Long-term sustainability of cork oak agro-forests in the Iberian Peninsula: A model-based approach aimed at supporting the best management options for the montado conservation

(1)

ContentslistsavailableatScienceDirect

Ecological

Modelling

jo u r n al ho m e p ag e :w w w . e l s e v i e r . c o m / l o c a t e / e c o l m o d e l

Long-term

sustainability

of

cork

oak

agro-forests

in

the

Iberian

Peninsula:

A

model-based

approach

aimed

at

supporting

the

best

management

options

for

the

montado

conservation

M.L.

Arosa

a,∗

_,

_R.

_Bastos

b

_,

_J.A.

_Cabral

b

_,

_H.

_Freitas

a

_,

_S.R.

_Costa

c,d

_,

_M.

_Santos

b

a_Centre_for_Functional_Ecology,_Department_of_Life_Sciences,_University_of_Coimbra,_P.O._Box_3046,_3001-401_Coimbra,_Portugal

b_Laboratory_of_Applied_Ecology,_CITAB₋_Centre_for_the_Research_and_Technology_of_{Agro-Environment}_and_Biological_Sciences,_University_of Trás-os-MonteseAltoDouro,5000-911VilaReal,Portugal

c_Mountain_Research_Center_(CIMO),_ESA,_Polytechnic_Institute_of_Braganc¸_a,_Campus_de_Santa_Apolónia,_5300-253_Braganc¸_a,_Portugal

d_CBMA₋_Centre_of_Molecular_and_{Environmental}_Biology,_Department_of_Biology,_University_of_Minho,_Campus_de_Gualtar,_4710-057_Braga,_Portugal

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received25March2016 Receivedinrevisedform 26September2016 Accepted10October2016 Availableonline31October2016 Keywords: Quercussuber Modelling Montadosustainability Populationdynamics

a

b

s

t

r

a

c

t

Thefutureofthemontado,ahumanshapedagro-forestryecosystemofSouthWesternEurope,is ques-tionedduetotheobservedlackofcorkoakhealthandlownaturalregeneration.WedevelopedaSystem DynamicsModellingapproachtopredictthelong-termsustainabilityofthisagro-forest,by recreat-ingcork-oakpopulationdynamics,managementpracticesandthemainenvironmentalandbiological constrainsassociatedwiththisecosystem.Ourresultsindicatethattheleadinglimitationstocorkoak regenerationinmontadoecosystemsresultfromtheintensityandinteractionoflandmanagement prac-tices,namelylivestockandtheuseofheavymachinery.Themainconclusionsindicatethatlimitingthe quantityoflivestockupto0.40LU.ha−1,andconsideringsoilploughingwithaminimumperiodicityof 5years,arecrucialtomaintainingsustainablecorkoakpopulations.Thisstudyrepresentsaﬁrststep tosupportstrategicoptionsforcorkoakmontadomanagementbyprovidingprojectionsoflong-term populationtrendsunderrealisticsocial-ecologicalchangescenarios.

1. Introduction

The largest cork oak woodlands are found in the Iberian

Peninsula,wherecorkharvesting,agricultural,pastoralandother

traditionaluseshavebeenpracticedatleastsincetheMiddleAges

(Joffreet al.,1999;Oleaetal.,2005; Bugalhoet al.,2011).This

agro-silvo-pastoralecosystem,knownasmontadoordehesa(the

PortugueseandSpanishnames,respectively),iscomposed

typi-callyofopenwoodlands(20–80treesha−1)withonlyoneorafew

treespeciessuchascorkoak(Quercussuber),holmoak(Quercus

rotundifolia)andpines(Pinusspp.)(Joffreetal.,1999;Pinto-Correia

andMascarenhas,1999;Pinto-CorreiaandFonseca,2009).

More-over,themontadoecosystemisconsideredbothaprotectedhabitat

withintheEUhabitatsdirective(92/43/EEC)andahighnaturevalue

farmingsystem,accordingtotheclassiﬁcationproposedbythe

EuropeanEnvironmentalAgency(Paracchinietal.,2008).

∗ Correspondingauthor.

E-mailaddress:merisinha@hotmail.com(M.L.Arosa).

Thedominanttreeofthemontado,thecorkoak(Quercussuber),

isasclerophyllousevergreenoakthatoccursinnon-carbonated

soils(Pausas et al.,2009)restricted to thewesternpart ofthe

MediterraneanBasin(Tutinetal.,1964).Thisisa slow-growing

tree with a high longevity like other oaks, with a lifespan of

250–300years.Earliestﬂoweringandfructiﬁcationoccurataround

15–20years ofage,producing both annualand biennial acorns

(Natividade, 1950; Elena-Rosselló et al., 1993; Díaz-Fernández et al.,2004; Pereira and Tomé,2004).Acorns can bedispersed

by animals or predated. The main seed dispersers are wood

mouse(Apodemussylvaticus)andEuropeanjay(Garrulus

glandar-ius), which are involved in short- and long-distance dispersal,

respectively(Herrera,1995;Pausasetal.,2009).Acornpredators

aremainlyseed-boringinsects,suchastheacornweevil(Curculio

spp.,Coleoptera)andtheacornmoth(Cydiaspp.,Lepidoptera)(Ceia

andRamos,2014).Inopenareas,predatorscanalsobelivestock,

deer,wildboar,birdsandrabbitswhereasunderashrubcanopy

acornpredationismostlycarriedoutbysmallrodents,since

nei-therlargemammalsnorbirdswillpenetratethedensevegetation

(Herrera, 1995).Corkoak trees growin areas withstrong

sea-sonalwaterdeﬁcitsbysurvivingtodroughtsinpartduetotheir

http://dx.doi.org/10.1016/j.ecolmodel.2016.10.008

(2)

Fig.1.Conceptualdiagramofthedynamicmodelusedto:B)simulatethecorkoakpopulationdynamics,assumingcorkoakagestratiﬁcation(seedlings,saplings,immature treesandadulttrees),A)simulatetheassociatedenvironmentalconstrainsandC)therespectiveeconomicbalancesub-model.

extensiveanddeeprootsystemandinsomecasesseedlingsshed

leavesandresproutwhenthedroughtisover(Acácioetal.,2007;

Gómez-Aparicioetal.,2008;Arosaetal.,2015).Thecorkoakhasa

thickinsulatingbarkwhichevolvedasaﬁreprotectionmechanism

(Pausas,1997;Catryetal.,2012)andwhichregrowsafterextraction

forcorkproduction(Pausasetal.,2009).Corkextractionrepresents

animportantresourceofhigheconomicvalueintheIberian

Penin-sulaascorkthicknessandpropertiesmakeitvaluablerawmaterial

forindustry.Aftertheremovaloftheouterbark,atraditional

prac-ticethatdoesnotharmthetreeandoccursevery9years,thetree

hasthecapacitytoproduceanewcorkbarkbyaddingnew

lay-ersofcorkeveryyearandthismayberepeatedthroughoutthe

treelifetime(Pausasetal.,2009).Theannualproductionofcorkis

about370000t,mostlyfromthecorkoaksofPortugalandSpain

thatproducerespectively51%and23%oftheworldtotal(Pereira

andTomé,2004).

Ourmaingoalwastheintegrationofdisperseinformationfrom

speciﬁcstudiesconcerningthedifferentlifestagesandstressors

inasingleuser-friendlyplatformforunderstandingandpredicting

thefuturesustainabilityofthemontadoecosystem.Weconsidered

asstressorsthelackofcorkoakhealthandthelownatural

regen-erationrates.Treehealthhasbeenaffectedbyintensivepruning,

exaggeratedcorkharvestingand theinﬂuenceofpestsand

dis-eases(Camilo-Alvesetal.,2013;AcácioandHolmgren,2014).The

limitationstonaturalregenerationhavebeenattributedtovarious

causes,including:a)poordispersalandshortageofviableacorns

(Brancoetal.,2002;PulidoandDíaz,2005;Acácioetal.,2007);

b)highpost-dispersiveacornlossesassociatedwithseedling

mor-talityduetoover-grazingbylivestockandwildanimals(Herrera,

1995;Plieningeretal.,2004;Acácioetal.,2007;Pulidoetal.,2013);

c)lowseedlingsurvivaltosummerdrought(Gimenoetal.,2009;

Smitetal.,2009).Besides,thedevelopmentoffarm

mechaniza-tion,includingthegeneraliseduseofwideplows,discharrows,and

scariﬁersdestroysyoungtreesandmaydamagerootsandweaken

establishedtrees,creatingmoresusceptibilitytotheattackofpests

anddiseases (Brancoand Ramos,2009; Arosaetal., 2015).The

ﬁrefrequencydetectedinthelastfewdecadesinthisecosystem

(PausasandVallejo,1999)hasalsobecomeanadditionalobstacle

tocorkoaksurvivaland recruitment(Acácioetal.,2009,2010).

Althoughcorkoakisabletoexhibitﬁre-resistancecharacteristics

due to thebark insulation and tothe mechanism of

resprout-ingafterwards,frequentorintensewildﬁresmaykilladulttrees,

especiallyiftheseeventsoccurimmediatelyaftercorkextraction

(Moreiraetal.,2007).Additionally,asconsequenceofconcomitant

changeofmeanannualtemperatureandrainfallextremesduring

thelastdecades,therecurrentdroughts willreducethecanopy

coverandlimitcorkoakregeneration(VivasandMaia,2008;Acácio

etal.,2009).

Consideringitsecologicalandeconomicsustainability,the

con-servationoftheseecosystemsrequiresthecomprehensionofits

capacitytowithstandandrecoverfromdisturbancesimposedby

naturalandanthropogenicfactors(Müller,2005;Kandzioraetal.,

2013;Stolletal.,2015).Ecologicalmodellingprovidesusefultools

tostudycomplexsystemsbypredictingtheoutcomeofalternative

scenarios,andmighthelpguidingthemostcorrectmanagement

optionsfromprojectedfutureoutcomes(e.g.Bastosetal.,2012,

2015;Fernandesetal.,2013;Santosetal.,2013).Actuallydynamic

modelscanbeusedtosupportthemechanisticunderstandingof

complexmultifactorialecologicalprocessesastheysimultaneously

integratethestructureandthecompositionofsystemsforaspeciﬁc

period(Jørgensen,1994,2001).Whenproperlydeveloped,tested

and applied withinsight and withrespect for theirunderlying

assumptions,dynamicmodelsarecapableofsimulatingconditions

thataredifﬁcultorimpossibletoproduceotherwise(Jørgensen,

2001).

WehavedevelopedaSystemDynamicsModellingapproachto

recreatethemanagementpracticesassociatedwithcorkoak

mon-tadoandthemainbiologicalandenvironmentaldriversinﬂuencing

thissemi-naturalecosystem,inordertoassessitslong-term

sus-tainability(Sterman,2001).Thisstudyistheﬁrsttointegrateina

singleapproachthemultifactorialfactorsassociatedwithcorkoak

populationdynamics.Speciﬁcally,ourobjectiveswere:(1)

inte-gratinginauser-friendlyplatformthebioticandabioticfactors,

actingasawhole,withholisticandinterrelatedinﬂuencesonthe

corkoakpopulationdynamics(2)simulaterealisticscenariosin

ordertounderstandthemostdeterminantfactorsimplicatedinthe

corkoakdeclineacrosstheIberianPeninsulaand(3)highlightthe

mosteffectivemanagementpracticesthroughquantitativemetrics

withimplicationsforthelong-termsustainabilityandconservation

(3)

Fig.2.Classesinthelifecycleofcorkoakinagroforestsconsideredforthedesignof theconceptualdiagramofthefoursub-models:1seedlings,2saplings,3immature trees,4adulttreesandtheprocessesinfluencingeachcohortofindividuals.Drought, fire,livestockandshrubclearingareprocessesthatmostaffect1and2classes.Death andfirehaveanimportanteffectover3and4classes.Individualsofaspecificclass transittothenextwhencompletingtheirdevelopment.

2. Methods

2.1. Dynamicmodelconceptualization

Corkoakpopulationdynamicsinagro-forestsweremodelled

toanticipatethepotentialeffectsof currentmanagement

prac-ticesunderdifferentenvironmentalscenariosexpectedforthe21st

century.Naturalregenerationisadynamicprocesswherenew

indi-vidualsarerecruitedintothematurepopulation,compensatingthe

lossesduetomortality,inducedbyseveralbioticandabiotic

fac-tors(Harper,1977;Holmgrenetal.,1997;PulidoandDíaz,2005; FeiandSteiner,2008).Therefore,wedevelopedaSystemDynamics

Modellingapproachcomposedbyfoursub-modelsconcerningthe

corkoakpopulationdynamicsbasedoncorkoakagestratiﬁcation

(classes)(Gonzálezetal.,2015),representedbythesumof

differ-entlifestagecohorts(seedlings,saplings,immaturetreesandadult

trees),andoneadditionalsub-modeltoestimatetheeconomic

bal-ancesassociated(Fig.1).

Theyearwaschosenastime unitand thesimulation period

wasestablishedfor100years,consideredsuitabletocapturethe

inﬂuenceofmainfactorsonthelongtermcorkoakpopulation

dynamics,namelythoseinducedbymanagementoptions.All

mod-ellingwas performedwiththe software STELLA,version 10.0.5

(IseeSystems,Inc.).Theoriginalconceptualdiagramofthe

over-allsub-models(AppendixIinSupplementarymaterial),variables

andequations includedinthemodel construction(Table1 and

AppendixIIinSupplementarymaterial)andfullexplanationof

pro-cesses(AppendixIIIinSupplementarymaterial),areavailableas

supplementaryelectronicmaterial.

3. Naturalregeneration

Inordertorecreatecorkoak naturalregeneration, four

sub-models (seedlings, saplings, immature trees, adult trees) were

designedtoreproducethemostrelevant classes,dynamics and

environmentalconstrainsassociatedtothecorkoakpopulations

(Fig.1,Table1,AppendixI,AppendixIIinSupplementary

mate-rial, Fig.3).For each sub-model, cork oak trees were grouped

intoclasses,consideringtheirstage,sizeanddiameteratbreast

height(DBH):seedlings(acornorcotyledonscattersstillattached,

height<50cm),saplings(height>50cm,DBH<10cm),immature

tree (height >50cm, DBH>10cm) and adult tree (DBH>25cm)

(MuickandBartolome,1987;MonteroandCa ˜nellas,2003).

Transi-tionsbetweenclassesandwithineachclasswereaffectedbybiotic

andabioticfactorsthatimplicatecause-effectrelationships.Data

usedinthemodelwascompiledfrompublicationsregardingthe

corkoak’lifecycleandthecorkoakmontadocharacterizationinthe

IberianPeninsula,rangingfrominlandregions,withlowerannual

precipitation(550mm)andhighertemperature(17◦C),tocoastal

regions,withhigherrainfall(1800mm)andmildertemperatures

(14◦C)(Table1).

3.1. Bioticandabioticprocessesinﬂuencingcorkoaknatural

regeneration

Severalbioticandabioticprocesseswereassumedasinﬂuent

driversonthecorkoakpopulationdynamics.Theseedlingsand

saplingsweremainlyaffectedbydroughts,ﬁreeffects,livestock

andgrazingpressure,andshrubclearingintensity(Fig.2).Onthe

otherhand,immatureandadulttreeswereaffectedbyﬁreand

naturaldeathbyaging.Besidestheeffectofexternalfactors,the

transitionof individuals throughoutclasses consideredintothe

modelisalsolimitedbyintraspeciﬁccompetitionandbythe

suc-cessofestablishedcorkoakplants.

3.1.1. Acorngermination

To model the germination of acorns, each adult tree was

assumedtoproducebetween1.32and33.6kgofacornsperyear

(Herrera, 1995;Ca ˜nellas etal.,2007;Pérez-Ramoset al.,2008).

Sincethegerminationofacornsislimitedbybioticfactorsrelated

withpredationandinsectdamages,theeffectivenumberofacorns

supplyingthesystemwasdependentonthepredation/survivalrate

andoftheproportionofviableacorns(Brancoetal.,2002;Acacio

etal.,2007).Furthermore,acornsthatdonotgerminateandremain

intheunderstorydryoutandbecomeinviablebeforethebeginning

ofanewyear.

3.1.2. Competition

Cork oak competition in montado involves an intraspeciﬁc

struggle for basic resources, suchas light, water and/or

nutri-ents(Flores-Martinezetal.,1994;Holmgrenetal.,1997;Callaway andWalker,1997).Ourdynamicmodelwasdesignedtorecreate

intraspeciﬁccompetitionwithinandbetweenthedifferentclasses,

fromthesaplingtotheadulttrees.Duringthetreedevelopment,

competitionvaries from soilresources tolight, this last

affect-ingtheshape andsize ofcrowns(Kenkel,1988; Deleuzeetal.,

1996).Withincreasingtreecoverandlightlimitation,theolder

individualsbegintodisproportionatelyobstructtheincomingsolar

radiationbecauseofitsunidirectionalnature,therebysuppressing

thegrowthof younger trees (Dolezal etal., 2004).In this

con-text,AssmannandGardiner(1970)deﬁnedthattheareaavailable

foratreeisrepresentedbyitscrownprojectionversusthe

cor-respondingfractionnotoccupied.Serrada(2002)estimatedthat

thismethodsuffersthedisadvantageofsupposingafullcanopy

cover(100%)butincorkoakmontadosthenormalcanopycoveris

incomplete,thusweusedtheDiBérenger´ısmethodfor

calculat-ingaveragecrownprojectionusingthediameteratbreastheight

(DBH).This methodenablesthecalculationof treedistribution

withinanarea,assumingamixtureoftreesbelongingtoallclasses.

Throughthefunctiony=0.0431•x1.6025_(x_is_DBH_in_cm)_the

maxi-mumcorkoakdensity(treesha−1)belongingtoeachsizeclasswas

limitedinthemodel(Natividade,1950;Marín-PageoandCamacho,

(4)

Table1

Speciﬁcationofthemainvariablesincludedinthemodel,respectivedescription,measureunitsandreferences.

Corkoakpopulationdynamics

Variable Description Unit Source

Seedlings

Randomacornspertree Averagekgofacornspertree Minimumacornspertree=1.32 Maximumacornspertree=33.6

kgofacorns Herrera(1995),Ca ˜nellasetal. (2007),Pérez-Ramosetal. (2008)

RandomacornsPredationrate Proportionofpredatedacornsbefore

germination

Minimumacornspredationrate=0.36 Maximumacornspredationrate=0.83

rate Acácioetal.(2007)

Randomacornsinsectdamagerate Proportionofacornsdamagedbyinsects

beforegermination

Minimumacornsinsectdamagerate=0.09 Maximumacornsinsectdamagerate=0.68

rate Brancoetal.(2002),Acácio etal.(2007)

RandomacornsAcorngerminationrate Proportionofacornssuccessfullygerminated

Minimumacornsgerminationrate=0.38 Maximumacornsgerminationrate=0.84

rate Herrera(1995),Acácioetal. (2007),Arosaetal.(2015)

Averageseedlingsﬁremortalityrate Proportionofseedlingmortalitybyﬁre

Fireseedlingsmortality=0.0008

rate Catryetal.(2012)

Randomseedlingdroughtmortalityrate Proportionofseedlingmortalitybydrought

Minimumseedlingsdroughtmortality rate=0.68

Maximumseedlingsdroughtmortality rate=0.70

rate Arosaetal.(2015)

Randomseedlingslivestockmortalityrate Proportionofseedlingmortalitybylivestock

Minimumseedlingslivestockmortality rate=0.28

Maximumseedlingslivestockmortality rate=0.65

rate Plieningeretal.(2004)

Randomseedlingsshrubclearingmortalityrate Proportionofseedlingmortalitybyshrub clearing

Minimumseedlingsshrubclearingmortality rate=0

Maximumseedlingsshrubclearingmortality rate=1

Averageseedlingsresproutingrate Proportionofseedlingsuccessfully

resprouted=0.239

Saplings

Averagesaplingsﬁremortalityrate Proportionofsaplingsmortalitybyﬁre

Firesaplingsmortality=0.0008

Randomsaplingscohort×drought mortalityrate Proportionofsaplingsmortalitybydrought Minimumsaplingsdroughtmortality rate=0.68

Maximumsaplingsdroughtmortality rate=0.70

Randomsaplingscohort×livestockmortalityrate Proportionofsaplingsmortalitybylivestock Minimumsaplingslivestockmortality rate=0.57

Maximumsaplingslivestockmortality rate=0.87

rate Plieningeretal.(2004)

Randomsaplingscohort×shrubclearingmortalityrate Proportionofsaplingsmortalitybyshrub clearing

Minimumsaplingsshrubclearingmortality rate=0

Maximumsaplingsshrubclearingmortality rate=1

DBHsaplings Saplingsaveragediameteratbreastheight=6 cm Pulidoetal.(2013)

Immaturetrees

Randomimmaturetreescohort×firemortalityrate Proportionofimmaturetreesmortalitybyfire Minimumimmaturetreescohort×fire mortalityrate=0.06

Maximumimmaturetreescohort×ﬁre mortalityrate=0.08

Randomimmaturetreescohort×deathrate Proportionofimmaturetreesdeath

Minimumimmaturetreescohort×death rate=0

Maximunimmaturetreescohort×death rate=0.000092

rate RibeiroandSurov ´y(2008)

DBHimmaturetrees Immaturetreesaveragediameteratbreast

height=10

cm Pulidoetal.(2013)

Adulttrees

Adulttreesperha Initialdensityofadulttrees=50 treesha−1 _Joffre_et_al.₍₁₉₉₉₎

Randomadulttreesﬁremortalityrate Proportionofadulttreesmortalitybyﬁre

Minimumadulttreesﬁremortality rate=0.038214

Maximumadulttreesﬁremortality rate=0.04465

(5)

Table1(Continued)

Corkoakpopulationdynamics

Variable Description Unit Source

Randomadulttrees deathrate Proportionofadulttreesdeath Minimumadulttreesdeathrate=0 Maximumadulttreesdeathrate=0.000092

rate RibeiroandSurov ´y(2008)

DBHadulttrees Adulttreesaveragediameteratbreast

height=89

cm Pulidoetal.(2013)

Corkoakmanagementincomeperyear

Randomcorkproductionpertreekg Averagekgofcorkproducedpertree

Minimumcorkproductionpertree=35.95534 Maximumcorkproductionpertree=69.56406

kgtree−1 _Pereira_and_Tomé₍₂₀₀₄₎

Randomincomefromcorkpertreeeuroskg AverageincomeinDperkgofproducedcork

Minimumincomefromcorkpertree=1.33 Maximumincomefromcorkpertree=3.33

Dkg−1 _Pinheiro_and_Ribeiro₍₂₀₁₃₎ Averagestrippingcostcorkextractioneurosperkg AveragecostofcorkextractioninDperkg

Strippingcostcorkextraction=0.23

Dkg−1 _Pinheiro_and_Ribeiro₍₂₀₁₃₎

Periodicitycorkextraction Averageyearsbetweencorkextraction

Periodicitycorkextraction=9

years PereiraandTomé(2004)

Averageshrubclearingcosteurosperha AveragecostinDperclearedha

Shrubclearingcost=120

Dha−1 _Pinheiro_et_al.₍₂₀₀₈₎ Averagerevenuefromlivestockeurosperha AverageincomeinDoflivestockperha

Revenueformlivestock=22.4

Dha−1 Pinheiroetal.(2008) 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Year 0 200 400 600 800 1000 No . of a co rn s/ s eed li ng s/ s a pl in gs pe r h a 0 10 20 30 40 50 60 No. of i mm atu re/ ad u lt tr ee s per ha A Acorns Seedlings Saplings Immature trees Adult trees 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Year 0 200 400 600 800 1000 1200 No . of a co rn s/ se ed lings/s ap li ngs pe r ha 0 10 20 30 40 50 60 No . of im ma tu re /a d u lt tr ee s p e r ha B Acorns Seedli

ngs

Sapli

ngs

Immature trees Adult trees 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Year 0 200 400 600 800 1000 1200 No . of a c o rn s/ se e d li ng s/ s a pl in gs pe r h a 0 10 20 30 40 50 60 N o. of i mm atu re/ ad u lt tr ee s pe r h a C Acorns Seedlings Saplings Immature trees Adult trees 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Year 0 200 400 600 800 1000 1200 No . of a co rns/ se e d li n gs/ sa p li n gs per ha 0 10 20 30 40 50 60 N o. of i mm at u re/ ad u lt tr e es pe r h a D Acorns Seedlings Saplings Immature trees Adult trees

Fig.3.Modelsimulationsofcorkoakpopulationaveragetrendsperlifestage(acorns,seedlings,saplings,immature,andadulttrees)from100independentsimulations throughout100yearsofsimulation,includingthestochasticoccurrenceofﬁresanddroughts,aswellascorkextractionevery9years.(A)Scenario1:shrubclearingevery4 yearsandpresenceoflivestock(0.24–0.40LUha−1_)._(B)_Scenario_2:_shrub_clearing_every₅_years_and_presence_of_livestock_(0.24–0.40_LU_ha−1_)._(C)_Scenario_3:_shrub_clearing every5yearsandthepresenceoflivestock(0.40LUha−1_)._(D)_Scenario_4:_shrub_clearing_every_year,_while_livestock_was_absent.

3.1.3. Fire

FiresarecommonphenomenainsouthernEuropeandcorkoaks

arevulnerabletoitseffects,especiallyseedlingsandsaplings,but

alsoimmatureandadulttreeswhenburningoccursimmediately

aftercorkextraction(Cabezudoetal.,1995;Pausas,1997;Moreira

etal.,2007;Catryetal.,2012).DatafromtheSpanishForest

Ser-vice(magrama.gob.es)wasusedtoparameterizeannualfrequency

andtomodeltheconsequencesofﬁresforeachcorkoaklifestage,

taking intoaccount therespective treediameters (DBH)(Catry

(6)

ﬁres.ha-1;mean±standarddeviation)obtainedperstudyunitin44years,

from100independentstochasticsimulations,wasconsidereda

reliablereproductionoftheregionalhistoricaltrendsofﬁreevents

(0.016±0.021ﬁresha-1)(magrama.gob.es).

3.1.4. Drought

Droughtscontributetocorkoakpopulationlossesasithampers

regenerationandincreasestreemortality(AcácioandHolmgren,

2014).Consideringthehistoricaltrendsofdroughtsinmainland

Portugal from 1976 to 2007 (INAG, 2007), thedrought effects

peryearwereconsideredasdirectconstrainsoverseedlingsand

saplingsviability,themorevulnerableclassestothis

environmen-talfactor.Afterthecalibrationprocedures,theaveragenumberof

0.24±0.078droughts.ha−1(mean±standarddeviation)obtained

perstudyunitin31years,from100independentstochastic

sim-ulations,wasconsidereda reliablereproductionoftheregional

historical trends of drought events (0.24±0.43 droughts.ha-1)

(INAG,2007).

3.1.5. Livestockgrazing

Theconsequences ofintensegrazing arethedeathof young

trees,soilcompaction,decreaseofwaterinﬁltrationandlocal

bio-diversitydecline(Plieningeretal.,2004;Plieninger,2006).Grazing

pressurewasdeﬁnedastheratiobetweenstockingrate,i.e.the

numberoflivestockunits(LU)perunitareaandunit time,and

biomass (totaldry weight of vegetationper unit areaand unit

time)(Allenetal.,2011;Sales-Baptistaetal.,2015).Fromthe50’s

topresenttimes,stockingratesinmontadoecosystemsincreased

from0.10–0.15to0.24–0.40LUha−1(Plieningeretal.,2004).Since

the presence of livestock affects the survival of seedlings and

saplingsduetostemconsumption,damagesfromlivestockgrazing

wereincludedinthemodeltoassesstheinﬂuenceoflivestock

pres-sureinthecorkoakregeneration.Thelivestockmortalityratewas

determinedbytheproportionofseedlingsandsaplingsexpected

todieduetolivestockgrazingperyearofsimulation,rangingfrom

0.28to0.87(Plieningeretal.,2004).

3.1.6. Shrubclearing

Weincludedtheeffectofshrubclearingasa negative

inﬂu-enceonseedlings andsaplingsviabilitysincethis isa common

managementpracticein montadotocontrol shrubinvasion and

topromote pasture growth(Pignatti, 1983; Pulido et al.,2001;

Plieningeretal.,2003,2004;Calvo etal.,2005).Contrarytothe

effect of shrub encroachment that enhances oak regeneration,

promotessoilrehabilitationandpreventserosion,shrubclearing

increasessoil erosionand can derailtheregenerating cork oak

seedlingsand saplings(Plieninger etal.,2004; Pulidoand Díaz,

2005;Pérez-Ramosetal.,2008;Simõesetal.,2009;Nunesetal., 2011;Arosaetal.,2015).Therefore,mortalityduetoshrubclearing

wasincludedinthemodeltoassesstheinﬂuenceofthis

manage-mentpracticeinthecorkoakregeneration.

3.1.7. Corkoaknaturalmortality

The cork oak natural mortality often represents a sum of

continuous environmental and intrinsic processes with

multi-ple contributors acting in the tree, such as age degeneration,

insectdamagesordiseases,withunobviouscause-effect

relation-shipsconcerningmanagement practicesand/orclimatechanges

(Franklinetal.,1987;Brancoetal.,2002;Camilo-Alvesetal.,2013).

Inthemodel,weconsideredtheeffectoftreenaturalmortalityover

immatureandadultcorkoaktreesusingestimatesfromRibeiro

andSurov ´y(2008)whodeterminedthemortalitytrendsofcorkoak

(treesha−1)inPortugalbasedonaerialphotographsinterpretation.

4. Economicbalance

Thecorkoakmontadoisamultifunctionalagro-forestry

ecosys-tem that can beneﬁt society with many goods and services.

However,theeconomicsustainabilityofcorkoakmontadodepends

mainlyonthepriceofcork,associatedwithitsqualityandﬁnaluse.

Corkoaktreesmaylive250–350yearsbutcorkthicknessand

qual-itydecreasesandthelimittousefulcorkproductionis150–200

years.Treeswithcircumferenceatbreastheightmorethan70cm

canbe debarked,which corresponds totrees with25–40years

ofage,dependingonsiteproductivityandtreedensity.Afterthe

ﬁrstcorkdebarking,theminimumperiodtolerablebetweencork

extractionsis9years(PereiraandTomé,2004;Pinheiroetal.,2008).

Furthermore,dependingofthemontadomanagementandlanduses

proﬁt,thebeneﬁtscanbealsoincreasedmainlybylivestock

pro-duction.Ontheotherhand,shrubclearingandcorkextractions

representthemain costsassociatedwiththemontado

manage-ment. Overall, themodelling of cork oak population dynamics,

underscenarioswheredifferentmanagementpracticeswere

sim-ulated,includestheassessmentofthepossibleeconomicincomes

frommontadobybalancingtheﬁnancialgainsandlossesbetween

scenarios.

5. Sensitivityanalysis

AccordingtoLeeetal.(2015)thepurposeofasensitivity

analy-sisisprovideameasureoftherobustnessofthemodel,measuring

thesensitivityoftheobtainedresultstochangesinparameters,

forcingfunctionsand/orsub-models.Localsensitivityanalysis(SA)

wasdonebyone-parameter-at-a-timetechnique(OAT),changing

thepopulationparametersofthemodelwith+/−10%and+/−50%

variation oftherespectivevalues andobserving changesin the

responseoftheselectedstatevariables(Ligmann-Zielinska,2013).

6. Scenarios

Ourmodelwasconstructedinordertosimulaterealistic

envi-ronmentalconditionsandmanagementpracticesconcerningcork

production,and assessitsinﬂuence onthecorkoak population

dynamics.Wedesignedthescenarios,consideringthemost

com-monmanagementpracticesinmontado.Thedesignincludedthe

morelimitingfactorsforcorkoakregeneration,thenumberof

live-stockunitsandthefrequencyofshrubclearing.Fourscenarioswere

designed,consideringﬁreanddroughtsasstochasticeventsand

corkoakextractionevery9years.Inallscenariosameannumber

of50adulttrees.ha-1wasconsideredinthebeginningofeach

sim-ulation(t=0),(Joffeetal.,1999),reproducingatypicalmontado.

Noseedlings,saplingsandimmaturetreeswereconsideredinthe

beginningofsimulations,sincethisisthetypicalsituationin

man-agedmontados.Scenario1recreatesthetraditionalmanagement

ofthemontado,withshrubclearingevery4yearsandannual

pres-enceoflivestockwithvaluesrangingbetween0.24–0.40LUha−1

(thebrowsingindexvariedbetween0.278–0.651onseedlingsand

between0.574–0.875onsaplings,Plieningeretal.,2004).

Scenario2wasbasedonthescenario1trends,maintainingthe

densityof livestockbutwithshrubclearingoccurrenceevery5

years.Thisallowedustotestthepossibleeffectsofasimple

man-agementalterationonthecorkoakpopulationdynamics.

Inscenario3,shrubclearingoccurrencewassetevery5years

asinscenario2butlivestockdensitywasincrease(0.40LUha−1).

Thebrowsingindexcorrespondedto0.651and0.875onseedlings

andsaplings,respectively(Plieningeretal.,2004).

Inordertorecreateamontadowithannualsoilmobilizationto

producefodderforlivestockattheendofsummer,thescenario4

(7)

Inallscenarios,mortalityduetoshrubclearingwasbasedonthe

resultsofArosaetal.(2015),assumingthatallemergingseedlings

andsaplingsdieduetoshrubclearing.

In order to detect possible effects of differentmanagement

practices on the sustainability of the cork oak population, the

Kruskal-Wallis non-parametric test was used since the

vari-ablestested didnot meetthenormalityassumption,evenafter

datatransformation(McDonald,2014):classes(i.e.average

den-sityofseedlings, saplings,immaturetrees and adulttrees) and

theincome from cork oak management (i.e.average economic

revenue),accordingto2timeframes(t=50andt=100)were

com-pared betweenscenarios do detect long-term changes. In case

ofsigniﬁcantdifferencesprovedbytheKruskal-Wallistest,post

hoc comparisons between the groups were done using mean

ranks(SiegelandCastellan,1988).Here,signiﬁcancelevelswere

correctedaccordingtothesequentialBonferonnitechnique.

Differ-enceswereconsideredsigniﬁcantiftheirprobabilityofoccurring

bychancewaslessthan5%.Statisticalanalyseswerecarriedout

usingthesoftwareStatisticaversion8.0(StatSoftInc.2007).

7. Results

7.1. Sensitivityanalysis

Theresults fromthe OATsensitivity analysis showthat the

parametersassociatedwithseedlings,saplings,immaturetreesand

adulttreecauseindividualchangesonthemodellingoutcomes

(Table2).Randomsaplingscohortxlivestockmortalityratewas

theparameterwiththemainkeyinﬂuenceontheoutputsofmost

surrogates.

7.2. Corkoakpopulationdynamics

Thescenario 1 involves themaintenance of montado

(agro-forestry)traditionalmanagement:adeclineofmostclasses(acorns,

seedlings,saplingsandadulttrees)waspredictedthroughoutthe

simulationperiod(Fig.3A).Anywaythemostlimitingfactortothe

overallpopulationisassociatedtotheimmaturetrees,constrained

bymortalityassociatedwithshrubclearing.Fires,droughts,

live-stockbrowsingandcorkextractionseemtoplaysecondaryroles.

Thedecreasein thenumber ofadulttrees willalsorestrictthe

amountofacornsavailableeachyear,andconsequentlythe

num-berofviableseedlingsandsaplings(Fig.3A):theinitialdensityof

50adulttreesha−1isexpectedtodecreasetocirca35treesha−1in

theendofsimulation(i.e.−30%ofadulttrees/100years).

Conversely,whentheimplementationof1-yeardelayinshrub

clearingwastestedinscenario2,theﬁnaldensityofadulttrees

isexpectedtosustainorevenincrease(52trees.ha−1)throughout

(i.e.+4%ofadulttrees/100years)(Fig.3B).Thisslightlynon-linear

increaseinthenumber ofadulttrees(Fig.3),shows thatsome

individualsreachtheimmaturetreesstage,compensatingadults

“naturalmortality”,raisingthenumberofacornsandconsequently

thenumberofseedlingsandsaplingsandcreatingapopulationof

immaturetreesthatwillsubstitutethedecliningadults.

Nevertheless,thecorkoakrecoveryachievedinscenario2by

increasingshrubclearingperiodicityforevery5years,was

com-pletely withdrew by increasing livestock pressure as tested in

scenario3,withanoveralldecreaseforallclasses(Fig.3C).The

smallamountofimmaturetrees(Fig.3C)isnotenoughtosustain

inthelong-termthepopulationofadulttrees.Theestimated

den-sityofadulttreesafter100yearsofsimulationdecreaseabout36

trees.ha−1(i.e.−28%ofadulttrees/100years).

Intensiveshrubclearingeffectsmodelledinscenario4,resulted

inneithersaplingsnorimmaturetrees,beingthemostaggressive

managementasdepictedinFig.3D.Adecreaseinthedensityof

adulttreesto35treesha−1wasestimatedattheendof100years

ofsimulation(i.e.−30%ofadulttrees/100years).

By comparing the four scenarios we didnot detect

signiﬁ-cantdifferencesinthenumberofseedlingsatt=50(H3,96=5.95,

p=0.114,Fig.4A),whileatt=100thenumberofseedlingsin

sce-nario2wassigniﬁcantlyhigher,withmeanvaluesof299±185.63

trees.ha−1(H3,96=58.56,p<0.001,Fig.4A).

Thelownumberofsaplingsinscenario3,withmeanvaluesof

67.99±64.05att=50and63.08±42.63trees.ha−1att=100,and

theabsenceofsaplingsinscenario4att=50andt=100was

insuf-ﬁcienttowithstandthepopulationofimmatureandadulttreesin

thelong-term(H3,96=32.68,p<0.001andH3,96=29.06,p<0.001,

forscenario3and4respectively.Fig.4A).

Theresultsforimmaturetreesaresimilaratt=50andt=100

(H3,96=94.29,p<0.001,Fig.4C).Infact,theobtainedlownumber

ofimmaturetreesinscenarios1,3and4seemstobeaconsequence

ofthelowavailabilityofsaplingsduetointensiveshrubclearing

andpresenceofhighnumberoflivestockunits.Ontheotherhand,

althoughimmaturetreeshavealsobeenprojectedatlowdensities

inscenario2(6.06±5.90and4.81±4.07trees.ha−1,att=50and

t=100respectively)itseemsenoughfortherecoveryofadulttrees

inthelongterm.Thismaybemainlyduetothetimetorecoverfrom

theextendedshrubclearingintervalandthenumberoflivestock

units.

Regardingadulttreestheresultsatt=50weresigniﬁcantly

dif-ferentbetweenscenarios(H3,96=46.65,p<0.001,Fig.4D),being

the scenario 2 the only one with increasing number of adult

treesfacingtheotherscenarios(46.53±3.63trees.ha-1).Att=100,

with51.68±8.19trees.ha−1inscenario2(H3,96=60.77,p<0.001m

Fig.4D)thosedifferencesweremaintained

8. Montadomanagement

Theeconomicbalancesassociatedwithscenario1,2,3and4

resultedfromtheincomesrelatedwithcorkextractionand

live-stockpresence(anddirectpaymentstofarmersenrolledunderthe

CAP)andlossesfromsoilmanagementandstrippingcorkcosts.

Interestingly,scenario2hadthehighestaveragepredictedrevenue

from corkoak (661.33±151.70Dha−1peryear) (H3,396=218.18,

p<0.001,Fig.5),whileaverageproﬁtsassociatedwithscenarios1

and3werealmostidentical(564.94±118.20and569.55±120.70

Dha−1peryear, respectively). Scenario 4 generated the lowest

averageincome(275.13±112.39Dha−1peryear).

9. Discussion

The noveltyof our studywas theinclusion of severalwork

previouslypublishedwithinasinglemodellingplatformwiththe

key-componentsof changingmontadoagro-forestry, with

holis-ticandecologicalrelevance,namelybyfocusingondriverswith

inﬂuenceonthelong-termsustainabilityoftheseecosystems.In

themontado, a dramaticmortality ofcork oak hasbeentaking

placewhentreeswerethinnedtoeasetheuseoffarmmachinery

andpromotewidercrownsthatyieldmoreacorns(Pulidoetal.,

2001;Plieninger,2007).Somestrategieswerecreatedtopreserve

themontadoandenhanceitsculturalandnaturalvaluesand

eco-logicalservices.Inthisperspective,legalprotectionofthetrees

andagroenvironmentalschemeshavebeenimplementedto

pro-tectand promotetreeplantation in themontado(Pinto-Correia

etal.,2011).Ontheotherhand,mostmontadoecosystemsinSpain

andPortugalareprivateandchangesintheruralsocioeconomic

condition,inpartinducedbythepreviousCommonAgricultural

Policy(CAP),havecontributedtotheruralexodusduetopoor

con-ditionsforagriculture,theperipherallocationoftheruralareas

(8)

There-Table2

Localsensitivityanalysis(one-parameter-at-a-time)ofthemainstatevariablesofthemodelto+/−10%and+/−50%variationoftheparametervalues.

StateVariable Parameter Sensitivity

−50% −10% +10% +50%

Seedlings Randomseedlingsshrubclearingmortalityrate 0.046 −0.120 −0.308 −0.308

Randomseedlingdroughtmortalityrate −0.184 −0.121 −0.036 a

Randomacornsgerminationrate −0.597 0.045 −0.045 0.234

Randomacornspredationrate −0.126 −0.004 −0.243 −0.174

Randomacornsinsectdamagerate −0.136 −0.234 −0.177 −0.166

Averageseedlingsﬁremortalityrate −0.121 −0.121 −0.397 0.197

Randomseedlingslivestockmortalityrate −0.010 0.190 0.155 −0.478

Randomsaplingscohort×shrubclearingmortalityrate 0.247 0.338 0.099 0.099

Randomsaplingscohort×droughtmortalityrate −0.184 −0.121 −0.036 a

Averagesaplingsﬁremortalityrate −0.121 −0.121 −0.397 0.197

Randomsaplingscohort×livestockmortalityrate 1.228 2.245 −0.408 −0.377

Randomimmaturetreescohort×ﬁremortalityrate 0.214 0.090 −0.102 −0.026

Randomimmaturetreescohort×deathrate −0.209 0.196 −0.099 0.025

Periodadulttreesdecay −0.079 0.034 0.120 −0.203

Adulttreesperha −0.559 −0.333 0.420 −0.196

Randomadulttreesdeathrate −0.039 −0.051 −0.325 −0.189

Randomadulttreesﬁremortalityrate 0.142 0.060 0.053 0.078

Randomacornspertree −0.663 −0.102 0.271 0.587

Saplings Randomseedlingsshrubclearingmortalityrate 1.658 1.599 1.225 1.225

Randomseedlingdroughtmortalityrate 2.595 1.193 1.360 a

Randomacornsgerminationrate 0.015 1.479 1.878 3.079

Randomacornspredationrate 1.538 2.054 1.405 0.257

Randomacornsinsectdamagerate 2.809 1.260 1.825 0.861

Averageseedlingsﬁremortalityrate 1.656 1.411 2.118 2.076

Randomseedlingslivestockmortalityrate 4.098 1.384 1.335 −0.038

Randomsaplingscohort×droughtmortalityrate 2.595 1.193 1.360 a

Averagesaplingsﬁremortalityrate 1.656 1.411 2.118 2.076

Randomsaplingscohort×livestockmortalityrate 11.535 7.816 0.502 0.286

Randomimmaturetreescohort×ﬁremortalityrate 2.437 1.809 2.291 2.426

Randomimmaturetreescohort×deathrate 1.643 1.729 1.277 2.287

Periodadulttreesdecay 1.629 1.019 2.932 1.979

Adulttreesperha 0.252 2.158 4.240 3.693

Randomadulttreesdeathrate 1.881 1.446 2.521 2.487

Randomacornspertree −0.458 1.599 3.249 5.096

Immaturetrees Randomseedlingsshrubclearingmortalityrate 0.710 0.746 −0.011 −0.011

Randomacornsgerminationrate −0.600 −0.020 0.367 1.614

Randomacornsinsectdamagerate 0.479 0.548 0.056 −0.071

Randomimmaturetreescohort×ﬁremortalityrate 0.568 0.032 0.785 −0.061

Randomimmaturetreescohort×deathrate −0.294 0.779 0.427 0.602

Periodadulttreesdecay −0.164 −0.207 0.033 0.483

Adulttreesperha −0.390 0.045 0.610 1.090

Randomadulttreesdeathrate 0.356 −0.061 0.389 0.111

Adulttrees Randomseedlingsshrubclearingmortalityrate 0.247 0.338 0.099 0.099

Randomacornsgerminationrate −0.179 0.103 0.332 0.351

Randomacornsinsectdamagerate 0.142 0.063 0.064 0.043

Randomimmaturetreescohort×ﬁremortalityrate 0.160 0.165 0.223 0.223

Randomimmaturetreescohort×deathrate 0.154 0.245 0.135 0.229

Periodadulttreesdecay −0.098 0.109 0.215 0.419

Adulttreesperha −0.421 0.138 0.504 0.795

Randomadulttreesdeathrate 0.248 0.086 0.251 0.255

(9)

1 2 3 4 Scenario 0 50 100 150 200 250 300 350 No. se ed lings p er ha A Initial Year 25 Year 50 Year 75 Year 100 1 2 3 4 Scenario 0 20 40 60 80 100 120 140 160 180 No. sa p li n gs p er ha B Initial Year 25 Year 50 Year 75 Year 100 1 2 3 4 Scenario 0 1 2 3 4 5 6 7 No . imma tu re t ree s pe r ha C Initial Year 25 Year 50 Year 75 Year 100 1 2 3 4 Scenario 0 10 20 30 40 50 60 No . a dult t ree s p er ha D Initial Year 25 Year 50 Year 75 Year 100

Fig.4.Averagedensityofseedlings(A),saplings(B),immaturetrees(C)andadulttrees(D)perhectare,from100independentsimulations,accordingto4timeframes(t=1, t=25,t=75andt=100)ineachscenarioconsidered.

fore,theimportanceofthesocioeconomicsituationinrelationto

landusechangeisgenerallyrecognized(Brandtetal.,1999):the

directaidtocerealareasand tosomegrazinganimalsinto

sin-glepayments tofarmers,whicharedecoupledfromproduction,

resultedinland ownersthat havetended toreplacesheepand

goatswithcattle,whichinlast16yearsincreaseditsnumbersby

2.5timesduetoCAPsupport(D ´yrmundsson,2004;Pinto-Correia

andGodinho,2013;Pinto-Correiaetal.,2014;Viegasetal.,2014).

Aftertheabandonmentofcerealcropsandlivestockrearing,during

the70’s,shrubclearingandrotationalploughingbecamecommon

practicestocontrolshrubinvasion,topromotepastureproduction

andpreventﬁre(Pignatti,1983;Pulidoetal.,2001;Plieningeretal.,

2003,2004,2005).Theheavycattlebreedsorhigherlivestockrates,

togetherwithdeepploughingofcultivatedcrops,isundoubtedly

affectingnatural regenerationand manyauthorsestimatethese

factorsasamajorthreattothemontadoecosystem(Pinto-Correia

andMascarenhas,1999;Bugalhoetal.,2011;MorenoandPulido, 2009).

10. Mainlimitationstoregeneration

Overall,ourresultsindicatethatthemain limitationstothe

lackof corkoak regeneration in montado ecosystemsrisefrom

landmanagementpractices,ratherthanenvironmentalfactors.In

fact,theobtainedresultssuggestthatdelayingshrubclearingand

limitingthequantityoflivestockunitsmightconstituteasuitable

measuretoallowthelong-termsustainabilityofthemontado.A

minimaldelayofone yearinsoil ploughingwillbeenoughfor

landownerstoensurethemaintenance/increaseinthenumberof

immatureand adulttrees aswellas thelong-term

sustainabil-ityof themontadoecosystem.Since corkisthemostproﬁtable

productobtainedfromthemontado,anincreaseinthedensityof

adulttreeswouldhavealsoadirectpositiveeffectontheeconomic

balanceofagroforestry“crop”.Ontheotherhand,byincreasing

livestockdensities,corkoakregenerationisseverelyconstrained

duetotheconsumptionofseedlingsandsaplings,andthe

regener-ationcyclewouldeventuallycollapse.Asimilarresultisobtained

whenthereisacontinuoususeofheavymachinery.Other

con-ventionalsemiquantitativestudiesthat investigatedtheeffects

ofmontadochangeshavecometosimilarconclusions,

corroborat-ingthatlivestockpressureactsasamaindriverdeterminingthe

ratesofmontadoloss,althoughnotexplainingwithdetailthe

pro-cessbeneath(Blondel,2006;Plieninger,2007;Gasparetal.,2008;

Bugalhoet al.,2011).Acornpredationand browsing,aswellas

tramplingofseedlingsarethemainpressuresofovergrazing,

asso-ciatedwiththeintensityandtypeofgrazing(livestocktype,breeds,

density,lengthoftime inpasture)(Pulido andDíaz,2005).Our

resultsidentifylimitstothecarryingcapacityofthemontadonear

0.40LUha-1,fromwhichthefutureofthemontadosustainability

wouldbeseriouslythreatened.Thisoutcomealsodemonstrates

thatcaremustbetakenwiththeCAPsupporttoincreasingthe

numberoflivestockperha.AlsoBaeza,2004andCalvoetal.,2012

limitthenumberofLU.ha-1near1,whichbasedonourresults,may

putthemontadoagroecosystematrisk.

Resultsrelative toshrubclearingdemonstrated thatmostof

(10)

1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Year 0 200 400 600 800 1000 1200 Ba la n ce i n € pe r ha S1 S2 S3 S4

Fig.5.Predictedaverageeconomicbalance(D.ha-1)forscenarios1,2,3and4,from 100independentsimulationsthroughout100yearsofsimulation.

Shrubclearingmighthinder landscapeheterogeneityand

dam-agethefungalcommunityinmontado,decreasingthemycorrhizal

community that is part of the symbiosismost intimately

con-nectedtothesoilandmostdirectlyinvolvedinuptakingnutrients

and inﬂuencing soil properties (Azul, 2002; Azul et al., 2010).

Theperiodbetweenconsecutiveshrubclearingssupports

differ-entmanagementobjectives:intensiveshrubclearingeverythree

yearswillfavourpasturelandfor livestockgrazingand

decreas-ingshrubgrowingwillneglectitspositiveprotector effectover

regeneration(Canteiroetal.,2011;Simõesetal.,2015).Thisagrees

withourresults, where shortperiodsbetweenclearings would

inhibittheregeneration of trees,since mechanical clearingis a

non-selectivetoolthatdestroysmostseedlings(Arosaetal.,2015),

thustheabsenceofimmaturetreesdirectlyaffectsnegativelythe

populationofadulttrees.Accordingtoourlongtermresults,we

assumethatextendingshrubclearingforlongerperiodswould

ben-eﬁtnaturalregeneration,whichisinagreementwithothershort

termstudies(Canteiroetal.,2011;Godinhoetal.,2014;Simões

etal.,2016).Infact,otherauthors’conclusionssuggest

maintain-ingshrubcoverforaperiodbetween7and12yearstohelpcork

oakregenerationandpromotetreenurseryandsoilrehabilitation.

Conversely,periodslongerthan12yearscouldfacilitateﬁrespread

duetotheaccumulationofdeadmaterial(Canteiroetal.,2011).

EUprogramshaveonlybeensubsidizingplantingyoungcorkoak

treesbutresultsobtainedbyArosaetal.(2015)conﬁrmedthat

directseeding(ornaturalregeneration)ofcorkoakisagood

alter-nativetofacilitateregeneration,becauseemergencesuccessranged

between43.5% and 63.9%. Agricultural policiesshouldconsider

thisinformation astheyplayanimportantrole inmanagement

decisionandcancontributetofurtherdegradationoftreecover

(Pinto-CorreiaandVos,2004).

11. Futureperspectives

Thedevelopeddynamic modelallowedpredictinglongterm

trendsin corkoak localpopulation, includingthoseattributable

to changes in habitat due to different sources of disturbance

andmanagementscenarios(Winkler,2006).Therefore,thisstudy

representsanearlysteptosupportstrategicoptionsforimpact

mitigationandmanagementbyprovidingprojectionsoflong-term

indicatortrendsunderrealisticsocial–ecologicalchangescenarios

(Bastosetal.,2015).Althoughtheprojectionsrecreates

realisti-callytheknowncorkoakregenerationpattersoccurringinmontado

regions,somelimitationsarisewhenconsideringvalidationasa

fundamentalprocesstoassesstherelativeaccuracyofthemodel

response facingindependent real data(Rykiel, 1996).The

vali-dationof thepredictionswasnot possiblein this studydue to

theimpracticalityofanimmediatevalidationoffuturescenarios,

whichcanonlybedoneafterseveralyearsofcollectingrelevant

site-speciﬁc informationunderwell-known localenvironmental

conditions(Glenzetal.,2001;Chaloupka,2002).

Ouracademicanduser-friendlymodel,afterspeciﬁccalibration

andvalidation,couldbeusedtosupportspeciﬁcfarmerdecisions

inthemanagementofcorkoaksavannasandsupportagroforestry

extensionservices.Furthermore,large-scaleanalysesare

manda-torytofullyunderstandtheroleofmulti-scalecumulativedrivers

inﬂuencingtheregenerationprocessandtheroleofpoliciesin

man-agingthis ecosystem(Godinhoet al.,2014).Inthis context,the

combinationoflocalpopulationdynamics(temporal)andregional

distributional(spatial)modelscanhelptoassessandpredicthow

anthropogenicand environmentalchangeswillaffectthe

abun-danceanddisplacementofvulnerablespeciesorcommunitiesin

disturbed ecosystems and regions (Guisan and Thuiller, 2005).

Besides, modelsare useful to identifyareas where the conﬂict

betweenthepreviousforecastedtrendsanddrivers ofpressure

are of major conservationist concern, by allowing the

integra-tionoftheecologicalconsequencesfromlocaltoregionallevels

(Bjørnstadetal.,1999;Vicenteetal.,2011).Model-basedresearch

resultedveryusefulasaninvestigativetooltopredicttheoutcome

ofalternativescenarios,guidingcurrentrestorationoptionsfrom

predictedfuturetargets(Bastosetal.,2012).Futureworkshouldbe

directedtodeveloptherestorationstrategiesofcorkoakmontados,

whichwillcontributetoﬁghtdesertiﬁcation,protectbiodiversity,

mitigateclimatechangeandpromotethesustainabledevelopment

oflocalcommunities.

12. Conclusion

Thecapacitytoaccuratelypredictresponsesofspeciesto

land-scapechangesiscrucialforconservationplanningandtosupport

key ecosystems management (Kandziora et al., 2013).In

mon-tadoecosystems,thereis alack ofevaluation onthelong-term

ecologicalchangesassociatedwithmanagementpractices,whose

consequencesmightrepresentmajorlimitingfactorstothe

ecolog-icalandeconomicsustainabilityofthisvaluableagro-ecosystem.

Thedevelopeddynamicmodelallowedintegratingandsimulating

themainfactorsaffectingthelong-termsustainabilityofthe

mon-tadosystem,namelythemajorlimitingfactorsdrivingthelackof

regenerationofcorkoaktreesinmontados.Thenumberoflivestock

unitsperhaandshortshrubclearingintervalswereidentiﬁedas

themainlimitingfactorsforthelongtermsustainabilityofthis

agroforestryecosystem.Thus,itshouldbeconsideredamaximum

quantityoflivestockof0.40LU.ha−1andtheminimumperiodicity

betweensoilploughingto5years.Speciﬁcmanagementpractices

aresuggestedtoovercomethesustainabilityoftheidentiﬁed

prob-lems.

Acknowledgements

This work was supported by the FCT (Portuguese

Foun-dation for Science and Technology) under a Ph.D. grant

[SFRH/BD/70708/2010].R.BastosisfundedbyFCTthroughthe

doc-toralgrantSFRH/BD/102428/2014.S.CostaisfundedbytheFCT

throughthepost-doctoralgrantSFRH/BPD/102438/2014.Weare

gratefultoAlfredoSendimfromHerdadedoFreixodoMeiowho

(11)

AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,in

theonlineversion,athttp://dx.doi.org/10.1016/j.ecolmodel.2016.

10.008.

References

Acácio,V.,Holmgren,M.,2014.PathwaysforresilienceinMediterraneancorkoak landusesystems.Ann.For.Sci.71,5–13.

Acácio,V.,Holmgren,M.,Jansen,P.A.,Schrotter,O.,2007.Multiplerecruitment limitationcausesarrestedsuccessioninMediterraneancorkoaksystems. Ecosystems10,1220–1230.

Acácio,V.,Holmgren,M.,Rego,F.,Moreira,F.,Mohren,G.M.,2009.Aredroughtand wildﬁresturningMediterraneancorkoakforestsintopersistentshrublands? Agrofor.Syst.76,389–400.

Acácio,V.,Holmgren,M.,Moreira,F.,Mohren,G.M.J.,2010.Oakpersistencein Mediterraneanlandscapes:thecombinedroleofmanagement,topography, andwildﬁres.Ecol.Soc.15,40.

Allen,V.G.,Batello,C.,Berretta,E.J.,Hodgson,J.,Kothmann,M.,Li,X.,McIvor,J., Milne,J.,Morris,C.,Peeters,A.,Sanderson,M.,2011.Aninternational terminologyforgrazinglandsandgrazinganimals.GrassForageSci.66,2–28.

Arosa,M.L.,Ceia,R.C.,Costa,R.S.,Freitas,H.,2015.Factorsaffectingcorkoak (Quercussuber)regeneration:acornsowingsuccessandseedlingsurvival underﬁeldconditions.PlantEcol.Divers.,http://dx.doi.org/10.1080/17550874. 2015.1051154.

Assmann,E.,Gardiner,S.,1970.ThePrinciplesofForestYieldStudy.Pergamon Press,Oxford.

Azul,A.M.,Sousa,J.P.,Agerer,R.,Martín,M.P.,Freitas,H.,2010.Landusepractices andectomycorrhizalfungalcommunitiesfromoakwoodlandsdominatedby QuercussuberL.consideringdroughtscenarios.Mycorrhiza20(2),0261–0262,

http://dx.doi.org/10.1007/s00572-009-0261-2.

Azul,A.M.,2002.DiversidadeDeFungosEctomicorrízicosEmEcossitemasDe MontadoPh.D.Thesis.DepartmentofBotany,TheUniversityofCoimbra, Coimbra,Portugal.

Baeza,M.J.,2004.Elmanejodelmatorralenlaprevencióndeincendiosforestales. In:Vallejo,V.R.,Alloza,J.A.(Eds.),AvancesEnElEstudioDeLaGestiónDel MonteMediterráneo.FundaciónCentrodeEstudiosAmbientalesdel Mediterráneo,Valencia.

Bastos,R.,Santos,M.,Ramos,J.A.,Vicente,J.,Guerra,C.,Alonso,J.,Honrado,J.,Ceia, R.S.,Timóteo,S.,Cabral,J.A.,2012.Testinganovelspatially-explicitdynamic modellingapproachinthescopeofthelaurelforestmanagementforthe endangeredAzoresbullﬁnch(Pyrrhulamurina)conservation.Biol.Conserv. 147,243–254.

Bastos,R.,Pinhanc¸os,A.,Santos,M.,Fernandes,R.,Vicente,J.,Morinha,F.,Honrado, J.P.,Travassos,P.,Barros,P.,Cabral,J.A.,2015.Evaluatingtheregional cumulativeimpactofwindfarmsonbirds:howspatially-explicitdynamic modellingcanimproveimpactassessmentsandmonitoring?J.Appl.Ecol.,

http://dx.doi.org/10.1111/1365-2664.12451.

Bjørnstad,O.N.,Ims,R.A.,Lambin,X.,1999.Spatialpopulationdynamics:analyzing patternsandprocessesofpopulationsynchrony.TrendsEcol.Evol.14, 427–431.

Blondel,J.,2006.ThedesignofMediterraneanlandscapes:amillennialstoryof humanandecologicalsystemsduringthehistoricperiod.Hum.Ecol.34, 713–729.

Branco,M.,Branco,C.,Merouani,H.,Almeida,M.H.,2002.Germinationsuccess, survivalandseedlingvigourofQuercussuberacornsinrelationtoinsect damage.For.Ecol.Manage.166,159–164.

Brandt,J.,Primdahl,J.,Reenberg,A.,1999.Ruralland-useandlandscape dynamics-analysisof‘drivingforces’inspaceandtime.In:Kronert,R.,Baudry, J.,Bowler,I.R.,Reenberg,A.(Eds.),Land-UseChangesandTheirEnvironmental ImpactinRuralAreasinEurope.UNESCO,Paris,pp.81–102.

Bugalho,M.N.,Caldeira,M.C.,Pereira,J.S.,Aronson,J.,Pausas,J.G.,2011.

Mediterraneancorkoaksavannasrequirehumanusetosustainbiodiversity andecosystemservices.Front.Ecol.Environ.9,278–286.

Ca ˜nellas,I.,Roig,S.,Poblacione,M.J.,Gea-Izquierdo,G.,Olea,L.,2007.Anapproach toacornproductioninIberiandehesas.Agrofor.Syst.70,3–9.

Cabezudo,B.,LaTorre,A.P.,Nieto,J.M.,1995.Regeneracióndeunalcornocal incendidadoenelSurdeEspa ˜na(Istán,Málaga).ActaBot.Malacitana20, 143–151.

Callaway,R.M.,Walker,L.R.,1997.Competitionandfacilitation:asynthetic approachtointeractionsinplantcommunities.Ecology78,1958–1965.

Calvo,L.,Tárrega,R.,Luís,E.,Valbuena,L.,Marcos,E.,2005.Recoveryafter experimentalcuttingandburninginthreeshrubcommunities.PlantEcol.180, 175–185.

Calvo,L.,Baeza,J.,Marcos,E.,Santana,V.,Papanastasis,V.P.,etal.,2012.Post-ﬁre managementofshrublands.In:Moreira,F.(Ed.),Post-ﬁreManagementand RestorationofSouthernEuropeanForest.ManagingForestEcosystems,Vol24. Springer,NewYork,pp.293–319.

Camilo-Alves,C.S.P.,daClara,M.I.E.,Ribeiro,N.M.C.A.,2013.Declineof MediterraneanoaktreesanditsassociationwithPhytophthoracinnamomi:a review.Eur.J.For.Res.132,411–432.

Canteiro,C.,Pinto-Cruz,C.,PaulaSimões,M.,Gazarini,L.,2011.Conservationof Mediterraneanoakwoodlands:understorydynamicsunderdifferentshrub management.Agrofor.Syst.82,161–171.

Catry,F.X.,Moreira,F.,Pausas,J.G.,Fernandes,P.M.,Rego,F.,Cardillo,E.,Curt,T., 2012.Corkoakvulnerabilitytoﬁre:theroleofbarkHarvesting,tree characteristicsandabioticfactors.PLoSOne7(6),e39810,http://dx.doi.org/10. 1371/journal.pone.0039810.

Ceia,R.S.,Ramos,J.A.,2014.Birdsaspredatorsofcorkandholmoakpests.Agrofor. Syst.,10.1007/s10457-014-9749-7.

Chaloupka,M.,2002.StochasticsimulationmodellingofsouthernGreatBarrier Reefgreenturtlepopulationdynamics.Ecol.Modell.148,79–109.

Díaz-Fernández,P.M.,Climent,J.,Gil,L.,2004.Biennialacornmaturationandits relationshipwithﬂoweringphenologyinIberianpopulationsofQuercussuber. TreesStruct.Funct.18,615–621.

D ´yrmundsson,O.R.,2004.SustainabilityofsheepandgoatproductioninNorth Europeancountries−fromtheArctictotheAlps.In:55thAnnualMeetingof theEuropeanAssociationforAnimalProduction,Bled,Slovenia.

Deleuze,C.,Herve,J.C.,Colin,F.,Ribeyrolles,L.,1996.Modellingcrownshapeof Piceaabies:spacingeffects.Can.J.For.Res.26,1957–1966.

Dolezal,J.,Ishii,H.,Vetrova,V.P.,Sumida,A.,Hara,T.,2004.Treegrowthand competitioninaBetulaplatyphylla–Larixcajanderipost-ﬁreforestincentral Kamchatka.Ann.Bot.94,333–343.

Elena-Rosselló,J.A.,Río,J.M.,GarcíaValdecantos,J.L.,Santamaría,I.G.,1993. Ecologicalaspectsoftheﬂoralphenologyofthecorkoak(Q.suberL.):whydo annualandbiennalbiotypeappear?Pp.114s-121s.In:Geneticsofoaks.proc. IUFROworkingpartymeeting,2–6Sept.1991.ArboretumNationaldesBarres, France.INRAParis.

Fei,S.,Steiner,K.C.,2008.Relationshipsbetweenadvanceoakregenerationand bioticandabioticfactors.TreePhysiol.28,1111–1119.

Fernandes,C.,Cabral,J.A.,Crespí,A.L.,Hughes,S.J.,Santos,M.,2013.Converting simplevegetationsurveysinfunctionaldynamics.ActaOecol.48,37–46.

Flores-Martinez,A.,Ezcurra,E.,Sanchez-Colon,S.,1994.EffectofNeobuxbaumia tetetzoongrowthandfecundityofitsnurseplantMimosaluisana.J.Ecol.82, 325–330.

Franklin,J.F.,Shugart,H.H.,Harmon,M.E.,1987.Treedeathasanecological process.Bioscience37,550–556.

Gómez-Aparicio,L.,Pérez-Ramos,I.M.,Mendoza,I.,Matías,L.,Quero,J.L.,Castro,J., Zamora,R.,Mara ˜nón,T.,2008.Oakseedlingsurvivalandgrowthalong resourcegradientsinMediterraneanforests:implicationsforregenerationin currentandfutureenvironmentalscenarios.Oikos117,1683–1699.

Gaspar,P.,Escribano,M.,Mesías,F.J.,RodriguezdeLedesma,A.,Pulido,F.,2008.

SheepfarmsintheSpanishrangelands(dehesas):typologiesaccordingto livestockmanagementandeconomicindicators.SmallRumin.Res.74,52–63.

Gimeno,T.E.,Pías,B.,Lemos-Filho,J.P.,Valladares,F.,2009.Plasticityandstress toleranceoverridelocaladaptationintheresponsesofMediterraneanholm oakseedlingstodroughtandcold.TreePhysiol.29,87–98.

Glenz,C.,Massolo,A.,Kuonen,D.S.,Schlaepfer,R.,2001.Awolfhabitatsuitability predictionstudyinValais(Switzerland).Landsc.UrbanPlann.55,55–65.

Godinho,S.,Guiomar,N.,Machado,R.,Santos,P.,Sá-Sousa,P.,Fernandes,J.P., Neves,N.,Pinto-Correia,T.,2014.Assessmentofenvironment,land

management,andspatialvariablesonrecentchangesinmontadolandcoverin southernPortugal.Agrofor.Syst., http://dx.doi.org/10.1007/s10457-014-9757-7.

González,D.,Cabral,J.A.,Torres,L.,Santos,M.,2015.Acohort-basedmodelling approachformanagingolivemothPraysoleae(Bernard,1788)populationsin oliveorchards.Ecol.Modell.296,46–56.

Guisan,A.,Thuiller,W.,2005.Predictingspeciesdistribution:offeringmorethan simplehabitatmodels.Ecol.Lett.8,993–1009.

Harper,J.L.,1977.PopulationBiologyofPlants.AcademicPress,London,UK.

Herrera,J.,1995.Acornpredationandseedlingproductioninalow-density populationofcorkoak(QuercussuberL.).For.Ecol.Manage.76,197–201.

Holmgren,M.,Scheffer,M.,Huston,M.A.,1997.Theinterplayoffacilitationand competitioninplantcommunities.Ecology78,1966–1975.

INAG,2007.PreliminaryCountryReport,Contribuic¸ãoportuguesaparaorelatório daComissãoEuropeiaWaterScarcity&Droughts−In-depthassessment, SecondInterimReport,InstitutodaÁgua,MinistériodoAmbiente OrdenamentodoTerritórioeDesenvolvimentoRegional.

Joffre,R.,Rambal,S.,Ratte,J.P.,1999.ThedehesasystemofsouthernSpainand Portugalasanaturalecosystemmimic.Agrofor.Syst.45,57–79.

Jørgensen,S.E.,1994.Modelsasinstrumentsforcombinationofecologicaltheory andenvironmentalpractice.Ecol.Modell.75,5–20.

Jørgensen,S.E.,2001.FundamentalsofEcologicalModelling,3rded.Elsevier, Amsterdam.

Kandziora,M.,Burkhard,B.,Müller,F.,2013.Interactionsofecosystemproperties, ecosystemintegrityandecosystemserviceindicators−atheoreticalmatrix exercise.Ecol.Indic.28,54–78.

Kenkel,N.C.,1988.Patternofself-thinninginJackpine:testingtherandom mortalityhypothesis.Ecology69,1017–1024.

Lee,J.S.,Filatova,T.,Ligmann-Zielinska,A.,Hassani-Mahmooei,B.,Stonedahl,F., Lorscheid,I.,Voinov,A.,Polhil,G.,Sun,Z.,Parker,D.C.,2015.Thecomplexities ofagent-basedmodelingoutputanalysis.J.Artif.Soc.Soc.Simulat.18(4),4.

Ligmann-Zielinska,A.,2013.Spatially-explicitsensitivityanalysisofan

agent-basedmodeloflandusechange.Int.J.Geogr.Inf.Sci.27(9),1764–1781.

Müller,F.,2005.Indicatingecosystemandlandscapeorganisation.Ecol.Indic.5, 280–294.

(12)

Marín-Pageo,F.,Camacho,V.,2011.DiBérenger´ısmethodapplicationinsouthwest iberiancorkoakstands.L’italiaForestaleeMontana66(1),31–40.

McDonald,J.H.,2014.HandbookofBiologicalStatistics,3rded.SparkyHouse Publishing,USA,pp.157–164.

Montero,G.,Ca ˜nellas,I.,2003.ElAlcornoqueManualDeReforestaciónYCultivo. Mundi-Prensa,Madrid.

Moreira,F.,Duarte,I.,Catry,F.,Acácio,V.,2007.Corkextractionasakeyfactor determiningpost-ﬁrecorkoaksurvivalinamountainregionofsouthern Portugal.For.Ecol.Manage.253,30–37.

Moreno,G.,Pulido,F.J.,2009.Thefunctioning,managementandpersistenceof dehesas.In:Rigueiro-Rodríguez,A.,McAdam,J.,Mosquera-Losada,M.R.(Eds.), AgroforestryinEurope,CurrentStatusandFutureProspects.Springer,The Netherlands,pp.127–160.

Muick,P.C.,Bartolome,J.W.,1987.AnAssessmentofNaturalRegenerationofOaks inCaliforniaDivisionofForestry.UniversityofCalifornia,Berkeley,CA,101pp.

Natividade,J.V.,1950–Subericultura.Dir.Gral.DosServ.Forestaira.Lisboa.ed. espa ˜nola,1992.MAPA,Madrid.

Nunes,A.N.,DeAlmeida,A.C.,Coelho,C.O.,2011.Impactsoflanduseandcover typeonrunoffandsoilerosioninamarginalareaofPortugal.Appl.Geogr.31, 687–699.

Olea,L.,López-Bellido,R.J.,Poblaciones,M.J.,2005.Europetypesofsilvopastoral systemsintheMediterraneanarea:dehesa.In:Mosquera,M.R.,Rigueiro,A., McAdam,J.(Eds.),SilvopastoralismandSustainableLandManagement.CABI Publishing,Lugo,Spain,pp.30–35.

Pérez-Ramos,I.M.,Urbieta,I.R.,Mara ˜nón,T.,Zavala,M.A.,Kobe,R.K.,2008.Seed removalintwocoexistingoakspecies:ecologicalconsequencesofseedsize, plantcoverandseeddroptiming.Oikos117,1386–1396.

Paracchini,M.L.,Petersen,J.E.,Hoogeveen,Y.,Bamps,C.,Burfield,I.,vanSwaay,C., 2008.HighNatureValueFarmlandinEurope.AnEstimateoftheDistribution PatternsontheBasisofLandCoverandBiodiversityData.OfficeforOfficial PublicationsoftheEuropeanCommunities,Luxemburg.

Pausas,J.G.,Mara ˜nón,T.,Caldeira,M.,Pons,J.,etal.,2009.NaturalRegeneration.In: Aronson(Ed.),pp.115–124,Chapter10.

Pausas,J.G.,1997.ResproutingofQuercussuberinNESpainafterﬁre.J.Veg.Sci.8, 703–706.

Pereira,H.,Tomé,M.,2004.CorkOakEncyclopediaofForestSciences.Elsevier, Oxford,pp.613–620.

Pignatti,S.,1983.HumanimpactonthevegetationoftheMediterraneanBasin.In: Holzner,W.,Werger,M.J.A.,Ikusima,I.(Eds.),Man’sImpactonVegetation,3. Geobotany,pp.151–161.

Pinheiro,A.C.,Ribeiro,N.A.,2013.Forestpropertyinsurance:anapplicationto Portuguesewoodlands.Int.J.Sustain.Soc.5,284–295.

Pinheiro,A.C.,Ribeiro,N.A.,Surov ´y,P.,Ferreira,A.G.,2008.Economicimplications ofdifferentcorkoakforestmanagementsystems.Int.J.Sustain.Soc.1, 149–157.

Pinto-Correia,T.,Fonseca,A.,2009.HistoricalperspectiveofMontados:theÉvora casestudy.In:Aronson,J.,SantosPereira,J.,Pausas,J.G.(Eds.),CorkOak WoodlandsinTransition:Ecology,Management,andRestorationofanAncient MediterraneanEcosystem.IslandPress,WashingtonDC,pp.49–58.

Pinto-Correia,T.,Godinho,S.,2013.Changingagriculturechanginglandscapes: whatisgoingoninthehighvaluedmontado.In:Ortiz-Miranda,D., Moragues-Faus,A.,ArnalteAlegre,E.(Eds.),AgricultureinMediterranean Europe:BetweenOldandNewParadigms(ResearchinRuralSociologyand Development,Vol.19.EmeraldGroupPublishingLimited,Bingley,pp.75–90.

Pinto-Correia,T.,Mascarenhas,J.,1999.Contributiontothe

extensiﬁcation/intensiﬁcationdebate:newtrendsinthePortuguesemontado. Landsc.UrbanPlann.46,125–131.

Pinto-Correia,T.,Vos,W.,2004.MultifunctionalityinMediterraneanlandscapes: pastandfuture.In:Jongman,R.(Ed.),TheNewDimensionsofthe

Europeanlandscape.WagenigenURFrontisSeries,Vol.4.Springer,Dordrecht, TheNetherlands,pp.135–164.

Pinto-Correia,T.,Ribeiro,N.,Sá-Sousa,P.,2011.Introducingthemontado,thecork andholmoakagroforestrysystemofSouthernPortugal.Agrofor.Syst.82, 99–104.

Pinto-Correia,T.,Menezes,H.,Barroso,F.,2014.Thelandscapeasanassetin SouthernEuropeanfragileagriculturalsystems:contrastsandcontradictions inlandmanagersattitudesandpractices.Landsc.Res.39,205–217.

Plieninger,T.,Pulido,F.J.,Konold,W.,2003.Effectsofland-usehistoryonsize structureofholmoakstandsinSpanishdehesas:implicationsforconservation andrestoration.Environ.Conserv.30,61–70.

Plieninger,T.,Pulido,F.J.,Schaich,H.,2004.Effectsofland-useandlandscape structureonholmoakrecruitmentandregenerationatfarmlevelinQuercus ilexL.dehesas.J.AridEnviron.57,345–364.

Plieninger,T.,2006.Habitatloss,fragmentation,andalteration—quantifyingthe impactofland-usechangesonaSpanishdehesalandscapebyuseofaerial photographyandGIS.Landsc.Ecol.21,91–105.

Plieninger,T.,2007.Compatibilityoflivestockgrazingwithstandregenerationin Mediterraneanholmoakparklands.J.Nat.Conserv.15,1–9.

Pulido,F.J.,Díaz,M.,2005.RegenerationofaMediterraneanoak:awholecycle approach.Écoscience12,92–102.

Pulido,F.J.,Díaz,M.,Trucios,S.J.H.,2001.Sizestructureandregenerationof SpanishholmoakQuercusilexforestsanddehesas:effectsofagroforestryuse ontheirlong-termsustainability.For.Ecol.Manage.146,1–13.

Pulido,F.,McCreary,D.,Ca ˜nellas,I.,McClaran,M.,Plieninger,T.,2013.Oak regeneration:ecologicaldynamicsandrestorationtechniques.In:Campos,P., Huntsinger,L.,OviedoPro,J.L.,Starrs,P.F.,Díaz,M.,Standiford,R.B.,Montero, G.(Eds.),MediterraneanOakWoodlandWorkingLandscapes.Springer Netherlands,TheNetherlands,pp.123–144.

Ribeiro,N.A.,Surov ´y,P.,2008.InventárioNacionalDeMortalidadeDeSobreiroNa FotograﬁaaéreaDigitalDe2004/2006.ICAM,MADRP,AFN.ÉvoraUniversity Publishings.

Rykiel,E.J.,1996.Testingecologicalmodels:themeaningofvalidation.Ecol. Modell.90,229–244.

Sales-Baptista,E.,Cancelad’Abreu,M.,Ferraz-de-Oliveira,M.I.,2015.Overgrazing intheMontado?Theneedformonitoringgrazingpressureatpaddockscale. Agrofor.Syst.,http://dx.doi.org/10.1007/s10457-014-9785-3.

Santos,M.,Bastos,R.,Cabral,J.A.,2013.Convertingconventionalecological datasetsindynamicanddynamicspatiallyexplicitsimulations:current advancesandfutureapplicationsoftheStochasticDynamicMethodology (StDM).Ecol.Modell.258,91–100.

Serrada,R.,2002.ApuntesDeSelvicultura.E.U.I.T.F,Madrid.

Siegel,S.,Castellan,N.J.,1988.NonparametricStatisticsfortheBehavioralSciences, 2nded.McGraw-Hill,NewYork,NY.

Simões,M.P.,Madeira,M.,Gazarini,L.,2009.AbilityofCistusL.shrubstopromote soilrehabilitationinextensiveoakwoodlandsofMediterraneanareas.Plant Soil323,249–265.

Simões,M.P.,Belo,A.F.,Fernandes,M.,Madeira,M.,2016.Agrofor.Syst.90,107, doi:10.1007/s10457-015-9818-6.

Smit,C.,Díaz,M.,Jansen,P.,2009.Establishmentlimitationofholmoak(Quercus ilexsubsp.ballota(Desf.)Samp.)inaMediterraneansavanna–forestecosystem. Ann.For.Sci.66,1–7.

Sterman,J.D.,2001.Systemdynamicsmodelling:toolsforlearninginacomplex world.Calif.Manage.Rev.43,8–25.

Stoll,S.,Frenzel,M.,Burkhard,B.,Adamescu,M.,Augustaitis,A.,Baeßler,C.,Bonet, F.J.,Cazacu,C.,Cosor,G.L.,Díaz-Delgado,R.,Carranza,M.L.,Grandin,U.,Haase, P.,Hämäläinen,H.,Loke,R.,Müller,J.,Stanisci,A.,Staszewski,T.,Müller,F., 2015.AssessmentofecosystemintegrityandservicegradientsacrossEurope usingtheLTEREuropenetwork.Ecol.Modell.295,75–87.

Tutin,T.G.,Heywood,V.H.,Burges,N.A.,Valentine,D.H.,Walters,S.M.,Webb,D.A., 1964.Floraeuropaea.LycopodiaceaetoPlatanaceae,Vol.1.Cambridge UniversityPress,Cambridge,UK.

VanDoorn,M.M.,Bakker,A.M.,2007.Thedestinationofarablelandinamarginal agriculturallandscapeinSouthPortugal:anexplorationoflandusechange determinants.Landsc.Ecol.22,1073–1087.

Vicente,J.,Randin,C.F.,Gonc¸alves,J.,Metzger,M.,Lomba,A.,Honrado,J.,Guisan, A.,2011.Wherewillconﬂictsbetweenalienandrarespeciesoccurafter climateandland-usechange?Atestwithanovelcombinedmodeling approach.Biol.Invasions13,1209–1227.

Viegas,I.,AguiarFontes,M.,LimaSantos,J.,2014.Produçãodecarnedebovinoem sistemassilvo-pastoristradicionais–umaboaopçãoparaaproteção

ambiental?-BeefProductionintraditionalsilvopastoralsystems–Asecondbest fortheenvironment?RevistaPortuguesadeCiênciasVeterinárias1091–10. Vivas,E.,Maia,R.,2008.CaracterizaçãodasPrinciaisSituaçõesdeSecaHistórica emPortugalContinental-AImportánciadaUtilizaçãodeIndicadores.Actasdo 2as_Jornadas_de_Hidráulica,_Recursos_Hídricos_e_Ambiente,_FEUP._Portugal. Winkler,E.,2006.Recenttrendsinplant-ecologicalmodelling:speciesdynamicsin