
DENNIS,R.L.H.,W.R.WILLIAMS,ANDT.G.SHREEVE.
1998.FaunalstructuresamongEuropeanbutter?ies:
Evolutionaryimplicationsofbiasforgeography,
endemismandtaxonomicaf?lation.Ecography21:
181-203.
FREY,J.K.1992.Responseofamammalianfaunal
elementtoclimaticchanges.J.Mammal.73:43-50.
HAROLD,A.S.,ANDR.D.MOOI.1994.Areasof
endemism:De?nitionandrecognitioncriteria.Syst.
Biol.43:261-266.
HENDERSON,I.M.1991.Biogeographywithoutarea?
Aust.Syst.Bot.4:59-71.
HOLLOWAY,J.D.,ANDN.JARDINE.1968.Twoap-
proachestozoogeography:Astudybasedonthe distributionsofbutter?ies,birdsandbatsintheIndo-
Australianarea.Proc.Linn.Soc.Lond.179:153-188.
HOVENKAMP,P.1997.Vicarianceevents,notareas,
shouldbeusedinbiogeographicalanalysis.
Cladistics13:67-79.
HUMPHRIES,C.J.,ANDL.R.PARENTI.1999.Cladistic
biogeography.OxfordUniv.Press,Oxford,U.K.
JARDINE,N.1972.Computationalmethodsinthestudy
ofplantdistributions.Pages381-393inTaxonomy, phytogeographyandevolution(D.H.Valentine,ed.).
AcademicPress,London.
Syst.Biol.51(4):652-664,2002
DOI:10.1080/10635150290102311
ResolutionofaSupertree/SupermatrixParadox
JOHNGATESY,1CONRADMATTHEE,2ROBDESALLE,3ANDCHERYLHAYASHI1
1DepartmentofBiology,UniversityofCalifornia,Riverside,California92521,USA
2DepartmentofZoology,UniversityofStellenbosch,Stellenbosch7602,SouthAfrica
3DepartmentofInvertebrates,AmericanMuseumofNaturalHistory,CentralParkWestat79thStreet,
NewYork,NewYork10024,USA
Inthetotalevidenceapproachtosystem-
atics,allcharactersandtaxaaremergedin asingle"supermatrix,"andthedataare analyzedsimultaneously(Miyamoto,1985;
Kluge,1989;NixonandCarpenter,1996).
Recently,however,apleahasbeenmade
KAISER,G.W.,L.P.LEFKOVITCH,ANDH.F.HOWDEN.
1972.FaunalprovincesinCanadaasexempli?edby
mammalsandbirds:Amathematicalconsideration.
Can.J.Zool.50:1087-1104.
LINDER,H.P.2001.Onareasofendemism,withanexam-
plefromtheAfricanRestionaceae.Syst.Biol.50:892- 912.
LINDER,H.P.,ANDD.M.MANN.1998.Thephylogeny
andbiogeographyofThamnochortus(Restionaceae).
Bot.J.Linn.Soc.128:319-357.
MAYR,E.1944.Wallace"slineinthelightofrecent
zoogeographicstudies.Q.Rev.Biol.19:1-14.
MAYR,E.1965.Whatisafauna?Zool.Jahrb.Syst.92:473-
486.
MORRONE,J.J.1994.Ontheidenti?cationofareasof
endemism.Syst.Biol.43:438-441.
MORRONE,J.J.,ANDJ.V.CRISCI.1995.HistoricalBio-
geography:Introductiontomethods.Annu.Rev.Ecol.
Syst.26:373-401.
NELSON,G.,ANDN.PLATNICK.1981.Systematicsand
biogeography:Cladisticsandvicariance.Columbia
Univ.Press,NewYork.
PALLARY,P.1909.Cataloguedelafaunemalacologique
del"´Egypte.Mem.Inst.´Egyptian6:1-92.
PETERS,J.A.1955.Useandmisuseofthebioticprovince
concept.Am.Nat.89:21-28.
PLATNICK,N.I.1991.Onareasofendemism.Aust.Syst.
Bot.4:xi-xii.
PLATNICK,N.I.,ANDG.NELSON.1978.Amethodof
analysisforhistoricalbiogeography.Syst.Zool.27: 1-16.
REBEL,H.1931.ZurFragedereuropäischenFaunenele-
mente.Ann.Naturhist.Mus.Wien46:49-55.
SIMPSON,G.G.1977.Toomanylines;thelimitsofthe
OrientalandAustralianzoogeographicregions.Proc.
Am.Philos.Soc.121:107-120.
UDVARDY,M.D.F.1969.Dynamiczoogeography.Van
NostrandReinhold,NewYork.
VANE-WRIGHT,R.I.1991.TranscendingtheWallaceline:
DothewesternedgesoftheAustralianregionandthe
Australianplatecoincide?Aust.Syst.Bot.4:183-197.
Firstsubmitted9May2001;reviewsreturned
22March2002;?nalacceptance2May2002
AssociateEditor:PeterLinder
fortheuseof"supertrees"insystemat- ics(Sandersonetal.,1998).Inthisframe- work,individualcharactersarenotinter- pretedasphylogeneticevidence.Instead, topologiessupportedbydifferentpublished
studies,thatis,sourcetrees,areencodedintoDownloaded from https://academic.oup.com/sysbio/article/51/4/648/1698781 by guest on 15 August 2023
Biogeography Documents PDF, PPT , Doc