The effects of two abundant ant species on soil nutrients and seedling recruitment in Brazilian Atlantic Forest

(1)

REVISTA

BRASILEIRA

DE

Entomologia

AJournalonInsectDiversityandEvolution

w w w . r b e n t o m o l o g i a . c o m

Biology,

Ecology

and

Diversity

The

effects

of

two

abundant

ant

species

on

soil

nutrients

and

seedling

recruitment

in

Brazilian

Atlantic

Forest

Fábio

Souto

Almeida

a,∗

_,

_Luciana

_Elizalde

b

_,

_Leticia

_Maria

_Souto

_Silva

c

_,

_Jarbas

_Marc¸

_al

_Queiroz

d

a_Universidade_Federal_Rural_do_Rio_de_Janeiro,_Instituto_Três_Rios,_Departamento_de_Ciências_do_Meio_Ambiente,_Três_Rios,_RJ,_Brazil b_Grupo_LIHO,_Laboratorio_Ecotono,_{INIBIOMA-CONICET,}_UNCOMA,_Bariloche,_Argentina

c_Universidade_Federal_de_Uberlândia_(UFU),_Departamento_de_Biologia,_MG,_Brazil

d_Universidade_Federal_Rural_do_Rio_de_Janeiro,_IF,_Departamento_de_Ciências_Ambientais,_Seropédica,_RJ,_Brazil

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received30December2018 Accepted5August2019 Availableonline14August2019 AssociateEditor:RicardoSolar Keywords: Clidemia Communitystructure Ecosystemengineers Insurancehypothesis Miconia

a

b

s

t

r

a

c

t

Antscaninﬂuencesoilfertilityandthespatialdistributionofseeds,withpossibleeffectsonseedling recruitment.TheantspeciesPachycondylastriataFr.Smith,1858andOdontomachuschelifer(Latreille, 1802)co-occurinmanyforestareasintheNeotropics.Weassessedsoilfertilityandseedbankstructure insoilsamplescloseanddistant(control)fromantnestsinforestfragments.Wealsoassessedtherichness andabundanceofseedlingsonnestsandcontrolsites.Insoilsamplesfromantnests,theconcentration ofphosphorusandpotassiumwererespectively55.6%and36%higherthanincontrolsites.Aluminium was11–15%lowerinsoilsamplesfromantnests.Inthegreenhouse,soilsfromantnestshadhigherplant abundanceandspeciesrichness,butthesamespeciescompositionincomparisonwithcontrolsites. AlthoughmoreplantsemergedfromsoilsamplesofO.chelifernests,intheﬁeld,thedensityandrichness ofseedlingsweresimilarforthetwoantspeciesstudied.Seedlingsinthenestsiteswere,onaverage, 1.8timesmoreabundantand1.6timesricherinspeciesthanincontrolsites.Ourresultsshowedthat antspeciescanplayakeyroleinseedlingrecruitmentinforestfragments,whereotheranimalswith equivalentandpositiveeffects,suchasmammals,aremissing.

Introduction

Seeddispersalandtheavailabilityofsuitablesitesforseedling

germinationandgrowthaffecttheseedlingrecruitmentprocess

(LeckandParker,2008).Theavailabilityofsuitablesitesdepends onseveralabioticfactors,suchastemperature,light,water,andsoil nutrientconcentration(Bruna,2002;Salinas-Pebaetal.,2014),as wellasbioticfactors,suchasseedpredators,herbivores,pathogens, andsoilbioturbators(AugspurgerandKelly,1984;Brussaard,1997; Benitez-Malvidoetal.,1999;ZamoraandMatias,2014).Many

ani-malsmoveseedsfromparentplantsorsoilsurfacetobettersites

for germinationand toseed banks (Chambers and MacMahon,

1994).Insoilseedbanks,viableseedsremainburiedandprotected

frompredatorsforsometime,waitingforfavourableconditions

togerminate(EscobarandCardoso,2015).Therefore,thedigging

andburrowingactivitiesofanimalsthatusefruitsandseeds,such

∗ Correspondingauthor.

E-mails:fbioalmeida@yahoo.com.br(F.S.Almeida),luelizalde@gmail.com (L.Elizalde),leticiamariasouto@hotmail.com(L.M.Silva),jarquiz@gmail.com (J.M.Queiroz).

mammalsandants,canbepositiveforplantrecruitmentthrough

theaccumulationandprotectionofseedsinseedbanks(O’Grady

etal.,2013;PiresandGaletti,2013).

Theantsareabundantorganismsintropicalforests(Fittkauand Klinge, 1973)and can directlyand indirectly affecttheprocess

ofseedlingrecruitment,thus,contributingtoforestsuccessional

dynamics (Folgarait, 1998).By carrying fruitsand seeds tothe

nests,antsactasimportantseeddispersers(PassosandOliveira,

2004),improvingthemovementofseedstosuitableandprotected

sites.Antscanalsoaffectthephysicalstructureandchemical com-positionofsoils(Farji-BrenerandWerenkraut,2017).Duringthe

constructionofnests,alargeamountof soilparticlesismoved,

alteringsoilphysicalstructureandformingbioporesthatincrease waterinﬁltrationrates(LobryDeBruynandConacher,1994).There isevidencethatthesoilnearantnestshasahighernutrient

con-centrationthanadjacentsoils(Farji-BrenerandGhermandi,2000;

Cammeraatetal.,2002;Folgaraitetal.,2002;PassosandOliveira, 2002,2004).Thisincreaseinsoilfertilityisprobablyduetothe behaviourofantscarryingplantmaterialtothenest(Briese,1982). Thus,theeffectsofantsonsoilcharacteristicsandplantcommunity

mayvaryaccordingtofoodpreferences,nestpopulation

character-istics,andneststructure,whichvaryamongantspecies.

https://doi.org/10.1016/j.rbe.2019.08.001

(2)

PachycondylastriataFr.Smith,1858andOdontomachuschelifer

(Latreille,1802)areabundantandecologicallyimportantspecies

inNeotropicalforests(Almeidaetal.,2013;AlmeidaandQueiroz,

2015).They have similar body sizes (head width: ca. 2.7mm;

Almeidaetal.,2013).Pachycondylastriatamakesnestscomposed

ofseveralshallowchambers,whereasO.chelifermakesdeepnests

(Bottcheretal.,2016).Althoughbothantspeciesareclassiﬁedas ground-dwellinggeneralistpredators(Grocetal.,2013),O. che-liferhasagreaterpropensitytointeractwithfruits(Bottcheretal., 2016).Bothspeciescantransportwholefruitsortheirﬂeshy

por-tionstothenest and discardseeds intact(Passosand Oliveira,

2002;PizoandOliveira,2000).ThematurenestsofO.chelifer

usu-allyshowsalargernumberofindividualsthanthoseofP.striata

(Medeiros1997;Guimarãesetal.,2018),whichcanresultinmore activityoftransportationofanimalandplantpartstothenestsof O.chelifer.

In the present study, we tested two hypotheses about the

relationshipamongants,plants,andsoilcompositioninforest frag-mentsofAtlanticForest,south-easternBrazil.(1)Soilfertilityand

speciesabundanceand richnessof seedandseedlingbanksare

higheronneststhaninadjacentareas.(2)Soilfertilityandplant

community variables are higheron nestsof O. chelifer than on

thoseofP.striata,duetodifferencesinfoodpreferencesandcolony characteristics.Toinvestigatethesehypotheses,weassessed infor-mationonsoilfertility,soilseedbank,andseedlingdensityinareas closetoanddistantfromnestsofP.striataandO.chelifer.

Materialsandmethods

Sitedescriptions

Thestudyplotswerelocatedinthreeareasofsemi-deciduous

tropicalforestnearthecityofVassouras,stateofRiodeJaneiro, south-easternBrazil.Theﬁrstplotwaslocatedina6haforest frag-ment(22◦2741S;43◦3857W);thesecondhadanareaof36ha (22◦273S;43◦3840W),andthethird,780ha(22◦2711S; 43◦3947W).Nativeforestscorrespondtoapproximately28%of thestudyregion(Pereiraetal.,2017).Theaveragetemperatureis

18◦Cinthecoldestmonthand23◦Cinthewarmestmonth.The

averageannualrainfallis1280mm(Francelinoetal.,2012).The averagealtitudeis600ma.s.l.,andtheregionalclimateisclassiﬁed asCWa,accordingtotheKöppensystem(Alvaresetal.,2013).The mainsoiltypeintheregionisHaplicCambisol(Pereiraetal.,2017). Datacollection

Weselectedtenmaturenests(i.e.,withaconsiderableamount

ofsoilexcavatedoutsideandhighantactivity)ofeachantspecies (O.cheliferand P.striata)in each ofthethree forestfragments.

Wecollectedsoilsamplestomeasuresoilchemicalpropertiesand

makeagerminationexperiment,andrecordedseedlingdensityand

richnessinapaireddesign,comparingnestswithcontrolsites10m distantfromthenest.

Soilchemicalanalysis

Ineach nestorcontrolsite (10mdistantfromthenest),we

collectedsixsubsamplesof5cmindiameterand10cmindepth

(193cm3_),_extracted_with_a _hollow_metallic_structure._Then, _we

transferredthreesubsamplestoplasticbagsintheﬁeldandtook

themtothelaboratoryforchemicalanalysis.WemeasuredsoilpH,

theconcentrationofavailableAl,Ca,K,Mg, andP,and the

per-centageoforganiccarbonandsaturationwithbases(V%),whichis

consideredaproxyforsoilfertility(Fialhoetal.,2006),following EMBRAPA(1997).

Seedbankandgerminationexperiment

Tostudytherelationshipbetweenantsandsoilseedbank,the

remainingthreesubsamplescollectedfromnestsofO.cheliferand

P.striata,andtheirrespectivecontrolsites,werecombinedand

spreadonplastictraysarrangedatrandominagreenhouse.The

traycontainingnestsoilwasplacedbesidethetraywithsoilfrom itscontrolsite.Inthegreenhouse,wemonitored120samplesfor14

consecutiveweeks.Duringthisperiod,werecordedspecies

abun-danceandrichnessoftheplantsgerminatedpersample.Everyten

days,werandomlychangedthepositionofthetraysonthebenchof

thegreenhouse.Tofacilitateplantidentiﬁcation,wetransplanted

someindividualstolargerpots,andmonitoredtheirdevelopment

there.Wealsocomparedgerminatedplantswithspecimensofthe

localherbarium(HerbariumoftheDepartmentofBotanyofthe

FederalRuralUniversityofRiodeJaneiro).

Seedlingcommunity

We recordedseedlingdensity and richness(5–30cm tall)in

50×50-cmplotsplacedoverthenestsofthetwoantspeciesand

theirrespectivepairedcontrolsites(N=10foreachantspeciesper forestfragment).

Monitoringantactivity

Antactivityisimportantbecausemoreactivityimpliesmore

seedscomingtothenest,resultingindifferencesbetweennests.We observedfor10minthenumberofantsattractedtobiscuit/sardine

baitsinthemorning(9:00–12:00)andafternoon(13:00–16:00)in

thesamemarkednests.Thebaitsweresetat15cmfromthenest

entrance.Weusedthemeannumberofantsfoundinthemorning

andafternoonasanestimationofantactivityinthesamplednests. Statisticalanalyses

Asourdatahaveanestedstructure(i.e.,thesoilfromnestsand controlsiteswassampledinpairsandreplicatedinthreedifferent forestsites),weusedmixed-effectsmodels.Webuiltatotalof12

models,changingthedependentvariablebutmaintainingthesame

independentvariablesandtherandomstructure.Theindependent

variablesweresoiltype(nestorcontrol),antspecies(P.striataorO. chelifer),theirinteraction,andantactivity.Weincludeddependent variablesrelatedtosoilchemistryandplantcommunity.Forthesoil

chemistry,weusedaluminiumconcentration,soilpH,percentage

ofsaturationwithbases,calcium,magnesium,phosphorus,

potas-sium,andorganiccarbonconcentrationasdependentvariablesin

separatemodels.Weusedplantspeciesrichnessandabundance

ofseedlingstotesttheeffectofantnestsonseedlingsandthesoil

seedbank(intheﬁeld),aswellasforthesoilseedbank

(green-houseexperiment).Logarithmicorsquare-roottransformationsof

dependentvariableswereusedwhennecessarytonormalizemodel

residuals.

Therandomstructureincludedsoiltype(nestorcontrol),nested

withinforestfragments.ThemodelselectionfollowedZuuretal.

(2009),i.e.,westartedwithamodelinwhichtheﬁxedcomponent containedalltheexplanatoryvariablestoﬁndtheoptimalstructure

oftherandomcomponent.Incasewefoundtheoptimalrandom

structure,weselectedtheﬁxed-effectcomponentsinthemodel,

usinglikelihoodratiotests(Zuuretal.,2009).

Foranalysingspeciescompositionofthegerminatedseedsinthe

greenhouseexperiment,weusedPERMANOVA,asimplemented

intheAdonisroutineoftheveganlibrary(Oksanenetal.,2013), accordingtosoiltype(nestvscontrol)orantspecies(O.chelifervs P.striata),separatelyforeachforestfragment.Wealsoanalysedthe

(3)

Table1

ResultsofmixedeffectmodelsforsoilchemicalcompositionandtheplantcommunityparametersinthesamplesofPachycondylastriataandOdontomachuschelifernestsas wellascontrolsites;†Controlsoilisthereference;*Odontomachusisthereference.

Nestvscontrolsoil† Antspecies* Antactivity

Dependentvariable Transformation used

Intercept Parameter L.Ratio P-value Parameter L.Ratio P-value Parameter L.Ratio P-value

Aluminium none 1.72 −0.19 4.92 0.03 — — >0.05 — — >0.05 Magnesium sqrt 1.06 0.14 6.44 0.01 — — >0.05 — — >0.05 Calcium log 0.28 0.07 7.67 0.006 — — >0.05 — — >0.05 %basesaturation(V) sqrt 5.03 — — >0.05 — — >0.05 — — >0.05 Phosphorous log 0.79 0.11 8.08 0.005 — — >0.05 — — >0.05 Potassium log 1.67 0.1 13.9 <0.001 — — >0.05 — — >0.05 pH log 0.54 — — >0.05 — — >0.05 — — >0.05 Organiccarbon sqrt 1.6 — — >0.05 — — >0.05 — — >0.05

Seedlingrichness log 0.83 0.17 50.01 <0.001 — — >0.05 — — >0.05

Seedlingabundance log 9.97 2.44 32.33 <0.001 — — >0.05 0.2 12.07 <0.001

Seedrichness sqrt 1.86 0.3 11.42 <0.001 — — >0.05 — — >0.05

Seedabundance log 0.81 0.15 8.65 0.003 −0.11 5.23 0.02 — — >0.05

antnestsandcontrolsitesinthegreenhouseexperimentusinga Chi-squaretest.

Statisticalanalysesandgraphicswerecarriedoutinthe statis-ticalsoftwareR(RDevelopmentCoreTeam,v.3.3.1)usingnlme, vegan,andggplot2packages.

Results

Samples from ant nests had different chemical properties, exceptsoilpH,percentageofsaturationwithbases,andorganic carbonpercentage(Table1).Theconcentrationofcalcium,

mag-nesium,phosphorus,andpotassiumwerehigherinantnestsoils

thanincontrolsites.Aluminiumcontent,though,wassigniﬁcantly

higherincontrolsites(Table1,Figs.1and2).Theconcentrationof

phosphorusandmagnesiumwere64%and32%higher,respectively,

inO.cheliferneststhanincontrolsites.Potassiumandmagnesium

were36%and30%higher,respectively,inP.striataneststhanin

controlsites.Ontheotherhand,aluminiumwas15%and11%lower,

respectively,inO.cheliferandP.striataneststhanincontrolsites.

Differencesbetweenantspeciesintheconcentrationofchemical

elementswerenotstatisticallysigniﬁcant(Table1).Differencesin antactivitybetweenP.striataandO.cheliferwerenotstatistically

signiﬁcant(Wilcoxonranksumtest;W=399.5;P=0.573).

Theanalysisof theseed bankin antnests andcontrol sites

resultedinthegerminationof980seedsof46plantspecies.We

countedatotalof509seedlings,withaminimumof1anda

maxi-mumof23persample.Seedlingspeciesrichnesspersamplevaried

from1to6.Themedianseedlingabundanceandrichnesswere,

respectively,66%and50%higherinantneststhanincontrolsites.

However,differencesinseedlingabundanceandrichnessbetween

nestsofthetwoantspecieswerenotstatisticallysigniﬁcant.We

foundapositiveandsigniﬁcantrelationshipbetweenantactivity

andseedlingabundance,butnotforspeciesrichness(Table1).

SeedlingabundanceinP.striatanestswas59%higherthanin

controlsites,and inO. chelifernests,it was29%higherthan in

controlsites.Plantspecies richnessand numberof germinating

plantsweresignificantlyhigherinantnestsoils(Table1). Differ-encesbetweenantspecieswerenotstatisticallysignificantforplant speciesrichness,butweresignificantforabundance,duetoahigher

abundanceofgerminatedseedsinsamplesfromnestsofO.chelifer

(Table1,Fig.2).

Plantspeciescomposition,analysedwitha PERMANOVA,did

notdifferbetweennestandcontrolsoilorantspeciesinthethree

forestfragments(P>0.05).Amongthesevenmostcommonplant

speciesrecordedintheseedbank(80%ofseedlings),onlyonewas

signiﬁcantlymorefrequentinsamplesfromcontrolsites,whereas

threepredominatedinsamplesfromantnests(Chi-squaredtest;

Table2).

Discussion

Wefoundstrongsupporttoourhypothesisthatposes

differ-encesonthesoilcharacteristicsand plantcommunity between

nest andcontrol sites.However,we only foundpartialsupport

toourhypothesisposingthatthesedifferenceswouldbe

depen-dentonantspecies.Soilcharacteristicsweresimilarinnestsof

O.cheliferandP.striata,withahigherconcentrationofcalcium,

magnesium,phosphorus,andpotassium,andalower

concentra-tionofaluminiumthanincontrolsites.Inaddition,theabundance

andrichnessofseedlingswerehigheronantneststhanin

con-trolsites,andsimilarforbothantspecies,accordingtothedata

collectedintheﬁeld.Plantrichnesswassimilarinnestsofboth

antspecies,accordingtodataobtainedfromtheseedbank.

How-ever,corroboratingoursecondhypothesis,wefounddifferencesin

seedbankabundancebetweenantspecies.SamplesfromO.

che-lifernestsshowedsigniﬁcantly moreseedlings emerging inthe

greenhouseexperimentthansamplesfromP.striatanests.

How-ever,plantspeciesrichnessandcompositionintheseedbankdid

notdifferbetweensoilsamplesfromthetwoantspecies.

Antstransporttotheirnestchambersanimalandplantparts

(leaves,ﬂowers,fruits,andseeds)thatwillincreasesoilfertility

oncetheydecompose(Cammeraatetal.,2002;Moutinhoetal.,

2003;Farji-BrenerandWerenkraut,2017).Moreover,physicaland

chemical effects emerging fromantcolony activitiescan

accel-eratelitter decomposition insidechambers (Wanget al.,2017).

Therefore,althoughwedonothavespeciﬁcinformationon

nest-siteselectionbythespeciesofourstudysystem,weassumethat

thedifferencesinsoilparametersclosetonestsresultfromsoil

bioturbationand foodaccumulation inside themcaused by ant

activities.Theconcentrationofaluminium candecreasein nest

soilsbythereleaseoforganicacidsbyplantmaterialsinsidenest

chambers(Madureiraetal.,2013).Despitedifferencesamongant

species, chemical characteristics of nest soil samples were not

speciesdependent.Therefore,thesedatadonotcorroborateour

secondhypothesis.However,itispossiblethatsoilsamplingatonly onedepthpreventedthedetectionofdifferencesinspecieseffects (Wangetal.,2017).

We foundhigher abundanceand diversity of plants in seed

banksamplesfromantnests,butplantspeciescompositionwas

similarinsoilsamplescloseanddistantfromantnests.Thefact

thatplantspeciescompositiondidnotdiffersuggeststhatboth

antspeciescollectfruitsandseedsaccordingtotheavailability

(4)

Odontomachus Pachycondyla

a

b

c

d

e

3 2 1 0 5 4 3 2 1 40 30 20 10 4 3 2 1 125 100 75 50 25 Control Nest Control Nest Control Nest Control Nest

Control Nest Control Nest

Control Nest

Fig.1. Chemicalelementsinsoilsamplesfromantnestsandcontrolsites. Con-centrationof(a)Aluminium,(b)Calcium,(c)Phosphorus,(d)Magnesium,and(E) Potassium.Al,Ca,andMgaregiveninCmolc/Kg,whereasPandKaregiveninmg/Kg. Solidhorizontalblacklinesintheboxesindicatethemedianvalues;thewhiskers representthesmallestandlargestnon-outlierobservations;dotsrepresentoutlying datapoints.

way(HorvitzandBeattie,1980;King,1977;Farji-BrenerandSilva, 1996).

Clidemiahirta(L.) D.Donand MiconiacalvescensDC.

(Melas-tomataceae),twoofthemostabundantspeciesinthegreenhouse

experiment, had higher frequencies in samples fromantnests.

Clidemia hirtais zoochoric,important for birds, and commonly

foundin earlysuccessionforests(Leitaoet al.,2010).The high

numberofgerminatedseedsofM.calvescensinthegreenhouse

experimentis relatedtothehighabundanceofitsdiasporesin

thehabitatsstudied(pers.obs.)andtheirusualcollectionbyant

Odontomachus Pachycondyla 20 15 10 5 0 6 4 2 30 20 10 0 10.0 7.5 5.0 2.5 0.0

Control Nest Control Nest

Control Control

Control

Nest Nest

Nest Control Nest

a

b

c

d

Fig.2. Seedlingabundance(a)andrichness(b)onnestsitesandcontrolsites;seed bankabundance(c)andrichness(d)inthegreenhouseexperiment.Solidhorizontal blacklinesintheboxesindicatethemedianvalues;thewhiskersrepresentthe smallestandlargestnon-outlierobservations;dotsrepresentoutlyingdatapoints.

species(DallingandWirth,1998;Farji-BrenerandMedina,2000;

Christianinietal.,2007).Theonlyplantspeciesthatshowedan

oppositepatterninnestandcontrolsites,withalowerabundance

nearnestsoil,wasPanicumpilosumSW,anon-forestgrassspecies. WesuggestthatP.pilosumseedsarelessattractivebecauseoftheir

smallsizeincomparisonwithotherseeds,andthelackofﬂeshy

tissue,whichseemstobepreferredbybothantspecies(Pizoand

Oliveira,2000).ItisalsopossiblethatthelownumberofP. pilo-sumplantsinnestsamplesisrelatedtoallelopathiceffectsofother seedsinthesoilofantnests,asitwasalreadydescribedforanother speciesofPanicum(Martinsetal.,2006).

(5)

Table2

Mostcommonplantspeciesandnumberofseedsgerminatingfromsamplestakenfromnests(N)ofOdontomachuschelifer(Latreille)andPachycondylastriataFr.(N)or controlsites(C).*ProbabilitiesfortheChi-squaretestoftheplantspeciesfrequenciesinant-nestvscontrolsites.

Taxa O.chelifer P.striata Total P*

N C N C

MiconiacalvescensDC. 80 57 58 25 220 0.0001

PanicumpilosumSW 46 73 29 45 193 0.0019

Clidemiahirta(L.)D.Don 51 34 47 26 158 0.0025

MikanialanuginosaDC 20 12 18 11 61 0.0547

PiptocarphabrasilianaCass. 22 14 16 6 58 0.0181

CecropiapachystachyaTrec. 13 8 16 10 47 0.1086

CaseariasylvestrisSwartz 11 10 16 7 44 0.1317

Thehigherdensityanddiversityofseedlingsemerginginsoil samplesfromantnestsinthegreenhouseexperimentisno guaran-teeofhigherseedlingrecruitmentcomparedtositeswithoutnests inforestfragments.Density-dependentmortalityfactorscouldact onnestareasandreduceseedlingrecruitment(Alvarez-Loayzaand Terborgh,2011).However,intheﬁeld,seedlingsinthenestsites

were,onaverage,1.8timesmoreabundantand1.6timesricher

inspeciesthancontrolsites.Therefore,evenifdensity-dependent

mortalityfactorswereoperating,theywerenotenoughtoreduce

thenumberofseedlingsinnestareastothesamelevelofcontrol

sites.Besides,seedlingsurvivalinnestareasmaybehigherdueto betterconditionsofantnestsoils,e.g.,highersoilfertility.

Antnestsmayalsoimprovethereceptionandtransmissionof

waterfromthesoilsurfacetothesubsoilforplantusage(Lobryde BruynandConacher,1994).Thisisparticularlyrelevantinforest

fragmentswhere poorsoilconditions,incombinationwithhigh

temperatures,reducedrought toleranceof seedlings(Zambrano

etal.,2014).Seedlingscanalsobeneﬁtfromantprotectionagainst naturalenemies(PassosandOliveira,2004).

InadditiontosomenichedifferencesbetweenO.cheliferand

P.striata,suchasnestarchitecture,characteristicsofcoloniesand feedingstrategies(Bottcheretal.,2016),theonlysigniﬁcant

dif-ferencebetweenantspecieswasthehigherabundanceofplants

germinatinginsoilsamplesfromO.chelifernests.Thisresultmay

beduetothehigheruseoffruitsbyO.chelifer,asdemonstrated

byBottcheretal.(2016).Despitethat,ﬁelddatadidnotconﬁrm

ahigher abundanceof seedlingsonO.chelifer nests,suggesting

thatdensity-dependentmortalitymayoccur.Therefore,ourresults

indicatethatthespecies studiedmaybefunctionallysimilarto

forestecosystems.DuetothehighabundanceofO.cheliferandP.

striatainforestfragments(AlmeidaandQueiroz,2015),the

popu-lationdynamicsofthetwospeciesmaybecoupledbycompetition

(MedeirosandOliveira,2009).Hence,iftheyhavesimilarfunctional

roles,thedecreaseinonespeciescanbefunctionallycompensated

forbyanincreaseintheother.Thiscompensationeffectis

pre-dictedbythe“insurancehypothesis”,whichlinksbiodiversityto

ecosystemstability(YachiandLoreau,1999).

Smallandlargemammalsdisperseseedsofspeciesofthegenus

Miconia(MagnussonandSanaiotti,1987;Groenendijketal.,1996).

Mammalsare important for seedlingrecruitment becausethey

buryseeds(PiresandGaletti,2013),protectingthemagainstseed

predators.Besides,somemammalspeciespromotesoilmovement

andcontributetoitsfertility(Clarketal.,2018),buttheycanbe absentfromsmallforestfragments(Redford,1992;Turner,1996).

Theabsenceofmammalsfromforestfragments,asisthecaseof

thefragmentsstudied(A.Pires,unpublisheddata),canaffectplant

recruitment.Wesuggestthatinthisscenario,plantrecruitment

canbehighlyinﬂuencedbyants,suchasO.cheliferandP.striata,

whichareabundantandcanprotecttheseedsfrompredationby

takingtheminsideorneartheirnests.Theymayalsofacilitateplant

growthduetoimprovedsoilconditions.In theforestfragments

studied,O.cheliferandP.striataarethemostfrequentantspecies removingseedsfromtheforestﬂoor(Almeidaetal.,2013).Dueto

theirwidespreadoccurrenceinNeotropicalforests,O.cheliferandP.

striatamayhaveastronginﬂuenceonplantcommunity,notonly

inrelativelywell-preservedtropicalforests(PassosandOliveira, 2002,2004),butalsoinforestfragmentswiththeabsenceofother

animalspecieswithsimilarecologicalroles,butmorevulnerable

tohabitatfragmentation.

Conﬂictsofinterest

Theauthorsdeclarenoconﬂictsofinterest.

Acknowledgements

ToEveraldoZonta(DepartamentodeSolos/UFRRJ),for

provid-ingthechemicalanalysisof soilsamples,andtoMSc.Marilena

Conde(Departamento de Biologia Vegetal/UFRRJ)for assistance

inplantidentiﬁcation.CNPqandFAPERJfundedourstudy

(pro-cesses471061/2008-5andE-26/112.121/2008,respectively).FSA,

LE, and JMQ thanks, respectively, CAPES, CONICET (Argentina)

andFAPERJ(proc.E-26/101.472/2010).WethankAlejandro

Farji-Brener,AlexanderChristianini,GabrielaPirk,andtwoanonymous

reviewersfortheirvaluablecommentsonanearlierversionofthe

manuscript. ¨ThisstudywasﬁnancedinpartbytheCoordenac¸ão

deAperfeic¸oamentodePessoaldeNívelSuperior-Brasil(CAPES)

-FinanceCode001”.

References

Almeida,F.S.,Queiroz,J.M.,2015.Formigasponeromorfascomoengenheirasde ecossistemas:impactossobreabiologia,estruturaefertilidadedossolos.In: Delabie,J.H.C.,Feitosa,R.M.,Serrão,J.E.,Mariano,C.S.F.,Majer,J.D.(Eds.),As FormigasPoneromorfasDoBrasil.Editus,Ilhéus,pp.437–446.

Almeida,F.S.,Mayhé-Nunes,A.J.,Queiroz,J.M.,2013.Theimportanceof ponero-morph ants for seed dispersal in altered environments. Sociobiology 60, 229–235,http://dx.doi.org/10.13102/sociobiology.v60i3.229-235.

Alvares,C.A.,Stape,J.L.,Sentelhas,P.C.,DeMoraes,G.,Leonardo,J.,Sparovek,G., 2013.Koppen’sclimateclassiﬁcationmapforBrazil.Meteorol.Z.22,711–728, http://dx.doi.org/10.1127/0941-2948/2013/0507.

Alvarez-Loayza,P.,Terborgh,J.,2011.FatesofseedlingcarpetsinanAmazonian ﬂoodplainforest:intra-cohortcompetitionorattackbyenemies?J.Ecol.99, 1045–1054,http://dx.doi.org/10.1111/j.1365-2745.2011.01835.x.

Augspurger,C.K.,Kelly,C.K.,1984.Pathogenmortalityoftropicaltreeseedlings -experimentalstudiesoftheeffectsofdispersaldistance,seedlingdensity,and lightconditions.Oecologia61,211–217,http://dx.doi.org/10.1007/BF00396763. Benitez-Malvido, J., Garc´ia-Guzmán, G., Kossmann-Ferraz, I.D., 1999. Leaf-fungal incidence and herbivory on tree seedlings in tropical rain-forest fragments: an experimental study. Biol. Conserv. 91, 143–150, http://dx.doi.org/10.1016/S0006-3207(99)00090-7.

Bottcher,C.,Peixoto,P.E.,Silva,W.R.,Pizo,M.A.,2016.Driversofspatialvariation intheroleofantsassecondaryseeddispersers.Environ.Entomol.45,930–937, http://dx.doi.org/10.1093/ee/nvw058.

Briese,D.T.,1982.Theeffectofantsonthesoilofasemi-aridsaltbushhabitat.Ins. Soc.29,375–386,http://dx.doi.org/10.1007/BF02228765.

Bruna, E., 2002. Effects of forest fragmentation on Heliconia acuminate seedling recruitment in central Amazonia. Oecologia 132, 235–243, http://dx.doi.org/10.1007/s00442-002-0956-y.

Brussaard,L.,1997. Biodiversityand ecosystemfunctioninginsoil.Ambio26, 563–570.

Cammeraat, L.H., Willott, S.J., Compton, S.G., Incoll, L.D., 2002. The effects of ants‘nest on the physical, chemical and hydrological

(6)

properties of a rangeland soil in semi-arid Spain. Geoderma 105, 1–20, http://dx.doi.org/10.1016/S0016-7061(01)00085-4.

Chambers,J.C.,Macmahon,J.A.,1994.Adayinthelifeofaseed:movementsand fatesofseedsandtheirimplicationsfornaturalandmanagedsystems.Annu. Rev.Ecol.S.25,263–292.

Christianini, A.V., Mayhe-Nunes, A.J., Oliveira, P.S., 2007. The role of ants in the removal of non-myrmecochorous diaspores and seed germination in a neotropical savanna. J. Trop. Ecol. 23, 343–351, http://dx.doi.org/10.1017/S0266467407004087.

Clark,K.L.,Branch,L.C.,Farrington,J.,2018.Bioturbationbymammalsandﬁre inter-acttoalterecosystem-levelnutrientdynamicsinlongleafpineforests.PLoSOne 13,e0201137,http://dx.doi.org/10.1371/journal.pone.0201137.

Dalling, J.W., Wirth, R., 1998. Dispersal of Miconia argentea seeds by the leaf-cutting ant Atta colombica. J. Trop. Ecol. 14, 705–710, http://dx.doi.org/10.1017/S0266467498000492.

EMBRAPA-EmpresaBrasileiradePesquisaAgropecuária,1997.Servic¸oNacional delevantamentoeconservac¸ãodesolos.In:ManualDeMétodosDeAnáliseDe Solo.EMBRAPA,RiodeJaneiro.

Escobar,E.D.F.,Cardoso,V.J.M.,2015.Longevityofseedsandsoilseedbankofthe CerradotreeMiconiachartacea(Melastomataceae).SeedSci.Res.25,386–394, http://dx.doi.org/10.1017/S0960258515000173.

Farji-Brener,A.G.,Ghermandi,L.,2000.Inﬂuenceofnestsofleaf-cuttingantsonplant speciesdiversityinroadvergesofNorthernPatagonia.J.Veg.Sci.11,453–460, http://dx.doi.org/10.2307/3236638.

Farji-Brener, A.G., Medina, C.A., 2000. The importance of where to dump the refuse: seed bank and ﬁne roots in nests of the leaf-cutting ants Atta cephalotes and Atta colombica. Biotropica 32, 120–126, http://dx.doi.org/10.1111/j.1744-7429.2000.tb00454.x.

Farji-Brener, A.G., Silva, J.F., 1996. Leaf cutter ants (Atta laevigata) aid to the establishment success of Tapirira velutinifolia (Anacar-diaceae) seedlings in a parkland savanna. J. Trop. Ecol. 12, 163–168, http://dx.doi.org/10.1017/S0266467400009378.

Farji-Brener, A.G.,Werenkraut,V., 2017. Theeffectsof antnests onsoil fer-tilityandplantperformance: ameta-analysis. J.Anim. Ecol.86, 866–877, http://dx.doi.org/10.1111/1365-2656.12672.

Fialho,J.S.,Gomes,V.F.F.,Oliveira,T.S.,Silva-Júnior,J.M.T.,2006.Indicadoresda qual-idadedosoloemáreassobvegetac¸ãonaturalecultivodebananeirasnaChapada doApodi-CE.Rev.Ciênc.Agron.37,250–257.

Fittkau, E.J., Klinge, H., 1973. On biomass and trophic structure of the Central Amazonian rain forest ecosystem. Biotropica 5, 2–14, http://dx.doi.org/10.2307/2989676.

Folgarait, P.F., 1998. Ant biodiversity and its relationship to ecosys-tem functioning: a review. Biodivers. Conserv. 7, 1221–1244, http://dx.doi.org/10.1023/A:1008891901953.

Folgarait, P.J., Perelman, S., Gorosito, N., Pizzio, R., Fernandez, J., 2002. Effects of Camponotus punctulatus ants on plant community composi-tion and soil properties acrossland-use histories. Plant Ecol. 163, 1–13, http://dx.doi.org/10.1023/A:1020323813841.

Francelino, M.R., Rezende, E.M.C., Silva, L.D.B., 2012. Proposta para zoneamento ambiental de plantio de eucalipto. Cerne 18, 275–283, http://dx.doi.org/10.1590/S0104-77602012000200012.

Groc,S.,Delabie,J.H.C.,Fernández,F.,Leponce,M.,Orivel,J.,Silvestre,R.,Vasconcelos, H.L.,Dejean,A.,2013.Leaf-litterantcommunities(Hymenoptera:formicidae) inapristineGuianeserainforest:stablefunctionalstructureversushighspecies turnover.Myrmecol.News19,43–51.

Groenendijk,J.P.,Bouman,F.,Cleef,A.M.,1996.Anexploratorystudy onseed morphologyofMiconiaruiz&Pavon(Melastomataceae),withtaxonomicand ecologicalimplications.ActaBot.Neerl.45,323–344.

Guimarães,I.C.,Pereira,M.C.,Batista,N.R.,Rodrigues,C.A.P.,AntonialliJúnior,W.F., 2018.ThecomplexnestarchitectureofthePonerinaeantOdontomachus che-lifer.PLoSOne13,e0189896,http://dx.doi.org/10.1371/journal.pone.0189896. Horvitz,C.C.,Beattie,A.J.,1980.AntdispersalofCalatheamarantaceaeseedsby car-nivorousponerines(Formicidae)inatropicalrain-forest.Am.J.Bot.67,321–326, http://dx.doi.org/10.2307/2442342.

King,T.J.,1977. Plant ecologyof ant-hillsin calcareousgrasslands.1. Patterns ofspeciesinrelationtoant-hillsinsouthernEngland.J.Ecol.65,235–256, http://dx.doi.org/10.2307/2259077.

Leck,M.,Parker,T.(Eds.),2008.SeedlingEcologyandEvolution.Cambridge Univer-sityPress,NewYork.

Leitao,F.H.M.,Marques,M.C.,Ceccon,E.,2010.Youngrestoredforestsincrease seedlingrecruitmentinabandonedpasturesintheSouthernAtlanticrainforest. Rev.Biol.Trop.58,1271–1282.

LobryDeBruyn,L.A.,Conacher,A.J.,1994.Theeffectofantbioporesonwater inﬁltrationinsoilsinundisturbedbushlandandinfarmlandinasemi-arid environment.Pedobiologia38,193–207.

Madureira, M.S., Schoereder, J.H., Teixeira, M.C., Sobrinho, T.G., 2013. Why does Atta robusta (Formicidae) not change soil features around their nests as other leaf-cutting ants do? Soil Biol. Biochem. 57, 916–918, http://dx.doi.org/10.1016/j.soilbio.2012.11.005.

Magnusson, W.E., Sanaiotti, T.M., 1987. Dispersal of Miconia seeds by de rat Bolomys lasiurus. J. Trop. Ecol. 3, 277–278, http://dx.doi.org/10.1017/S0266467400002169.

Martins, D.,Martins,C.C., Costa,N.V., 2006.Potencial alelopático desoluc¸ões desolocultivadocomBrachiaria brizantha:efeitossobreagerminac¸ãode gramíneasforrageiraseplantasdaninhasdepastagens.PlantaDaninha24, 61–70,http://dx.doi.org/10.1590/S0100-83582006000100008.

Medeiros,F.N.S.,Oliveira,P.S.,2009.Season-dependentforagingpatterns:casestudy ofaneotropicalforest-dwellingant(pachycondylastriata;Ponerinae).In:Jarau, S.,Hrncir,M.(Eds.),FoodExploitationbySocialInsects:Ecological,Behavioral, andTheoreticalApproaches.CRCPress,NewYork,pp.81–95.

Medeiros,F.N.S.,1997.EcologiacomportamentaldaformigaPachycondylastriata Fr.Smith(Formicidade:Ponerinae)emumaﬂorestadosudestedoBrasil.Tese (MestradoemCiênciasBiológicas),UniversidadeEstadualdeCampinas. Moutinho, P., Nepstad, D.C., Davidson, E.A., 2003.

Inﬂu-ence of leaf-cutting ant nests on secondary forest growth and soil properties in Amazonia. Ecology 84, 1265–1276, http://dx.doi.org/10.1890/0012-9658(2003)084[1265:IOLANO]2.0.CO;2. O’Grady,A.,Breen,J.,Harrington,T.J.,Courtney,R.,2013.Theseedbankinsoil

fromthenestsofgrasslandantsinauniquelimestonegrasslandcommunityin Ireland.Ecol.Eng.61,58–64,http://dx.doi.org/10.1016/j.ecoleng.2013.09.011. Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara,

R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H., http://CRAN.R-project.org/package=vegan,2013.

Passos,L.,Oliveira,P.S.,2002.AntsaffectthedistributionandperformanceofClusia criuvaseedlings,aprimarilybird-dispersedrainforesttree.J.Ecol.90,517–528, http://dx.doi.org/10.1046/j.1365-2745.2002.00687.x.

Passos, L., Oliveira, P.S., 2004. Interactions between ants and fruits of Guapira opposita (Nyctaginaceae) in a Brazilian sand plain rain for-est: ant effects on seeds and seedling. Oecologia 139, 376–382, http://dx.doi.org/10.1007/s00442-004-1531-5.

Pereira,M.P.S.,Francelino,M.R.,Queiroz,J.M.,2017.Acoberturaﬂorestalem pais-agensdomédiovaledoRioParaíbadosul.Floresta&Ambiente24,1–11, http://dx.doi.org/10.1590/2179-8087.134115.

Pires,A.,Galetti,M.,2013.TheagoutiDasyproctaleporina(Rodentia:dasyproctidae) asseeddisperserofthepalmAstrocaryumaculeatissimum.Mastozool.Neotrop. 19,147–153.

Pizo,M.A.,Oliveira,P.S.,2000.TheuseoffruitsandseedsbyantsintheAtlantic forestofsoutheastBrazil.Biotropica32,851–861.

Redford,K.H.,1992.Theemptyforest.BioScience42,412–422.

Salinas-Peba,L.,Parra-Tabla,V.,Campo,J.,Munguía-Rosas,M.A.,2014.Survivaland growthofdominanttreeseedlingsinseasonallytropicaldryforestsofYucatan: siteandfertilizationeffects.J.PlantEcol.7,470–479.

Turner,I.M.,1996.Specieslossinfragmentsoftropicalrainforest:areviewofthe evidence.J.Anim.Ecol.33,200–209,http://dx.doi.org/10.2307/2404743. Wang,S.,Wang,H.,Li,J.,Zhang,Z.,2017.Antscanexertadiverseeffectonsoilcarbon

andnitrogenpoolsinaXishuangbannatropicalforest.SoilBiol.Biochem.113, 45–52,http://dx.doi.org/10.1016/j.soilbio.2017.05.027.

Yachi,S.,Loreau,M.,1999.Biodiversityandecosystemproductivityinaﬂuctuating environment:theinsurancehypothesis.Proc.Nati.Acad.Sci.96,1463–1468. Zambrano, J., Coates, R., Howe, H.F., 2014. Effects of forest

fragmen-tation on the recruitment success of the tropical tree Poulsenia armata at Los Tuxtlas, Veracruz, Mexico. J. Trop. Ecol. 30, 209–218, http://dx.doi.org/10.1017/S0266467414000108.

Zamora,R.,Matias,L.,2014.Seeddispersers,seedpredators,andbrowsersact synergisticallyasbioticﬁltersinamosaiclandscape.PLoSOne9,e107385, http://dx.doi.org/10.1371/journal.pone.0107385.

Zuur,A.,Ieno,E.N.,Walker,N.,Saveliev,A.A.,Smith,G.M.,2009.MixedEffectsModels andExtensionsinEcologyWithR,StatisticsforBiologyandHealth. Springer-Verlag,NewYork.