Ecology and Evolution. 2019;00:1–13.  | 1www.ecolevol.org

Received:6August2019  |  Revised:17September2019  |  Accepted:15September2019DOI:10.1002/ece3.5740

O R I G I N A L R E S E A R C H

Highlighting convergent evolution in morphological traits in response to climatic gradient in African tropical tree species: The case of genus Guibourtia Benn.

Félicien Tosso1,2,3  | Jean‐Louis Doucet1 | Kasso Daïnou3 | Adeline Fayolle1 | Alain Hambuckers4 | Charles Doumenge5 | Honoré Agbazahou5 | Piet Stoffelen6 | Olivier J. Hardy2

ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium,providedtheoriginalworkisproperlycited.© 2019 The Authors. Ecology and EvolutionpublishedbyJohnWiley&SonsLtd.

1ForestisLife,TERRATeachingandResearchCentre,GemblouxAgro‐BioTech,UniversityofLiège,Gembloux,Belgium2EvolutionaryBiologyandEcologyUnit,FacultédesSciences,UniversitéLibredeBruxelles,Brussels,Belgium3Nature+asbl,s/cForestisLife,TERRATeachingandResearchCentre,GemblouxAgro‐BioTech,UniversityofLiège,Gembloux,Belgium4URSPHERES,BehavioralBiology,UniversityofLiege,Liege,Belgium5CentreInternationaldeRechercheAgronomiquepourleDéveloppement,TAC‐105/D,CampusInternationaldeBaillarguet,Montpellier,France6Herbarium,BotanicGardenMeise,Meise,Belgium

CorrespondenceFélicienTosso,ForestisLife,TERRATeachingandResearchCentre,GemblouxAgro‐BioTech,UniversityofLiège,2PassagedesDéportés,B‐5030Gembloux,Belgium.Email:tfelicien@yahoo.fr

Funding informationFondspourlaFormationàlaRecherchedansl'Industrie et l'Agriculture—Fonds National pourlaRechercheScientifique(FRIA‐FNRS,Belgium)

Abstract1. Adaptiveevolutionisamajordriveroforganismdiversification,butthelinksbe‐tween phenotypic traits and environmental niche remain little documented intropicaltrees.Moreover,trait‐nicherelationshipsarecomplexbecauseacorrela‐tionbetweenthetraitsandenvironmentalnichesdisplayedbyasampleofspe‐ciesmayresultfrom(a)convergentevolutionifdifferentenvironmentalconditionshaveselecteddifferentsetsoftraits,and/or(b)phylogeneticinertiaifnicheandmorphological differences between species are simply function of their phylo‐geneticdivergence, inwhichcasethetrait‐nichecorrelationdoesnotimplyanydirect causal link.Here,we aim to assess the respective roles of phylogeneticinertiaandconvergentevolutioninshapingthedifferencesofbotanicaltraitsandenvironmentalnichesamongcongenericAfricantreespeciesthatevolvedindif‐ferentbiomes.

2. This issue was addressed with the tree genus GuibourtiaBenn.(LeguminosaeandDetarioideae),whichcontains13Africanspeciesoccupyingvariousforesthabitattypes,fromrainforesttodrywoodlands,withdifferentclimateandsoilconditions.Tothisend,wecombinedmorphologicaldatawithecologicalnichemodellingandusedahighlyresolvedplastidphylogenyofthe13AfricanGuibourtiaspecies.

3. First,wedemonstratedphylogeneticsignalsinbothmorphologicaltraits(Manteltestbetweenphylogeneticandmorphologicaldistancesbetweenspecies:r=.24,p=.031)andenvironmentalniches(Manteltestbetweenphylogeneticandnichedistancesbetweenspecies:r=.23,p=.025).Second,wefoundasignificantcor‐relationbetweenmorphologyandniche,atleastbetweensomeoftheirrespec‐tivedimensions(Mantel'sr=.32,p=.013),evenafteraccountingforphylogeneticinertia(PhylogeneticIndependentContrast:r=.69,p=.018).Thiscorrelationoc‐curredbetweensomeleafletandflowertraitsandsolarradiation,relativehumid‐ity,precipitations,andtemperaturerange.

2  |     TOSSO eT al.

1  | INTRODUC TION

Historicalandenvironmentalfactorscontributebothtovariationintraitsacrossspecies (Freckleton&Jetz,2008). Inthebasictheoryofquantitativegenetics, a rapidevolutionarychange in traits thatareselectivelysensitivetolabileenvironmentalvariablesisexpected(Labra,Pienaar,&Hansen,2009).Buttheseexpectationsmetcom‐plicationssinceitisdifficulttodiscardrelativeroleofenvironmentaldeterminantsandphylogeneticeffectsinplantspeciestraitsvaria‐tion(Grafen,1989;Desdevises,Legendre,Azouzi,&Morand,2003;Diniz‐Filho,deSant'Ana,&Bini,1998;Westoby,Leishman,&Lord,1995;Wiens&Graham,2005).Inmodernphylogeneticcomparativemethods,thisproblemisparticularlycomplexwhenrelatedspecieshaverelatedenvironmentalniches(Labraetal.,2009;Losos,2008;Price,1997;Wiens&Graham,2005).Itisthencrucialtodistinguishthesimilaritybetweentraitsofrelatedspeciesandnichethatcanbeattributabletocommonancestry(phylogeneticinertia)fromsimilar‐itytoconvergentevolutionarychange.

Speciestraitscanevolveinresponsetoselectivepressuresbutalso neutrally (Freckleton & Jetz, 2008). Adaptive processes aresuspectedwhenparticulartraitsareassociatedwithparticularen‐vironmentalconditions.However,demonstratingcausallinksisdif‐ficultinmacroevolution.Tounderstandwhatdrivestheevolutionofspeciestraitsandniches,phylogeneticcomparativemethodsassessthecorrelationbetweencharactersamongspecies,whileaccountingfortheirphylogeneticrelationships,todecipherwhethertraitsimi‐laritybetweenspeciesreflectsphylogenetic inertia (Figure1Sc.3)or convergent evolution (Figure 1 Sc.2; Labra et al., 2009). Theseapproacheshaverarelybeenappliedtounderstandthetrait‐nicherelationshipsoftropicaltrees,partlybecausewell‐resolvedphylog‐eniesareoftenlacking.

Convergent evolution is the independent evolutionof a set ofsimilartraitsinorganismsfromdifferentlineagessubjecttosimilarabioticorbioticagentsofnaturalselection(Cody&Mooney,1978).Itisconsideredasanevidenceofadaptation(McLennan&Brooks,1993; Pagel, 1994) and implies causal traits‐niche relationships.Alternatively to adaptation, phylogenetic inertia has been viewedastaxon‐specific limitationsthat forcea taxon intocertaincombi‐nationsoftraitsregardlessofthenicheinwhichthattaxonoccurs(Morales,2000).Anapparenttrait‐nicheassociationinasetofspe‐ciesmaythusoccurifthemostphylogeneticallyrelatedspeciestendtosharebothsimilartraitsandsimilarenvironmentalnichesdueto

many possible processes, including genetic drift, stabilizing selec‐tion,andconstraints.

Todate,themechanismsunderlyingplantspeciesdiversificationand the evolution of their morphological traits in tropical forestsremain littleknown.Botanistsseek traits thatallowdistinguishingspeciesorhighertaxabuttheriskthatconvergentevolutionaffectsthesetraitsisrarelyassessed.Moreover,theroleofenvironmentalvariationasapotentialdriverofadaptivemorphologicalevolution(Ortiz‐Medrano, Scantlebury, Vázquez‐Lobo, Mastretta‐Yanes, &Piñero, 2016) has received little attention in African tropical treespecies. Understanding these mechanisms can provide importantinsightsintotheevolutionofAfricanbiomesandguidestrategiesfortheirconservation(Linder,2014).

Ifnaturalselectionbyabioticfactorsisamaindriveroftheevo‐lutionofspecies traits,aclose relationshipbetweenspecies traitsandtheenvironmentalconditionscanbeexpected(Ricklefs,1987;Schluter, 1988). However, this statement remains to be demon‐strated, especially for congeneric tropical tree species, becausemorphologically similar tree species can occur in contrasted envi‐ronments, forexample fromdense rain forests todrywoodlands/savannahs, so that the relative importance of adaptive processesinshapingthemorphologicaltraitsthatdiscriminatethemremainsunclear.

Acomparativeapproachtoexploreevolutionarypatternsdrivenby environmental variation (Freckleton & Jetz, 2008) requires awell‐resolvedphylogenyofthecomparedspeciestodistinguishcon‐vergentevolutionfromspurioustrait‐nicheassociationduetophylo‐geneticinertia(Giarla&Esselstyn,2015).Tostudytherelationshipsbetweentraitsandenvironmentaldriversfollowinganecomorpho‐logicalapproach(Motta&Kotrschal,1991;Wiens&Graham,2005),Broennimann et al. (2012) proposed multivariate techniques toquantitativelycomparetheenvironmentalnichesbetweenpairsofspecies (e.g., indicesofnicheoverlap,nicheequivalencyandnichesimilarity).Thesenewnichemodellingtoolscombinedwithphylo‐genetic information (a)helptoassesshowtherealizedniches (seeSoberón, 2007) of closely related species have evolved over time(e.g.,Guisan&Thuiller,2005;Ortiz‐Medranoet al., 2016), and (b)allowtoexplaintheobservedbiogeographicalpatternsandtotestecologicalhypotheses(Ratoetal.,2015).

The present work aimed to understand the relationships be‐tweenenvironmentalnicheandmorphological(dis)similarityamongcongenerictropical treespecieswhileaccountingforphylogenetic

4. Ourresultsdemonstratetheconvergentevolutionofsomemorphological traitsinresponsetoclimaticfactorsincongenerictreespeciesandhighlighttheactionofselectiveforces,alongwithneutralones, inshapingthedivergencebetweencongenerictropicalplants.

K E Y W O R D S

evolutionaryecology,Guibourtia,nichecomparison,phenotypicadaptation,PhylogeneticIndependentContrast,phylogeneticsignal,speciation,taxonomy

     |  3TOSSO eT al.

inertia.Morespecifically,weaddressedthefollowingquestions:(a)Do speciesmorphological traits and environmental niches displayphylogeneticsignals?,(b)aremorphologicalandenvironmental(dis)similarity between species correlated?, (c) are morphological andenvironmental (dis)similarity still correlatedafter factoringout theimpactofphylogeneticinertiasuggestingconvergentevolution?Byansweringthesequestions,wetestedwhethertheassociationbe‐tweenmorphologicaltraitsandenvironmentalnichewithinconge‐nericspeciessupportsoneofthefourscenariosdetailedinFigure1.

We addressed this issue using the genus Guibourtia Benn.(Leguminosae and Detarioideae), which includes 13 African treespecies(Tosso,Daïnou,Hardy,Sinsin,&Doucet,2015).Thisgenusrepresentsagoodmodel forourquestionsbecause (a) its speciesoccuroverawide rangeofhabitat types (rainforests,dry forests/savannas), and (b) awell‐resolved phylogeny of this genus is nowavailable(Tossoetal.,2018)andshowsthreeclades,eachincludingspeciesoccupyingcontrastingenvironments.

2  | METHODS

2.1 | Study species

The AfricanGuibourtia species occur in a variety of vegetationtypes,across theSudano‐SahelianRegion, theGuineo‐congolianRegion,andtheZanzibar‐InhambaneregionalMosaic(Tossoetal.,2015).The13speciescanberoughlycategorizedintospeciesin‐habitingrelativelydryandseasonalclimatesharboringtropicaldryforests(densedryforestsandwoodlands)orwoodedsavannahs,hereafteridentifiedbythe#symbol,andspeciesinhabitingwet/

moistandweakly seasonal climates favouringdensewet forests(specieswith * symbol). Six species are associatedwith dry andseasonal climates:G. carrissoana# (M.A.Exell) J.Leonard,G. cole-osperma# (Benth.)Leonard,G. conjugata# (Bolle)J.Leonard,G. co-pallifera#Benn.,G. schliebenii# (Harms) J.Leonard, andG. sousae# J.Leonard;andsevenspeciestowet/moistclimates:G. arnoldiana* (DeWild.&T.Durand)J.Leonard,G. demeusei*(Harms)J.Leonard,G. dinklagei*(Harms)J.Leonard,G. ehie*(A.Chev.)J.Leonard,G. le-onensis* J.Leonard, G. pellegriniana* Leonard, and G. tessmannii* (Harms)J.Leonard(FigureS1).

2.2 | Morphological traits

A total of 281 georeferenced herbarium samples were usedfor the morphological analyses. They were collected between1889 and 2010 and conserved in museums and botanical gar‐dens (the National Museum of Natural History in France, theBotanic GardenMeise in Belgium, the Royal Botanic Garden ofEdinburg inUK, theNairobiUniversityHerbarium inKenya,andtheHerbariumofMaputoinMozambique).Amongthesesamples,57%werefertile(withflowers,fruits,and/orseeds),andonlythespecimens forwhich thedeterminationwasvalidatedbyexperi‐encedbotanists in Leguminosae (mostly J. Leonard, J.Wieringa,M.Fougère‐Danezan,orR.Letouzey)weretakenintoaccount.Alistof45morphologicalcharactersweremeasuredonall(oronlyfertile)samples,usingthedeterminationkeysofLéonard (1950),Aubreville (1970), Tosso et al. (2015) and Fougère‐Danezan,Herendeen,Maumont,andBruneau(2010)(Table1).Concerningflowers,onceremoved,thepartstobemeasuredwererehydrated

F I G U R E 1  Fourhypotheticalscenarios(Sc1toSc4)regardingtheassociationbetweenamorphologicaltrait(M)andadimensionoftheenvironmentalniche(E)amongsixcloselyrelatedspeciesofknownphylogeny(P).Thetraitandnichevaluesarerepresentedbythesizeofeachgray(M)orblack(E)circle.UnderSc1,speciesmorphologicaltraitsandenvironmentalnichesevolvedquickly,showingnophylogeneticsignal,andindependently,showingnoM‐Ecorrelation.UnderSc2,morphologicaltraitandenvironmentalnicheevolvedquicklybutshowstrongM‐Ecorrelationduetoconvergentevolution(sameMtraitsselectedinsimilarEniches).UnderSc3,bothmorphologicaltraitandenvironmentalnicheevolveslowly,leadingtophylogeneticsignalsandresultinginaM‐Ecorrelation,butthereisnoremainingM‐EcorrelationwhenapplyingaPIC(PhylogeneticIndependentContrast)analysisfactoringoutthecorrelationduetophylogeneticdependencebecausetheyevolvedindependently.UnderSc4,astrongM‐Ecorrelationisduebothtophylogeneticinertiaandconvergentevolution(significantPICanalysis).WeexpecttheM‐EcorrelationtobestrongerthantheM‐PandE‐Pcorrelationsinscenarios2and4implyingconvergentevolution,andtobelessstrongthantheM‐PandE‐Pcorrelationsinscenario3implyingonlyphylogeneticdependence

4  |     TOSSO eT al.

in boilingwater at 90°C for 3min. The flowers (1–3)were thendissectedandobserved.Thefloralpiecesweremeasuredwithamicrometerincorporatedintoabinocularmicroscope(NachetGLI154),atmagnification×10–40.Themicroscopewasalsousedtocheckforthepresenceofglandsandhairsontheleaflets,flowers,andfruits(Table1).

2.3 | Environmental data

Environmental data corresponding to occurrence points of the 13Guibourtia specieswere used tomodel the species environmentalniches.To this end, in addition to thegeographical coordinatesofthe281specimensusedforthemorphologicalcharacterization,weextractedgeoreferenceddataofreliablydeterminedspecimensfromKewherbarium(K)andNaturalisherbarium(L),givingusatotalof401‐presence records after removing duplicates. Climate and soildatawereextracted fromtheClimaticResearchUnit (CRU)of theUniversityofEastAnglia (Mitchell& Jones, 2005;New,Hulme,&Jones,1999)and“FAODigitalSoilMapoftheWorld,version3.6.”Bothclimaticandsoildatawereinterpolatedat0.5°spatialresolu‐tion.Toavoidredundancyinenvironmentalinformation,aprincipalcomponentanalysis (PCA)wasusedasadatareductiontechnique(Heikkinenetal.,2006).WechosetheleastcorrelatedvariablesthatbestexplainthedistributionofGuibourtiaspecies:monthlymeansoftemperature(°C),precipitation(mm),solarradiation(w/m2),relativehumidity (%), temperature range (°C), Potential evapotranspiration(mm),windspeed(m/s),andsoilpH.

2.4 | Phylogenetic data

Inapreviousstudy,webuiltahigh‐resolutionphylogenyofthegenusGuibourtiausingthewholechloroplastgenome(FigureS2,Tossoetal.,2018).Wereconstructedthephylogenyandperformedmolecu‐lardatingforthe13AfricanspeciesusingaBayesianMCMCanalysis

TA B L E 1  The45morphologicalcharacters(including15vegetative,and30reproductivecharacters)consideredforcharacterizingGuibourtia individuals

Morphological traits

Character states or measure‐ment unit in case of quantita‐tive variables

Vegetativecharacters

Positionofprimaryleafveins Submarginal,marginal,median

Apexleaflets Obtuse,acuminate

Glandsontheabaxialsideofthelimb

Absent,present

Limb Membranous,coriaceous,(sub)coriaceous

Petiolehairiness Glabrous,notglabrous

Glandonpetioles Absent,present

Stipule Obsolete,persistent

Sizeofstipules Absent,tiny,foliaceous

Numberofleafletsperleaf 1,2

Numberofsecondaryleafveinssuprabasilar

1,2,3,4,5

Numberofsecondaryleafveinsbasilar

1,2,3

Lengthofleaflets cm

Widthofleaflets cm

Lengthofacumen cm

Petiolelength cm

Reproductivecharacters

Inflorescenceposition Axillary,axillaryandterminal,terminal

Typeofinflorescence Panicle=>cob,cluster

Pedicel Absent,present

Bracts Obsolete,persistent

Bractshairiness Absent,present

Glandonbracts Absent,present

Bracteoles Obsolete,persistent

Shapeofbracts Orbicular,linear

Hairinessoftheexternalfaceofthebracts

Absent,present

Shapeofflowerbud Cylindrical,globular,ellipsoid

Hairinessbudontheexternalfaceofthesepals

Absent,present

Hairinessontheinnersideofthesepals

Absent,present

Glandonsepals Absent,present

Hairinessofdisc Absent,present

Pilosityofovary Glabrous,pilose

Stipeoftheovary Sessile,stiped

Hairinessofthestipe'sovary Glabrous,pilose

Typeoffruit Indehiscentfruit,dehiscentfruit

(Continues)

Morphological traits

Character states or measure‐ment unit in case of quantita‐tive variables

Glandonfruit Absent,present

Veinsonthefruitoutersurface Absent,present

Stipeofthefruit Absent,present

Arillus on the seed Absent,present

Lengthofsepals mm

Widthofsepals mm

Lengthofstipeoftheovary mm

Lengthofthefruit cm

Widthofthefruit cm

Thicknessofthefruit mm

Lengthofthefruitstipe cm

Numberofseedsperfruit 1,2

TA B L E 1   (Continued)

     |  5TOSSO eT al.

implemented in BEAST v1.8.2 (Bayesian Evolutionary Analysis bySamplingTrees;Drummond&Rambaut,2007).

2.5 | Morphological similarity

To describe the morphological similarity between species, weproceeded in the following steps. First, a morphometric distancematrixbetweenthe281analyzedspecimenswasobtainedbycal‐culatingtheGowerdistance(Gower,1971)usingtheRpackage“FD”(Laliberté,Legendre,&Shipley,2014).Second,weappliedaPCoA(namelyPCoA‐basedmorphologicaldistance)onthisdistancematrixto synthesizemorphological variation on four axes that explained85.19%ofthetotalvariation.Pearson'scorrelationcoefficientsbe‐tween originalmorphological variables and axis scoreswere usedto interpret each PCoA axis in terms of morphological gradients.Third,wecomputedthemeansandstandarderrors(SE)ofthePCoAscoresofthespecimensforeachspeciesalongeachaxis.Finally,wecomputedtheEuclideandistancebetweenthemeanspeciesscoresalongthefirstfourPCoAaxestodescribemorphologicaldistancesbetweenspecies.

2.6 | Environmental niche comparison

The environmental niche of specieswas described using a prin‐cipalcoordinatesanalysis,asdoneformorphologicaldata,tofa‐cilitate the analysis of themorphology‐niche relationship.Here,we considered the environmental variables extracted from the281specimensforwhichmorphologicaldatawereavailable.Wethen computed the Gower distance between specimens to per‐formaPCoAandkept the three first axesexplaining90.10%ofthe total variation inenvironmental variables.Themean speciesscoresalongtheseaxeswereusedtodescribetheenvironmentalniche of the species, and the Euclidean distance betweenmeanspeciesscoresalongthethreeaxesprovidednichedistancesbe‐tweenspecies.

Weconductedcomplementaryenvironmentalnichecomparisonanalysesusingthe401‐presencerecordstofurtherdescribethede‐greeofnichesimilarityandnicheoverlapbetweenspecies,applyingtheapproachesdevelopedbyBroennimannetal.(2012).First,tode‐terminethelimitsoftheareaoccupiedbythe401‐presencerecords(Burgman&Fox,2003),weapplied theα‐hullpolygons techniqueusing the R package Alphahull (Pateiro‐López & Rodrıguez‐Casal,2010). We then obtained the background data (Warren, Glor, &Turelli, 2010) for the 13 species taken together by using a bufferzoneof2,000kmaroundthisarea (except intheocean).Thebuf‐ferextentwasbasedonpreliminaryanalysisfollowingthemethodofVanDerWal,Shoo,Graham,andWilliams(2009)tooptimizethepredictive powerof the explaining factors.Webuilt severalmod‐els for each species from a sample, varying buffer extent. Then,theareaunder theROCcurve (AUC)wasplottedasa functionofbufferextentandweselected thebufferextentcorresponding totheminimalvaluestillproducinganappreciableAUCincrease.Thebackgroundwasusedtoperformasmoothingtechniquedescribed

inBroennimannetal.(2012),whichimplementedanordinationtech‐nique(PCA‐ent)thatdividestheenvironmentalspaceincellsandap‐pliedakerneldensityfunctiontodeterminethe“smoothed”densityofeachspeciesoccurrenceineachcell.

Todeterminethedegreeofsharedenvironmentalnichebetweenapairofspecies,wecalculatedanindexofnicheoverlap.ThisindexquantifiestheproportionofnichesharedbytheconsideredcoupleofspeciesandcanbecomputedbymeansoftheSchoener'sD statis‐tic(D=1–0.5∑|pX,i−pY,i| where pX,i and pY,istandfortheprobabil‐ityassignedbytheenvironmentalnichemodellingforspeciesXandY,respectively,tocelli;andthesumistakenoverallthecellsofthebidimensionalenvironmentalspace;Schoener,1968;Warren,Glor,&Turelli, 2008).D varies from0 (completedisjunction) to1 (fullyoverlappingniches).

TheSchoener'sDstatisticwasthenusedtoperformtwotestsofnichedifferentiation:nichesimilarityandnicheequivalencytestsasdetailed inBroennimannetal. (2012).ThecomparisonofnichesimilarityandnicheequivalencytestsisgiveninDataS1.Allanaly‐seswereperformedintheRplatform(RDevelopmentCoreTeam,2013).

2.7 | Testing phylogenetic signal in morphology and environmental niche

We inferredpairwisephylogeneticdistancebetween species [P](patristicdistance;function“cophenetic.phylo”inRpackage“ape”;Paradis,Claude,&Strimmer,2004)basedonthephylogenyoftheGuibourtia species. To test the existence of phylogenetic signal(Losos,2008),weusedtwomethods.First,forunivariatedata,wecomputedBlomberg'sK (Blomberg,Garland,& Ives,2003)usingthe function “phylosig” (R package “phytools”; Revell, 2012) forenvironmental niche and morphology separately using (a) spe‐ciesmeansofeachquantitativemorphometricandenvironmentalvariableand(b)meanspeciesscoresalongeachmorphologicalandnichePCoAaxis.Toincreasethepowerofthetest,weintegratedalso the estimated standard errors (SE) of the mean values foreachspeciesfollowingIves,Midford,andGarland(2007).Second,forpairwisedistances,weperformedManteltests(Mantel,1967)between phylogenetic distance and PCoA‐based morphologicalornichedistances.Weincreasedthepowerofthetestusingthesquarerootofphylogeneticdistancematrix(P1/2)assuggestedbyHardyandPavoine(2012).

2.8 | Assessing the morphology—environmental niche relationship

To assess whether environmental niche divergence (PCoA‐basednichedistance) andenvironmentalnicheoverlap (D statistic)wererelated to morphological divergence (PCoA‐based morphologicaldistance),weusedManteltests.Tofurtherassessthemorphology‐nicherelationshipamongGuibourtiaspecies,weappliedtwo‐blockpartial least squares (2B‐PLS;Rohlf&Corti, 2000) analysis to ex‐plorepatternsofcovariationbetweenthemeanspeciesscoresalong

6  |     TOSSO eT al.

thesamenumberofPCoAmorphologicalandPCoAenvironmentalnicheaxes.We retained tenaxes to summarize100%of the totalvariation in the two datasets. The Rvcoefficient(Robert&Escoufier,1976)wasusedtosummarizetheamountofcovarianceineachdata‐setthatisaccountedfortheotherdataset.Thesignificanceofall2B‐PLSsummarystatisticswasassessedviapermutationoftherowsofeach dataset.

2.9 | Morphological adaptation and niche divergence accounting for phylogenetic inertia

Furthermore,toverifywhetherthemorphology‐nicherelationshipisnotsimplyduetophylogeneticinertia,weperformedtwotypesofanalyses.First,partialManteltestexaminedwhetherenvironmen‐tal nichewas still related tomorphological traits after accountingfor phylogenetic relationships (using amatrix of phylogenetic rawdistance as covariate). Second,we employed the PhylogeneticallyIndependentContrasts(PIC)approach(Felsenstein,1985)totestthecorrelationbetweenspeciesscoresalongeachofthefirsttwomor‐phologicalPCoAaxeswitheachofthefirsttwonichePCoAaxes.PICanalyseswerealsodoneoneachmorphologicaltraitweightinginthefirsttwonicheaxesinordertoidentifythespecificmorpho‐logicaltraitsselectedbyeachnicheaxis.ThenullhypothesisforPICtestassumesnoevolutionarylinkbetweentwotraits(evolutionbypurelyrandomgeneticdriftwithoutselection).Alltheanalyseswereperformed in the R platform (R Development Core Team, 2013),using the functions “pic” (R package “ape”) and “mantel.partial” (Rpackage“vegan”).

3  | RESULTS

3.1 | Morphological variation among species and phylogenetic signal

ThefirsttwocomponentsofthePCoA‐basedmorphologicaldis‐tance explained 32.50% and 26.40% of the total variation, re‐spectively(Figure2).Theresultsshowedthatspeciesassociatedwithwet/moistclimates(G. ehie and G. arnoldianaexcepted)werecharacterized by the combination of the followingmain charac‐ters: largeand long leafletsexhibitingmarginalvenation,subco‐riaceous limb, apex of leaflet acuminate, axillary inflorescence,persistentandpilosebracteoles,andnopedicelsandnovenationon the fruit (Table S1).By contrast, the species associatedwithdryandseasonalclimatesshowedsmallleafletswithsubmarginalvenation,coriaceouslimb,apexofleafletobtuse,axillaryandter‐minalinflorescence,obsoleteandglabrousbracteoles,andvena‐tiononthefruit.

We detected significant phylogenetic signal for six out of 45morphological traits (Table S1). Overall morphological similaritywas significantly correlated tophylogenetic distance (Mantel test:r= .24,p= .031).ThistrendwasalsoconfirmedwhenBlomberg'sKtestwasperformedonPCoA‐basedmorphologicalaxes:Thefirstaxisshowedsignificantphylogeneticsignal(K=1.2,p=.017)whilethesecondonenot(K=0.80,p=.130).Amongthe11quantitativemorphologicaltraits,onlyfourdisplayedsignificantKvalues(num‐berofsecondarysuprabasilar leafveins, lengthof leaflets,petiolelength,andwidthofsepals;TableS1).

3.2 | Pattern of environmental niche evolution

The principal component analysis of environmental niche proper‐ties(PCA‐ent)appliedon401occurrencesdatagaveafirstaxisex‐plaining51.7%of the totalvariancegeneratedby theeight testedvariablesandmainlyloadedbymonthlymeanthermalamplitudeandmonthlymeansolar radiationtowardpositivevalues,andmonthlymeanpotentialevapotranspirationandmonthlymeanprecipitationtowardnegativevalues(Figure3).Thesecondaxisexplainedabout19.9%ofthevariationandwasmainlyloadedbymonthlymeanwindspeedandpHofthesoiltowardpositivevalues,andmonthlymeantemperaturetowardnegativevalues.AllrainforestspeciesshowedadistributionintheenvironmentalspacewithverylowvaluesonbothPCA‐entaxis1(around−4)andaxis2(around−1,exceptforG. leon-ensis),whichcorrespondtohighprecipitationandlowsolarradiation.Dryforestandwoodlandspeciesgenerallydisplayedhighestdensityatlessnegativevaluesalongaxis1(between−3and0)andtendedtodifferentiatealongaxis2(highestdensitybetween−1and1depend‐ingonspecies).Therightpartofaxis1(scores>0)wouldcorrespondtothedesertzoneswhereGuibourtiaisabsent.Similarly,foraxis2scores>2maycorrespondtolow‐temperatureareas),inappropriateforthesurvivalandgrowthofGuibourtiaspecies.

As expected, niche overlap values showed that most pairsof rainforest versus dry forest and woodland species occupy

F I G U R E 2  PrincipalCoordinateAnalysison45morphologicaltraits.Thesymbolsinblackrepresentspeciesassociatedwithwet/moistclimateswhereasthoseingrayillustratespeciesassociatedwithdryandseasonalclimates

     |  7TOSSO eT al.

different environmental niches (Figure 3, S2). The results high‐lighted a more marked overlap between the niches of speciesassociatedwithwet/moistclimatesthanbetweenspeciesassoci‐atedwithdryandseasonalclimatesasconfirmedbytheresultsofnicheequivalency(strictnicheidentity)andnichesimilaritytestsdetailedinDataS3.

Wedetected aphylogenetic signal for species environmen‐talnichethroughthecorrelationbetweenphylogeneticdistance

and PCoA‐based niche distance (Mantel test: r = .16, p = .04)ornicheoverlap (Mantel test: r=−.23,p= .025).However, theunivariate Blomberg'sK tests seemed to lack power (K = 0.99,p = .139; K=0.76,p= .895forthefirsttwonichePCoA‐basedniche distance axes, Figure 4) andwhen testing each environ‐mentalvariableinturn,onlymonthlymeanprecipitationshoweda marginally significant signal (Blomberg's K test: K = 0.87,p=.08;TableS2).

F I G U R E 3  Environmentalnichesofthe13AfricanGuibourtiaspeciesintwomainenvironmentalaxesproducedbytheprincipalcomponentanalysis(PCA‐ent)appliedon401occurrences.Foreachspecies,thegray‐to‐blackshadingrepresentsthegridcelldensityofthespecies'occurrences(blackbeingthehighestdensity).Thedashedandsolidlinesdelimitrespectively50%and100%oftheavailableenvironmentconditionsinthestudyarea.ThelastpanelpresentsthecontributionofvariablesforloadingthemainPCA‐entaxesandthepercentageofinertiaexplainedbyaxesoneandtwo.ThegeographicaldistributionofeachspeciesispresentedbeloweachPCA‐ent

* * * *

* * * #

# # # #

#

8  |     TOSSO eT al.

3.3 | Relationships between morphology and environmental niche without accounting for phylogenetic inertia

We found statistically significant correlations between speciesmorphological similarityandnicheoverlap (Mantel test: r=−.44,p < .001) or PCoA‐based niche distance (Mantel test: r = .67,p = .001). The 2B‐PLS analysis confirmed the strong associationbetweenmorphologyandenvironmentvariation:40%of theco‐variance inmorphology isexplainedbyenvironmentalconditions(Rv=0.40,p<.001).

3.4 | Morphological similarity and environmental niche resemblance accounting for phylogenetic inertia

The partial Mantel test between PCoA‐based niche distance andmorphological similarity using phylogenetic distances as a thirdmatrix of covariateswas statistically significant (r = .24,p < .030; Figure4).Thisresultmustbeinterpretedwithcautionbecausepar‐tialManteltestsareliberal(Guillot&Rousset,2013).Interestingly,whenusingonlyquantitativemorphologicaltraits,whichdisplayednosignificantphylogenetic signal (stipe lengthof theovary, thick‐ness of the fruit, lengthof sepals, lengthof acumen, andnumberofsecondary leafveinsat thebase,TableS1), therelationshipbe‐tweenmorphologicalsimilarityandPCoA‐basednichedistancewasstronger(Manteltestbetweenspecies:r=.28,p=.039)thanusingonlytraitsthatexhibitedphylogeneticsignal(Manteltestbetweenspecies:r=.12,p=.787),supportingthatpartofthetrait‐nichecor‐relationdoesnotresultfromphylogeneticinertia,asdetailedinPICanalyses.

Phylogenetically Independent Contrasts analysis of the fourpairwise comparisons between the two first PCoA‐based envi‐ronmentalnicheaxesandmorphologicalaxeswassignificantonlybetween Axis 2‐morphology and Axis 1‐ environmental nicheandmarginally significant betweenAxis 1‐morphology andAxis2‐ environmental niche (Table 2). The morphological PCoA axis2 gathered principally flower traits (inflorescence position, typeof inflorescence,pedicel,bracts,bractshairiness,andpilosityon

theexternalfaceofthesepalsofflowerbuds)andsomeleafletsandfruittraits(leafletsdimensionandvenation,veinsonthefruitsurface)while axis 1 essentially summarized fruit traits (type offruit,thicknessofthefruit,glandonfruit,stipeofthefruit),seeds(arillusontheseed)andsomeleafletsandflowerstraits(numberofleafletveins,aspectofbracteoles,andsepals;TablesS1andS3).Concerningtheenvironmentalnicheaxes,nichePCoAaxis1wascorrelated with solar radiation, relative humidity, precipitations,andtemperaturerange.Thisaxisexhibitsrainforest‐drywoodlandgradientwhereasaxis2wasrelatedtowindspeedandpotentialevapotranspiration.

4  | DISCUSSION

Inthisstudy,wedemonstratedthatmorphologicaldifferencesbe‐tweenAfricanGuibourtia species are significantly associatedwithnichedivergence, even after factoringout the impactof phyloge‐netic inertia.This trendwasparticularlyevidentwhen floral traitsandleafdimensionswereconsideredalongarainforest—drywood‐landgradient.Thesefindingsvalidateasetofscenarioshighlighting(a)bothphylogenetic inertiaandconvergentevolution(scenario4)and(b)anindependentevolutionofsomemorphologicaltraitsand

F I G U R E 4  PhylogenyofAfricanGuibourtiaspecieswithspeciesscoresalongthefirsttwoPCoAaxesofmorphologicaldata(Ma1andMa2)andenvironmentalnichedata(En1andEn2)

TA B L E 2  CorrelationbetweenmorphologicalandenvironmentalnichePCoAaxesbyphylogeneticallyindependentcontrasts

Niche optimum

Axis 1 (solar radia‐tion, relative humid‐ity, precipitations and temperature range)

Axis 2 (wind speed and poten‐tial evapotranspi‐ration)

Morphology

Axis1(fruit,seed,leafletandflowertraits)

r=−.33;p = .290 r=−.53;p=.075

Axis2(flower,leafletandfruittraits)

r = .69; p = .018 r=−.44;p=.144

Theboldvaluesshowsignificanceatthe5%level.

     |  9TOSSO eT al.

environmentalniche(scenario1).Theysuggestthatbothselectionforces and neutral ones contributed to the morphological diver‐gencesandsimilaritiesamongGuibourtiaspecies.

4.1 | Relationship between morphology, phylogeny, and environmental niche: evidence of adaptive forces

The recent incorporation of phylogenetic and functional infor‐mation into biogeographical analyses provides amore completeunderstandingofevolutionaryandecologicalprocesses (Molina‐Venegas, Aparicio, Slingsby, Lavergne, & Arroyo, 2015). In thisstudy,sixmorphologicaltraits,andonemorphologicalPCoAaxisrepresenting c. a third of the morphological variation, showedsignificant phylogenetic signals. This finding is consistent withworks of Lee and Collins (2001), Blomberg et al. (2003) andValverde‐Barrantes, Smemo, and Blackwood (2015) who foundthatsimilarityinmorphometricformslargelyreflectsgenealogicalrelationships.

WealsoprovedthatsomeGuibourtiamorphologicaltraitswerecorrelated with environmental features. This correlation was notonly due to phylogenetic inertia, as revealed by three arguments.(a)PICanalysiswassignificantbetweentwoparticularMandEaxes(PIC:r=.69,p=.018).(b)Themantelcorrelationbetweenenviron‐mentalnicheandmorphologytraits(Manteltest:r=−.44,p<.001)wasstrongerthantheonebetweenthemorphologyandphylogeny(Manteltest:r=.24,p=.031)butalsobetweenthephylogenyandenvironmentalniche(Manteltest:r=.16,p=.04).(c)Themorphol‐ogy‐environmentalnichecorrelationremainedsignificantwhenthemorphologicaltraitsexhibitingnophylogeneticsignalwereconsid‐ered(Manteltest:r= .31,p= .008). Infact,thetraitssignificantlycorrelatedtoanenvironmentalnicheaxisdidnotshowphylogeneticsignals.

The overall signal between morphology and environmentalniche was significant in rainforest species especially when onlythereproductivetraitswereconsidered. Itwasalsosignificant forwoodlandspecieswhenallthemorphologicalcharactersweretakeninto account together.Among species having dehiscent fruits andarillate seeds dispersed by animals, three are found in rainforests(G. tessmannii,G. pellegriniana,andG. leonensis)andoneinwoodlands(G. coleosperma).Exceptthelatterone,dryforestandwoodlandspe‐cieshaveindehiscentfruitswithnonarillateseedsdispersedeitherbywindorwater.ThisfindingwashighlightedbyTurner(2001)andChazdon,Careaga,Webb,andVargas(2003)andconfirmsthedom‐inationofanimaldispersedplantsinrainforest.

Besides, the relation between PCoA‐based niche distance andmorphologicaldissimilaritieswasstrongerforthePCoAaxesconsider‐ingprincipallyflowertraitsandleafletsdimensionalongtherainforest‐drywoodlandgradient,whichevokesselectionforcesactingonthesephenotypiccharacters.Rainforestspecies (e.g.,G. tessmannii*,G. pel-legriniana*,G. leonensis*etc.)havelongerandlargerleafletsthanwood‐landspecies(G. sousae#,G. conjugata#,G. carrissoana#etc.).Asspecificleafareaisamongthebestindicatortoassesstheadaptationofspecies

tolightconditionsandwaterstressforphotosynthesis(Hoffmannetal.,2005),arelationshipbetweenleavesizeandenvironmentalnicheisexpected.Specieslivinginrainforestsinvestmuchmoreinleafarea(adaptation for a competitive lightenvironment) in comparisonwithspeciesofmoreopenwoodlands,whilethelatterinvestmoreinrootdevelopmenttoimprovetheircaptureofsoilwaterduringthedrysea‐son(adaptationtowaterstress,Hoffmann&Franco,2003).Wealsoobserved that floral traits could be selected. Rainforest Guibourtia specieshaveaxillaryinflorescenceswhilethewoodlandspecieshaveaxillaryandterminalinflorescences.Herrera(1996)explainedthisphe‐nomenonbyanadaptationtoinsectpollinators.Thiskindofadapta‐tiondeservestobemoredeeplyinvestigatedinAfricanbiomes.

Eventhoughitisknownthateachspecieshasspecificmorpholog‐ical responsesdependingon theenvironmental conditions towhichitissubjectedto(Gratani,Meneghini,Pesoli,&Crescente,2003),ourresultsareamongthefewinAfricathatprovethatpartofthemor‐phological differences between congeneric tree species result fromenvironmentaladaptation.Thisisinlinewithsomestudiesconductedwithinvariousgenera,whichhavefoundthesameassociation(Cicero&Koo,2012;Fontanella,Feltrin,Avila,Sites,&Morando,2012;Fort,Jouany,&Cruz,2015).Ribeiro,Lloyd,Dean,Brown,andBowie(2014)also demonstrated that niches differences among species are anevolutionaryforceshapingdiversification,butstudiesdidnotclearlydemonstrate a direct correlation between morphology and envi‐ronmental niche. Besides, Couvreur, Porter‐Morgan, Wieringa, andChatrou (2011)byexaminingwhether speciationwasdominatedbynichechangesfoundthatadaptationtoclimaticdifferencesbetweensisters specieshasnotbeenamajordriverof speciation in treesofAfricantropicalrainforests.

However,asshowninourresults,therelationshipbetweenmor‐phologyandenvironmentalnichewasnotevidentforallmorpholog‐icaltraits.Whensomeleafletsandflowerstraits(numberofleafletveins, aspect of bracteoles and sepals) are considered along therainforest‐woodlandgradient,no significant linkwasdetectedbe‐tweenmorphologyandenvironmentalniche.Thisresultunderlinesthepossibleactionofgeneticdriftasanevolutionaryforce,althoughonecannotruleoutthehypothesisthatselectionactedagainbutinresponsetootherfactorsandclimate(Felsenstein,1985).

4.2 | Speciation and evolutionary hypotheses within the genus Guibourtia

Considering that geographical andecological factors could con‐tribute to species divergence, two main hypotheses could beproposed to further understand species evolution within thegenus Guibourtia. The first hypothesis is disruptive ecologi‐cal adaptation (Knouft, Losos, Glor, & Kolbe, 2006), which as‐sumes that ecological niches tend to diverge in near relatives,reducing interspecific competition (Lososet al., 2003). The sec‐ond one deals with the fragmentation of favorable habitat ofrelated species described by Acevedo,Melo‐Ferreira, Real, andAlves (2014). This hypothesis supposes that the ancestor of re‐latedspeciesoccupieda large rangethatbecamefragmented in

10  |     TOSSO eT al.

different allopatric refuges following past climatic fluctuations.Tosso et al. (2018) demonstrated thatGuibourtia started to di‐versify~14–24MyrandcontinueduntilLateMioceneandEarlyPlioceneasconfirmedbyFougère‐Danezan(2005).Thetwofirstdivergenceeventsthatoccurred~14.8–16Myr(Tossoetal.,2018)ultimately lead to three clades corresponding to the subgeneraGorskia, Guibourtia, and Pseudocopaiva described by Léonard(1949).Thisperiodcoincideswiththesecondmajorenvironmen‐talperturbationthatoccurredfromtheEarlytotheMid‐Miocene(23–15Mya): Humid vegetation disappeared in the Sahara, thecontinentmovednorthward, downpositioning theEquator, andthe rainforest belt shifted southward (Maley, 1996). During theLateMioceneandEarlyPliocene,andespeciallywiththeonsetoftheglacial–interglacialcyclesoftheQuaternary,theAfricanrain‐forestprobablybecamefragmentedwhiledrierecosystemtypes(dry dense forest, woodland, and savannah) expanded (Dupont,Rommerskirchen,Mollenhauer,&Schefuß,2013).Ourresults, incombinationwiththeabove‐mentionedwork,couldhelphypoth‐esize that the common ancestor ofGuibourtia species probablyoccupied a large range in Africa. However, the common ances‐tormayalsohavehada restrictedareaandspeciationoccurredmainlyduringgeographicalexpansions.

Inadditiontothetwomainhypothesesmentionedpreviously,it is important to underline three important points. First, eachsubgenus or clade (Figure S2) presents awide distribution fromWestAfricatoSouthernAfricaviaCentralAfrica.Thus,thereisnobiogeographic signal between these three clades. Second, thereis a diversityof configurationsbetween sister species: (a) Therearecasesofparapatricdistributionalongenvironmentalgradientsthatmaysuggestecological speciation (G. demeusei vs. G. carris-soana and G. coleosperma vs. G. tessmannii),(b)alsocasesofpara‐patrictosympatricdistributionwithoutstrongdivergenceofnichesuggestingallopatricspeciationwithsecondarycontact(G. conju-gata—G. schliebenii; G. arnoldiana—G. ehie), and (c)acaseofallo‐patric distribution,West Africa versus Southern Africa, withoutstrongnichedivergence(G. sousae vs. G. copallifera,butwhichdi‐vergedformerly).Third,itseemsthathabitattransitionsoccurredas well from dry environments to moist forest environments(G. demeuseibasedonphylogenyandthedryenvironmentoccu‐piedbyG. carrissoana,G. copallifera,and G. sousaeinthesubgenusGuibourtia)asintheoppositedirection(caseofG. coleosperma in subgenusPseudocopaiva,andofthecladeformedbyG. conjugata and G. schliebeniiinthesubgenusGorskia).Basedonalltheseob‐servations,itisdifficultwiththecurrentdatatofavoraparticularmodeofspeciationwithinAfricanGuibourtiataxaalthoughacer‐tainconvergentevolutioninmorphologicaltraits.

5  | CONCLUSION

This study is the first in Africa that clearly demonstrated thatsimilarphenotypesevolveindependently indifferent lineagesbyusing new genomic data set and environmental nichemodelling

techniques with morphological characterization using herbariacollections.We characterized the main environmental variablesthatconstrainthepotentialdistributionofAfricanGuibourtiaspe‐cies.Wealsoassessedthediversityandsimilarityoftheenviron‐mental niches of these species and demonstrated phylogeneticsignalsofenvironmentalniche,atsomemorphologicaltraitsandsignificantcorrelationbetweennichedivergenceandmorphologi‐caldivergence,evenafteraccountingforphylogenetic inertia,atleast for someof their respective dimensions. These results arecongruentwiththescenario4(Figure1,Sc.4)suggestthatconver‐gent evolutionhasoccurred.The significantdifferences in envi‐ronmentalnichesspacesalsoreflectthemorphologicaldistanceswithinAfricanGuibourtia species due both to neutral processes(e.g.,drift)andselectionforcesatacertainlevel.

The demonstrated link betweenmorphology and the environ‐mental niche for AfricanGuibourtia species could also serve as abasis for predicting long‐term phenotypic changes at the specieslevel in the genus.

ACKNOWLEDG MENTS

ThisworkwasfinanciallysupportedbytheFondspourlaFormationàlaRecherchedansl'Industrieetl'Agriculture—FondsNationalpourlaRechercheScientifique(FRIA‐FNRS,Belgium),theFondsNationalpour laRechercheScientifique (F.R.S.‐FNRS,grantT.0163.13),andthe DynAfFor project (funded by FFEM‐AFD). We thank LouisFrançoisandMarieDuryforprovidingtheclimaticdataaswellasallherbariummaterial; collectors,DanielP.Silva,andBrunoVilelafortheirvariouscontributions.WealsoacknowledgeNature+foritssupportandNationalMuseumofNaturalHistoryinParis(France),Botanic Garden Meise (Belgium), Royal Botanic Garden Edinburg(U.K.),UniversityofNairobiHerbarium (Kenya), andHerbariumofMaputo (Mozambique) for permitting sampling and studying theherbariumcollections.

CONFLIC T OF INTERE S T

None declared.

AUTHOR CONTRIBUTIONS

F.T., J‐L.D.,K.D., andO. J.H.conceived the ideasanddesignedmethodology;F.T.,A.H.,C.D.,andH.A.contributedtothedatacollection;F.T.analyzedthedataunder thesupervisionofJ‐L.D.andO.J.H.;andF.T.,J‐L.D.,K.D.,A.F.,A.H.,C.D.,H.A.,P.S.,andO. J.H.wrote themanuscript.All authors read and approved thefinalmanuscript.

DATA AVAIL ABILIT Y S TATEMENT

Datasourceusedtobuildahigh‐resolutionphylogenyiscitedinthetext.Morphologicalandclimatedataaswellassampling locationsareavailableviaDryadhttps://doi.org/10.5061/dryad.cjsxksn1r.

     |  11TOSSO eT al.

ORCID

Félicien Tosso https://orcid.org/0000‐0002‐7228‐9961

R E FE R E N C E S

Acevedo,P.,Melo‐Ferreira,J.,Real,R.,&Alves,P.C.(2014).EvidencefornichesimilaritiesintheallopatricsisterspeciesLepus castroviejoi and Lepus corsicanus. Journal of Biogeography,41,977–986.

Aubreville,A.(1970).Légumineuses‐Césalpinioı ̈ dées.Flore du Cameroun,9,339.

Blomberg, S. P.,Garland, T.,& Ives,A. R. (2003). Testing for phyloge‐netic signal in comparative data: Behavioral traits aremore labile.Evolution, 57, 717–745. https://doi.org/10.1111/j.0014‐3820.2003.tb00285.x

Broennimann, O., Fitzpatrick, M. C., Pearman, P. B., Petitpierre, B.,Pellissier, L., Yoccoz, N. G., … Guisan, A. (2012). Measuring eco‐logical niche overlap from occurrence and spatial environmentaldata. Global Ecology and Biogeography, 21, 481–497. https://doi.org/10.1111/j.1466‐8238.2011.00698.x

Burgman,M.A.,&Fox,J.C.(2003).Biasinspeciesrangeestimatesfromminimum convex polygons: Implications for conservation and op‐tionsforimprovedplanning.Animal Conservation,6,19–28.https://doi.org/10.1017/S1367943003003044

Chazdon,R.L.,Careaga,S.,Webb,C.,&Vargas,O.(2003).Communityandphylogeneticstructureofreproductivetraitsofwoodyspeciesinwettropicalforests.Ecological Monographs,73,331–348.https://doi.org/10.1890/02‐4037

Cicero, C., & Koo, M. S. (2012). The role of niche divergence andphenotypic adaptation in promoting lineage diversificationin the Sage Sparrow (Artemisiospiza belli, Aves: Emberizidae).Biological Journal of the Linnean Society, 107, 332–354. https://doi.org/10.1111/j.1095‐8312.2012.01942.x

Cody, M., & Mooney, H. A. (1978). Convergence versus nonconver‐gence in Mediterranean‐climate ecosystems. Annual Review of Ecology and Systematics,9,265–321.https://doi.org/10.1146/annurev.es.09.110178.001405

Couvreur,T.L.P.,Porter‐Morgan,H.,Wieringa,J.J.,&Chatrou,L.W.(2011). Little ecological divergence associated with speciation intwoAfricanrainforesttreegenera.BMC Evolutionary Biology,11(1),296.

Desdevises,Y.,Legendre,P.,Azouzi,L.,&Morand,S.(2003).Quantifyingphylogenetically structured environmental variation. Evolution, 57,2647–2652.https://doi.org/10.1111/j.0014‐3820.2003.tb01508.x

Diniz‐Filho,J.A.F.,deSant'Ana,C.E.R.,&Bini,L.M.(1998).Aneigen‐vectormethodforestimatingphylogenetic inertia.Evolution,52(5),1247–1262.

Drummond,A.J.,&Rambaut,A.(2007).BEAST:Bayesianevolutionaryanalysisbysamplingtrees.BMC Evolutionary Biology,7,1.https://doi.org/10.1186/1471‐2148‐7‐214

Dupont,L.M.,Rommerskirchen,F.,Mollenhauer,G.,&Schefuß,E.(2013).MiocenetoPliocenechangesinSouthAfricanhydrologyandvegeta‐tioninrelationtotheexpansionofC4plants.Earth and Planetary Science Letters,375,408–417.https://doi.org/10.1016/j.epsl.2013.06.005

Felsenstein, J. (1985). Phylogenies and the comparative method. The American Naturalist,125,1–15.https://doi.org/10.1086/284325

Fontanella, F. M., Feltrin, N., Avila, L. J., Sites, J.W., &Morando,M.(2012).Earlystagesofdivergence:Phylogeography,climatemodel‐ing,andmorphologicaldifferentiationintheSouthAmericanlizardLiolaemus petrophilus(Squamata:Liolaemidae).Ecology and Evolution,2,792–808.

Fort,F.,Jouany,C.,&Cruz,P.(2015).HierarchicaltraitsdistancesexplaingrasslandFabaceaespecies'ecologicalnichesdistances.Frontiers in Plant Science,6,63.

Fougère‐Danezan,M.(2005).Phylogénie moléculaire et morphologique des Detarieae résinifères (Leguminosae, Caesalpinioideae): Contribution à l'étude de l'histoire biogéographique des légumineuses.Toulouse3,329p.

Fougère‐Danezan, M., Herendeen, P. S., Maumont, S., & Bruneau,A. (2010). Morphological evolution in the variable resin‐produc‐ing Detarieae (Fabaceae): Do morphological characters retain aphylogenetic signal? Annals of Botany, 105, 311–325. https://doi.org/10.1093/aob/mcp280

Freckleton,R.P.,&Jetz,W.(2008).Spaceversusphylogeny:Disentanglingphylogeneticandspatialsignalsincomparativedata.Proceedings of the Royal Society B: Biological Sciences,276(1654),21–30.

Giarla, T. C., & Esselstyn, J. A. (2015). The challenges of resolving arapid, recent radiation: Empirical and simulated phylogenomics ofPhilippine shrews. Systematic Biology, 64(5), 727–740. https://doi.org/10.1093/sysbio/syv029

Gower,J.C.(1971).Ageneralcoefficientofsimilarityandsomeofitsprop‐erties. Biometrics,27,857–871.https://doi.org/10.2307/2528823

Grafen,A.(1989).Thephylogeneticregression.Philosophical Transactions of the Royal Society B, 326, 119–157. https://doi.org/10.1098/rstb.1989.0106

Gratani,L.,Meneghini,M.,Pesoli,P.,&Crescente,M.(2003).Structuraland functional plasticity of Quercus ilex seedlings of differentprovenances in Italy. Trees, 17, 515–521. https://doi.org/10.1007/s00468‐003‐0269‐8

Guillot, G., & Rousset, F. (2013). Dismantling the mantel tests.Methods in Ecology and Evolution, 4, 336–344. https://doi.org/10.1111/2041‐210x.12018

Guisan,A.,&Thuiller,W.(2005).Predictingspeciesdistribution:Offeringmorethansimplehabitatmodels.Ecology Letters,8,993–1009.https://doi.org/10.1111/j.1461‐0248.2005.00792.x

Hardy, O. J., & Pavoine, S. (2012). Assessing phylogenetic signal withmeasurementerror:Acomparisonofmanteltests,Blombergetal'.sKandphylogeneticdistograms.Evolution,66,2614–2621.https://doi.org/10.1111/j.1558‐5646.2012.01623.x

Heikkinen,R.K.,Luoto,M.,Araújo,M.B.,Virkkala,R.,Thuiller,W.,&Sykes,M.T.(2006).Methodsanduncertaintiesinbioclimaticenve‐lopemodellingunderclimatechange.Progress in Physical Geography,30,751–777.https://doi.org/10.1177/0309133306071957

Herrera,C.M.(1996).Floraltraitsandplantadaptationtoinsectpollina‐tors:Adevil'sadvocateapproach.InD.Llyod&S.C.H.Barrett(Eds.),Floral biology(pp.65–87).NewYork,NY:Chapman&Hall.

Hoffmann,W.A.,daSilvaJr,E.R.,Machado,G.C.,Bucci,S.J.,Scholz,F.G.,Goldstein,G.,&Meinzer,F.C.(2005).SeasonalleafdynamicsacrossatreedensitygradientinaBraziliansavanna.Oecologia,145,306–315.https://doi.org/10.1007/s00442‐005‐0129‐x

Hoffmann,W.A.,&Franco,A.C.(2003).Comparativegrowthanalysisoftropicalforestandsavannawoodyplantsusingphylogeneticallyindependentcontrasts.Journal of Ecology,91,475–484.https://doi.org/10.1046/j.1365‐2745.2003.00777.x

Ives, A. R.,Midford, P. E., & Garland, T. (2007).Within‐species varia‐tionandmeasurementerror inphylogeneticcomparativemethods.Systematic Biology, 56, 252–270. https://doi.org/10.1080/10635150701313830

Knouft,J.H.,Losos,J.B.,Glor,R.E.,&Kolbe,J.J.(2006).Phylogeneticanalysisof theevolutionof theniche in lizardsof theAnolis sagrei group.Ecology,87,S29–S38.

Labra,A.,Pienaar,J.,&Hansen,T.F.(2009).Evolutionofthermalphysiol‐ogyinLiolaemuslizards:Adaptation,phylogeneticinertia,andnichetracking.The American Naturalist,174(2),204–220.

Laliberté,E.,Legendre,P.,&Shipley,B.(2014).FD: Measuring functional diversity from multiple traits, and other tools for functional ecology. R package version 1.0-12.

Lee,D.W.,&Collins,T.M.(2001).Phylogeneticandontogeneticinflu‐encesonthedistributionofanthocyaninsandbetacyaninsinleaves

12  |     TOSSO eT al.

oftropicalplants. International Journal of Plant Sciences,162,1141–1153.https://doi.org/10.1086/321926

Léonard,J. (1949).Notulae systematicae IV (Caesalpiniaceae-Amherstieae africanae americanaeque). Bulletin du Jardin Botanique de l'État,Bruxelles/BulletinvandenRijksplantentuin,Brussel,383–408.

Léonard,J. (1950).Notulae systematicae IX. Nouvelles observations sur le genre Guibourtia (Caesalpiniaceae). Bulletin du Jardin botanique del'ÉtataBruxelles,20(Fasc.2),269–284.

Linder,H.P.(2014).TheevolutionofAfricanplantdiversity.Frontiers in Ecology and Evolution,2,38.

Losos,J.B.(2008).Phylogeneticnicheconservatism,phylogeneticsignalandtherelationshipbetweenphylogeneticrelatednessandecolog‐icalsimilarityamongspecies.Ecology Letters,11,995–1003.https://doi.org/10.1111/j.1461‐0248.2008.01229.x

Losos,J.B.,Leal,M.,Glor,R.E.,deQueiroz,K.,Hertz,P.E.,Schettino,L.R.,…Larson,A.(2003).NichelabilityintheevolutionofaCaribbeanlizard community. Nature, 424, 542–545. https://doi.org/10.1038/nature01814

Maley,J.(1996).TheAfricanrainforest–maincharacteristicsofchangesin vegetation and climate from the Upper Cretaceous to theQuaternary.Proceedings of the Royal Society of Edinburgh. Section B. Biological Sciences,104,31–73.

Mantel,N.(1967).Thedetectionofdiseaseclusteringandageneralizedregressionapproach.Cancer Research,27,209–220.

McLennan,D.A.,&Brooks,D.R.(1993).Thephylogeneticcomponentofcooperativebreedinginperchingbirds:Acommentary.The American Naturalist,141(5),790–795.

Mitchell,T.D.,&Jones,P.D.(2005).Animprovedmethodofconstruct‐ingadatabaseofmonthlyclimateobservationsandassociatedhigh‐resolution grids. International Journal of Climatology, 25, 693–712.https://doi.org/10.1002/joc.1181

Molina‐Venegas,R.,Aparicio,A.,Slingsby,J.A.,Lavergne,S.,&Arroyo,J.(2015).InvestigatingtheevolutionaryassemblyofaMediterraneanbiodiversityhotspot:Deepphylogeneticsignalinthedistributionofeudicotsacrosselevationalbelts. Journal of Biogeography,42,507–518.https://doi.org/10.1111/jbi.12398

Morales, E. (2000). Estimating phylogenetic inertia in Tithonia(Asteraceae): A comparative approach. Evolution, 54(2), 475–484.https://doi.org/10.1111/j.0014‐3820.2000.tb00050.x

Motta,P.J.,&Kotrschal,K.M.(1991).Correlative,experimental,andcom‐parative evolutionary approaches in ecomorphology. Netherlands Journal of Zoology, 42, 400–415. https://doi.org/10.1163/156854291X00414

New,M.,Hulme,M.,&Jones,P.(1999).Representingtwentieth‐centuryspace‐timeclimatevariability.PartI:Developmentofa1961–90meanmonthlyterrestrialclimatology.Journal of Climate,12,829–856.https://doi.org/10.1175/1520‐0442(1999)012<0829:RTCSTC>2.0.CO;2

Ortiz‐Medrano, A., Scantlebury, D. P., Vázquez‐Lobo, A., Mastretta‐Yanes,A.,&Piñero,D.(2016).Morphologicalandnichedivergenceof pinyon pines. Ecology and Evolution, 6, 2886–2896. https://doi.org/10.1002/ece3.1994

Pagel,M.(1994).Detectingcorrelatedevolutiononphylogenies:Agen‐eral method for the comparative analysis of discrete characters.Proceedings of the Royal Society of London B: Biological Sciences,255,37–45.

Paradis, E.,Claude, J.,&Strimmer,K. (2004).APE:Analysesof phylo‐genetics and evolution in R language. Bioinformatics,20, 289–290. https://doi.org/10.1093/bioinformatics/btg412

Pateiro‐López,B.,&Rodrıguez‐Casal,A. (2010).Generalizing the con‐vexhullofa sample:TheRpackagealphahull.Journal of Statistical Software,34,1–28.

Price, T. (1997). Correlated evolution and independent contrasts.Philosophical Transactions of the Royal Society of London. Series

B: Biological Sciences, 352, 519–529. https://doi.org/10.1098/rstb.1997.0036

R Development Core Team (2013). RA Lang Environ Stat Comput, 55,275–286.

Rato, C., Harris, D. J., Perera, A., Carvalho, S. B., Carretero,M. A., &Rödder,D. (2015). A combination of divergence and conservatismin thenicheevolutionof theMoorish gecko,Tarentola mauritanica (Gekkota: Phyllodactylidae). PLoS ONE, 10, e0127980. https://doi.org/10.1371/journal.pone.0127980

Revell,L. J. (2012).phytools:AnRpackage forphylogeneticcompara‐tivebiology(andotherthings).Methods in Ecology and Evolution,3(2),217–223.https://doi.org/10.1111/j.2041‐210X.2011.00169.x

Ribeiro,Â.M.,Lloyd,P.,Dean,W.R.J.,Brown,M.,&Bowie,R.C.(2014).Theecologicalandgeographiccontextofmorphologicalandgeneticdivergenceinanunderstorey‐dwellingbird.PloS ONE,9(2),e85903.

Ricklefs,R.E. (1987).Communitydiversity:Relative rolesof local andregionalprocesses.Science,235(4785),167–171.

Robert,P.,&Escoufier,Y. (1976).Aunifyingtoolfor linearmultivariatestatisticalmethods:TheRV‐coefficient.Applied Statistics,25,257–265.https://doi.org/10.2307/2347233

Rohlf,F.J.,&Corti,M.(2000).Useoftwo‐blockpartialleast‐squarestostudycovariationinshape.Systematic Biology,49,740–753.https://doi.org/10.1080/106351500750049806

Schluter,D.(1988).Characterdisplacementandtheadaptivedivergenceof fincheson islands and continents.The American Naturalist,131,799–824.https://doi.org/10.1086/284823

Schoener, T. W. (1968). The Anolis lizards of Bimini: Resource par‐titioning in a complex fauna. Ecology, 49, 704–726. https://doi.org/10.2307/1935534

Soberón,J. (2007).GrinnellianandEltoniannichesandgeographicdis‐tributions of species. Ecology Letters, 10, 1115–1123. https://doi.org/10.1111/j.1461‐0248.2007.01107.x

Tosso,F.,Daïnou,K.,Hardy,O.J.,Sinsin,B.,&Doucet,J.‐L. (2015).LegenreGuibourtiaBenn.,untaxonàhautevaleurcommercialeetso‐ciétale(synthèsebibliographique).Biotechnologie, Agronomie, Société et Environnement,19,71–88.

Tosso,F.,Hardy,O.J.,Doucet,J.L.,Daïnou,K.,Kaymak,E.,&Migliore,J.(2018).EvolutionintheAmphi‐AtlantictropicalgenusGuibourtia(Fabaceae,Detarioideae), combiningNGSphylogenyandmorphol‐ogy.Molecular Phylogenetics and Evolution,120, 83–93. https://doi.org/10.1016/j.ympev.2017.11.026

Turner, I. M. (2001). The ecology of trees in the tropical rain forest. Cambridge,UK:CambridgeUniversityPress.

Valverde‐Barrantes,O.J.,Smemo,K.A.,&Blackwood,C.B.(2015).Finerootmorphology isphylogeneticallystructured,butnitrogen is re‐latedtotheplanteconomicsspectrumintemperatetrees.Functional Ecology,29,796–807.https://doi.org/10.1111/1365‐2435.12384

VanDerWal,J.,Shoo,L.P.,Graham,C.,&Williams,S.E.(2009).Selectingpseudo‐absencedataforpresence‐onlydistributionmodeling:Howfarshouldyoustrayfromwhatyouknow?Ecological Modelling,220,589–594.https://doi.org/10.1016/j.ecolmodel.2008.11.010

Warren, D. L., Glor, R. E., & Turelli, M. (2008). Environmental nicheequivalency versus conservatism: Quantitative approachesto niche evolution. Evolution, 62, 2868–2883. https://doi.org/10.1111/j.1558‐5646.2008.00482.x

Warren,D.L.,Glor,R.E.,&Turelli,M.(2010).ENMTools:Atoolboxforcomparativestudiesofenvironmentalnichemodels.Ecography,33,607–611.https://doi.org/10.1111/j.1600‐0587.2009.06142.x

Westoby,M., Leishman,M. R., & Lord, J.M. (1995). Onmisinterpret‐ing the ‘phylogenetic correction’. Journal of Ecology, 83, 531–534. https://doi.org/10.2307/2261605

Wiens,J.J.,&Graham,C.H.(2005).Nicheconservatism:Integratingevo‐lution,ecology,andconservationbiology.Annual Review of Ecology,

     |  13TOSSO eT al.

Evolution, and Systematics, 36, 519–539. https://doi.org/10.1146/annurev.ecolsys.36.102803.095431

SUPPORTING INFORMATION

Additional supporting information may be found online in theSupportingInformationsection.

How to cite this article:TossoF,DoucetJ‐L,DaïnouK,etal.HighlightingconvergentevolutioninmorphologicaltraitsinresponsetoclimaticgradientinAfricantropicaltreespecies:ThecaseofgenusGuibourtiaBenn..Ecol Evol. 2019;00:1–13. https://doi.org/10.1002/ece3.5740