SPECIAL PAPER Why are there so many species in the tropics James H - PDF document

SPECIAL PAPER Whyaretheresomanyspecies inthetropics? JamesH.Brown DepartmentofBiology,UniversityofNew Mexico,Albuquerque,NM87131,USA Correspondence:JamesH.Brown,Department ofBiology,UniversityofNewMexico, Albuquerque,NM87131,USA. E-mail:jhbrown@unm.edu Thisisanopenaccessarticleundertheterms oftheCreativeCommonsAttribution- NonCommercialLicense,whichpermitsuse, distributionandreproductioninanymedium, providedtheoriginalworkisproperlycited andisnotusedforcommercialpurposes. Knownforcenturies,thegeographicalpatternofincreasingbiodiversityfromthe polestotheequatorisoneofthemostpervasivefeaturesoflifeonEarth.Along- standinggoalofbiogeographershasbeentounderstandtheprimaryfactorsthat generateandmaintainhighdiversityinthetropics.Many‘historical’and‘ecolog- ical’hypotheseshavebeenproposedanddebated,butthereisstilllittleconsen- sus.Recentdiscussionshavecentredaroundtwomainphenomena:phylogenetic nicheconservatismandecologicalproductivity.Thesetwofactorsplayimportant roles,butaccumulatingtheoreticalandempiricalstudiessuggestthatthesingle mostimportantfactoriskinetics:thetemperaturedependenceofecologicaland evolutionaryrates.Therelativelyhightemperaturesinthetropicsgenerateand maintainhighdiversitybecause‘theRedQueenrunsfasterwhensheishot’. Keywords Ecologicalinteractions,evolutionaryrates,Janzen – Connelldynamics,latitudi- naldiversitygradient,metabolictheory,RedQueen,speciesdiversity,tropics. Thereis,however,onenaturalfeatureofthiscountry,theinter- estandgrandeurofwhichmaybefullyappreciatedinasingle walk:itisthe‘virginforest’.Herenoonewhohasanyfeeling ofthemagnicentandthesublimecanbedisappointed;the sombreshade,scarceilluminedbyasingledirectrayevenof thetropicalsun,theenormoussizeandheightofthetrees, mostofwhichriselikehugecolumnsahundredfeetormore aroundthebaseofsome,thespinyorfurrowedstemsofoth- ers,thecuriousandevenextraordinarycreepersandclimbers whichwindaroundthem,hanginginlongfestoonsfrom branchtobranch,sometimescurlingandtwistingonthe groundlikegreatserpents,thenmountingtotheverytopsof thetrees,thencethrowingdownrootsandbreswhichhang wavingintheair,ortwistingroundeachotherformropesand cablesofeveryvarietyofsizeandoftenofthemostperfectreg- ularity.These,andmanyothernovelfeatures – theparasitic plantsgrowingonthetrunksandbranches,thewonderfulvari- etyofthefoliage,thestrangefruitsandseedsthatlierotting ontheground – takenaltogethersurpassdescription,andpro- ducefeelingsinthebeholderofadmirationandawe.Itishere, too,thattherarestbirds,themostlovelyinsects,andthemost interestingmammalsandreptilesaretobefound.Herelurk thejaguarandtheboa-constrictor,andhereamidthedensest shadethebell-birdtollshispeal. AlfredRusselWallaceontropicalforestinBrazilinhis1849 lettertothemembersoftheMechanics’Institution,pub- lishedinWallace,1905(p.270). INTRODUCTION FormorethanthreecenturiesWesternsciencehasknownthat biodiversityisgreatestinthetropics.Europeanexplorersand tradersreturnedfromAfrica,AsiaandtheAmericaswith malsandplants.ManyweresenttoCarlLinnaeus,whobyhis lasteditionof SystemaNatura ein1758,hadcatalogued7700 speciesofplantsand4400speciesofanimals,including17spe- ciesofhummingbirdsfromSouthAmericaandtheCaribbean (Linnaeus,1758).Explorer-naturalistswhoaccompaniedvoy- agestothetropicswereawedbythevarietyofspecies,form andfunction,andwroteaccountscomparabletoWallace’slet- terfromBrazil(above).Manygiantsof19thcenturynatural science,includingJosephBanks,ThomasBelt,Alexandervon Humboldt,JosephDaltonHooker,CharlesDarwin,Henry WalterBatesandAlfredRusselWallace,wereindeliblyinu- encedbytheirexposuretotropicalbiodiversity. Inthemid-20thcentury,mostofthegreatsynthesizers wholaidthefoundationsofmodernevolution,systematics, biogeographyandecologycommentedonthepatternof increasingbiodiversityfromthepolestotheequator(e.g. TheodosiusDobzhansky,ErnstMayr,CharlesElton,Evelyn Hutchinson,PhilipDarlington,AlfredFischer,GeogeGay- lordSimpson,RobertMacArthurandEdwardO.Wilson). 8 http://wileyonlinelibrary.com/journal/jbi ª 2013TheAuthorsJouralofBiogeographyPublishedby JohnWiley&SonsLtd doi:10.1111/jbi.12228 JournalofBiogeography ( J.Biogeogr. )(2014) 41 ,8–22 Butthedatawerestillsketchyandtherewaslittleconsensusastothecausalprocesses.Withinthelastfewdecades,bioge-ographershavetakenadvantageofnewbiologicalinventoriesofpreviouslypoorlystudiedtaxaandregionsandoftechno-logicaladvancesincomputers,electronicdatabasesandgeo-graphicalinformationsystemstoquantifyandclarifytheempiricalpatterns.Itisnowclearthatthetropicsharbournotonlythemostspeciesofplantsandanimals,butalsothemostdiversegenomes,cladesofhighertaxa(e.g.Willigetal.,2003;Lomolinoetal.,2010),andevenlanguagesandculturesofsubsistencehumansocieties(Collard&Foley,2002;Pagel&Mace,2004;Gavinetal.,2013).Andthepat-ternisancient,apparentinthefossilrecorddatingbackhundredsofmillionsofyears(e.g.Stehlietal.,1969;Crane&Lidgard,1989;Crame,2001).Evenasthepatternshavebecomeclearer,however,theexplanationshaveremainedelusiveandcontroversial.Sowhyislifemostdiverseinthetropics?Thenumberofhypothesestoexplainthelatitudinaldiversitygradient(LDG)hasonlyincreasedinrecentdecades.Pianka(1966)listed6,Brown(1988)8,Rohde(1992)23,Willigetal.(2003)27,andLomolinoetal.(2010)32.Theseproliferatinghypothesesareamixedbag,rangingfromspecicideasforrestrictedtaxonomicorfunctionalgroupstogeneralphe-nomenapotentiallyapplicabletoallorganisms.Theyinvokedifferentkindsandlevelsofexplanationfromproximatetoultimate,randomtodeterministic,historicaltoecological,abiotictobiotic.Mostimportantly,manyarenotalternativesinthesensethattheyoffermutuallyexclusiveexplanations.Therearetoomanyformetoreviewandevaluateallofthemhere.Instead,Iofferapersonaloverview:highlightingsomethe-oreticalandempiricaladvances,evaluatingthepresentstateoftheart,andofferingaunifying,butadmittedlyincom-plete,synthesis.IfocusmytreatmentontheLDGofspeciesrichnessacrossspatialscalesfromlocalcommunitiestoregionalbiotas.Patterns:diversityofspecies,cladesandculturesTheLDGispervasive.Itoccursinnearlyallkindsoforgan-plants,animalsandmicrobesandenvironmentsterrestrial,freshwaterandmarine.Itoccursatalllevelsofevolutionarydifferentiation,notonlyatthespecieslevel,butalsoforintraspecicgeneticandphenotypicdifferentiationandforlineagesandhighertaxaofmultiplespecies.ManyexamplesareshowninLomolinoetal.(2010;seealsoWilligetal.,2003;Hawkinsetal.,2012).AlthoughtheLDGispervasive,itisdenitelynotuniver-sal.Therearemanyclearexceptions.Forexample,speciesofconifers,amphipods,craysh,ichneumonidwasps,volesandpenguinsaremostdiverseatmid-orhighlatitudesandgreatlyreducedorabsentinthetropics(e.g.Willigetal.2003;Lomolinoetal.,2010).Theseexceptionsareamongthe‘naturalexperiments’thatofferpotentiallypowerfulinsightsintothemechanismsthatgeneratethemoregeneralProcesses:historicalandecologicalEffortstoexplainboththecasesthatexhibitLDGsandthecasesthatareexceptionshavetypicallyfocusedoneffectsofeither‘history’or‘ecology’.ThehistoricalhypothesessuggestthattheLDGsarelegaciesofpastgeological,climaticandevolutionaryevents,mostofwhichoccurredthousandstomillionsofyearsago.Somehistoricalhypothesesarenon-equilibrialinthesensethattheyproposethatdiversityisstillchanginginalaggedresponsetopastperturbations.AnexampleistheideathattheLDGinatleastsomegroupsinNorthAmericaandEurasiaisalegacyofpastglaciations:thehighlatitudeswereuninhabitableorinhospitableduringtheglacialepochs,andtherehasnotbeensufcienttimeforanimalsandplantstodisperseandadapttothehabitatsthatbecameavailableduringtheinterglacialperiods,includingthecurrentone(e.g.Fischer,1960;Hortaletal.,2011).OtherhistoricalhypothesessuggestthatmanyLDGsreectalongstanding,approximatelysteady-staterelationshipbetweentheabiotictemplateoftheEarthandtheevolution-aryprocessesthathaveshapedbiodiversity.Anexampleistheoutofthetropicshypothesiswhichisdiscussedinmoredetailbelow.Incontrasttohistoricalhypotheses,ecologicalhypothesesassumethat,regardlessoftheirevolutionaryorigin,mostLDGswereoriginallycausedandarenowmaintainedbybio-logicalresponsestotheEarth’sabiotictemplate,especiallyvariablesthatareduetosolarradiation.Thepoles-to-equatorgradientsoftemperatureandseasonalityandthecorrespond-inggradientofbiodiversityhaveexistedforhundredsofmillionsofyears,althoughthedetailsofthepatternshaveuctuatedovertimeinresponsetotectonicevents,Milanko-vichorbitalcyclesandotherfactors.Oneclassofecologicalhypothesesinvokesproductivity:thereisalatitudinalgradientofprimaryproduction,andthemoreproductivetropicssupportmoreindividualsapportionedamongmorespecies.Asecondclassinvokesnicherelationships:adaptationstosomecombinationofabioticconditionsandbioticinteractionsallowtropicalspeciestobemorespecialized,dividingresourcesmorenelyamongmorespecies.THEROLEOFHISTORYOutofthetropicsBothfossilandphylogeneticreconstructionsprovidecompel-lingevidencethatmostlineagesoriginatedinthetropics.Therelevantdynamicsareperhapsmostsimplyandcogentlypresentedinthe‘outofthetropics’model(Jablonskietal.2006;Royetal.,2009;Cavender-Baresetal.,2011;Bowenetal.,2013)andsupportingempiricalevidence,mostlyfromthefossilrecord.Hereismydepiction:(1)ratesofJournalofBiogeography,8–222013TheAuthorsJouralofBiogeographyPublishedbyJohnWiley&SonsLtdWhyaretheresomanyspeciesinthetropics? originationofnewspeciesarehighestinthetropics;(2)higherratesofspeciationthanextinctiongeneratehighdiversityofspeciesandcladeswithinthetropics;(3)mostspeciesandcladesoftropicaloriginremainconnedtolowlatitudes,becauseabioticenvironmentalconstraintsinhibitcolonizationandrangeexpansionoutofthetropics;(4)aminorityoftropicalspeciesovercometheseconstraintsandexpandtheirrangestocolonizeandsometimesdiversifysec-ondarilyathigherlatitudes;and(5)attheselatitudeshighratesofextinctionresultinlowerstandingstocksofspeciesandclades.Accordingtothismodelthetropicsarebotha‘cradle’,becausemostlineagesoriginatethere,anda‘museum’,becausesomeoftheselineagessurviveforlongperiods.Progressivelyhigherlatitudestendtocontainpro-gressivelyyoungerspeciesandclades,becauseofthehigherextinctionrates.Thepredictedpatternsappeartobewellsupportedbythefossilrecord,especiallywhensamplingeffortandothercomplicationsaretakenintoaccount(Jab-etal.,2006).NicheconservatismAcomplementaryhistoricalhypothesisinvokes‘phylogeneticnicheconservatism’(Wiens&Graham,2005;Hawkinsetal.2007,2012;Donoghue,2008;Wiensetal.,2010;Buckleyetal.2010;Cavender-Baresetal.,2011).Myinterpretationisasfol-lows:(1)closelyrelatedspeciestendtosharesimilartraitsinheritedfromtheircommonancestors;(2)amongthesephy-logeneticallyconservativetraitsarenicheattributes,require-mentsandtolerancesforenvironmentalconditions;(3)tropicalenvironmentswithrelativelywarm,aseasonalclimaticregimeshavebeenpresentthroughoutmostofEarth’shistory,whereasmoreextremeconditions,includingperiodsofconti-nentalglaciation,havebeenmoreintermittent;(4)becauseofalongevolutionaryhistoryinrelativelyequableenvironments,mosttropicalspeciesandlineagescannottoleratetheabioticstressesathigherlatitudesespeciallycoldtemperature,lowwater(interrestrialenvironments),andextremeseasonalityandsotheyarerestrictedtothetropics;(5)noveladaptivetraitsarerequiredtotoleratestressfulabioticconditionsandexpandrangestohigherlatitudes;and(6)theincreasingsever-ityofstressactsasalter,resultinginadecreasingnumberofspeciesandlineageswithincreasinglatitude.Thepredictedpatternsappeartobegenerallysupportedbyphylogeneticanalysesoftraitevolution,climaticnichemodels,palaeocli-maticconditionsandgeographicalrangelimitsofnaturalizedexoticspecies.Tropicalspeciesexperienceandareadaptedtoonlyanarrowandequablerangeofabioticconditions(e.g.Janzen,1967;Terborgh,1973;Gaston&Chown,1999;Colwell,2011).So,nicheconservatismoffershistorical,ecologicalandevolutionarymechanismstoexplainwhyonlyaminorityofspeciesandlineageshaveexpandedoutofthetropicstocolo-nizeandsometimesdiversifyinthemorevariableandstressfulenvironmentsathigherlatitudes.Together,theoutofthetropicsandnicheconservatismhypothesesprovideacompellingaccountofthehistoricaldynamicsoftheLDG.Idonotquestionmostpartsofthisexplanation.Inmyjudgement,however,severalissuesstillneedtobeaddressed.Whyareratesofspeciationhighestinthetropics?HowdoesvariationinspeciationandextinctionratesandtheseverityofabioticconditionsacrosslatitudesgenerateandmaintainthestandingstocksofspeciesrichnessseenintheLDGs?Whataretheimplicationsoftheexcep-tions,suchasthediversicationintropicalSouthAmericaoflineagesofplacentalmammalsandculturesofaboriginalhumansthatcolonizedtheNewWorldrelativelyrecentlyviathecold,seasonalenvironmentoftheBeringlandbridge?THEROLEOFECOLOGYItisapparentthattheabove‘historical’hypothesesulti-matelyrely,implicitlyandexplicitly,on‘ecology’.G.E.Hutchinson(1959,p.347)perceptivelyaddressedtheseissuesinhisfamousHomagetoSantaRosalia:IfwecanhaveoneortwospeciesofalargefamilyadaptedtotherigorsofArcticexistence,whycanwenothavemore?Itisreason-abletosupposethatthetotalbiomassmaybeinvolved.Ifthefun-damentalproductivityofanareaislimitedbyashortgrowingseasontosuchadegreethatthetotalbiomassislessthanundermorefavorableconditions,thentherarerspeciesinacommunitymaybesorarethattheydonotexist.Itisalsoprobablethatcer-tainabsolutelimitationsongrowth-formsofplants,suchasthosethatmakethedevelopmentofforestimpossibleaboveacertainlatitude,mayinsoacting,severelylimitthenumberofniches.HereHutchinson,injustafewelegantlywordedsentences,addressestheroleofecologyingeneratingandmaintainingtheLDG:(1)howproductivityultimatelylimitsthetotalbiomassoflivingmatter;(2)howthatbiomassisappor-tionedamongthe‘numberofniches’andhenceamongspe-cies;and(3)howrarespecieswithspecializednichespersistinthefaceofextinction.Forthelastfewdecades,themainecologicalexplanationfortheLDGhasbeenthatregionsofhighproductivityhavehigherbiodiversitybecausemorespeciescanobtainsufcientresourcestomaintainviablepopulations.Thisexplanationisbasedonthewell-documentedpatternthatterrestrialnetprimaryproduction,whichiscontrolledlargelybytempera-tureandseasonality,increasesfromeffectivelyzeroatthepolestoamaximuminthelowlandwettropics.Conse-quently,resourcestosupportorganismsofalltrophiclevelsandmostlifestylesaremostabundantinthetropics,sotheseresourcescanbedividedamongmorespecieswitheachget-ting‘alarge-enoughpieceofthepie’topersistinthefaceofstochasticextinction.Theproductivityhypothesishasbeenadvancedinseveralformsbymultipleinvestigators(e.g.Hutchinson,1959;Connell&Orias,1964;MacArthur,1965,1972;Pianka,1966;Brown,1981;Wright,1983;Currie,1991;O’Brienetal.,1998).TherecanbenodoubtthattheLDGisultimatelydue,atleastinpart,toproductivity.ThepolesanddriestdesertsareJournalofBiogeography,8–222013TheAuthorsJouralofBiogeographyPublishedbyJohnWiley&SonsLtdJ.H.Brown nearlydevoidoflifebecausetheyaresimplytoocoldortoodryfororganismstosurviveandreproducethere.Funda-mentalphysical,chemicalandbiologicalconstraintslimitthecapacitiesoforganismstoconvertenergyandnutrientsintobiomass(e.g.O’Brienetal.,1998).Incontrast,thetropicsteemwithlifebecausethewarmmoistenvironmentoffersrelativelybenignabioticconditionsandabundantresources.Recently,themetabolictheoryofecology(MTE;Brownetal.,2004;Siblyetal.,2012)andempiricalresearchineco-systemecologyhavemadeprogressinsynthesizingourunderstandingofthelinkagesbetweenphysiochemicallimitsonbiologicalmetabolismandthemajorenvironmentalvari-ablesthataffectphotosynthesisandrespiration.Nearlyalloftheenergythatsupportslifecomesfromthesun,andtherateofphotosynthesisornetprimaryproduction(NPP)islimitedprimarilybytemperatureandwaterinterrestrialenvironmentsandbynutrientsandsolarradiationinmarineenvironments.NPPsetsabsolutelimitsontotalresourceuse,biomassandnumberofindividualsinanecosystem,althoughtherearetradeoffs,largelyduetobodysize,inhowenergyandbiomassareapportionedamongindividuals.StrongcorrelationsbetweenspeciesrichnessandNPPinterrestrialenvironmentsledmanyecologists,myselfincluded,toinferthatproductivityaffectsbiodiversityinthewaythatHutchinsonsuggested,bylimitingthenumberofindividualsperspeciesthatcouldpersistinthefaceofextinction(e.g.Brown,1981;Wright,1983;Currie,1991;Wrightetal.1993;Fraser&Currie,1996;Francis&Currie,1998;Kasparietal.,2000;Kilpatricketal.,2006).Now,however,severallinesofevidencesuggestthatthismechanism,byitself,isinadequatetoexplaintheubiquityandmagnitudeofLDGsindifferentorganismsandhabitats.In1992KlausRohde(seealsoFischer,1960)reviewedthehypothesesandsupportingevidencefortheLDG,andpro-posedarelativelynovelmechanism:‘Itisconcludedthatgreaterspeciesdiversityisduetogreater“effective”evolu-tionarytime(evolutionaryspeed)inthetropics,probablyastheresultofshortergenerationtimes,fastermutationrates,andfasterselectionatgreatertemperatures’(Rohde,1992,p.514).Rohdesuggestedthattheeffectoftemperatureonphysiologicalprocessescausesfasterratesofevolutionandmorerapidresponsestoselectioninthetropics.Empiricalstudieshaveprovidedstrongsupportfortheideathatbioticinteractionsare‘moreimportant’inthetropics,andthatthedensity-dependentrelationshipsplayamajorroleingeneratingandmaintainingtheLDG(e.g.Janzen,1970;Connell,1971;MacArthur,1972;Sax,2001;Mittelbachetal.,2007;Schemskeetal.,2009;Comitaetal.,2010;Ricklefs,2010;Swamy&Terborgh,2010;Johnsonetal.2012;Terborgh,2012).Butitisnotclearhowtemperatureorotherabioticfactorsthatvarywithlatitudeaffecttheratesandoutcomesofinteractionssothat‘diversitybegetsTheMTEhasmodelledexplicitlyandquantitativelythelinkagesbetweentemperature,metabolicbiochemistry,andphysiological,ecologicalandevolutionaryrates(e.g.Allenetal.,2002,2006;Brownetal.,2004;Savageetal.,2004;etal.,2005;Gillooly&Allen,2007;Siblyetal.2012).Thetheoryassumesthatoverabiologicallyrealisticrangeoftemperatures,biologicalratesincreaseexponentiallywithtemperature:,whereistherateofsomeprocesssuchasmetabolism,populationgrowthorspecia-tion,eistherootofthenaturallogarithm,isan‘activa-tionenergy’thatgivesthetemperaturedependence,Boltzmann’sconstant(8.62),andistem-peratureinKelvin(Gilloolyetal.,2001;Brownetal.,2004;etal.,2012).Thetheoryisstillincomplete,butitpro-videsaframeworkforanalysingtemperaturedependenceofphenomenarelatedtobiodiversity.Ishowanexample,usingpreliminarydataonforestsfromourNationalScienceFoun-dation(NSF)-supportedMacrosystemsresearchproject(Table1).TheupperpanelofFig.1showsproductivityasafunctionofaverageannualtemperatureinan‘Arrheniusplot’,wherethelogarithmofNPPisplottedasafunctionofinversetemperature,1/.Therelationshipishighlysigni-cant,with0.46,whichisintermediatebetweenempiricalvaluesforratesofphotosynthesisandsecondarysuccession0.33)andrespiration(0.65;Allenetal.,2005;etal.,2006).ThelowerpanelofFig.1showsthetemperaturedependenceoftreeabundanceandspeciesrich-ness.Bothrelationshipsaresignicant,buttherelationshipismuchstrongerforspeciesthanforindividuals(and0.16,respectively).Similarrelationshipsforspeciesrich-nesshavebeenpublishedelsewhere(e.g.Allenetal.,2002,2006;Allen&Gillooly,2006;Wangetal.,2009).Theimportantpointtobemadehere,however,isthatspeciesdiversityincreaseswithincreasingtemperaturemuchmorerapidlythanthenumberofindividualsandNPP.ThisimpliesthatproductivityaloneisnotasufcientexplanationfortheLDG.Instead,itsuggeststhattheLDGisdueinlargeparttorelativelydirecteffectsofkinetics.Thehighertemper-aturesinthetropicscausehigherratesofmetabolism,eco-logicaldynamicsandcoevolutionaryprocesses,whichgenerateandmaintainhigherbiodiversity.IofferthesyntheticframeworkoutlinedinFig.2andeshedoutbelow.Theowofenergythroughanecosystemdeterminesitscapacitytosupportlife.Netprimaryproduction(NPP)givesthewhole-ecosystemrateofenergysupply.NPPsetspower-fulconstraintsontotalabundanceandbiomassofallorgan-isms.GivenaxedNPP,anecosystemcansupporteitherahighbiomasscomposedofarelativelyfewlargeindividualsoralowbiomasscomposedofmanysmallindividuals(e.g.JournalofBiogeography,8–222013TheAuthorsJouralofBiogeographyPublishedbyJohnWiley&SonsLtdWhyaretheresomanyspeciesinthetropics? treesinaforestcomparedtoplanktonicalgaeintheocean).Therateofphotosynthesisincreaseswithincreasingtempera-tureinbothterrestrialplantsandaquaticalgae(Allenetal.2005;Yvon-Durocheretal.,2010;Anderson-Teixeira&Vito-usek,2012).However,geographicalvariationinNPPiscomplicatedbywaterlimitationinterrestrialecosystems(e.g.Lieth,1975;O’Brienetal.,1998)andnutrientlimitationinmarineandfreshwaterecosystems(e.g.Bunt,1975;Smith,1979;Mooreetal.,2013).Theeffectofproductivityondiversityispositive,butmodest:toosmalltoaccountforthemagnitudeoftheLDG.Thenumberofspeciesincreasesmuchmorerapidlywithdecreasinglatitudeandincreasingtemperaturethantheincreaseintotalecosystemenergysupply(NPP),biomassorabundance(Fig.1;seealsoEnquist&Niklas,2001;Currieetal.,2004).Theeffectofproductivityonbiodiversityisaddressedbyspeciesenergytheoryanditsextensionsandempiricaltests(Wright,1983;Wrightetal.,1993;Currieetal.,2004;Hurlbert,2006).Speciesenergyrelationshipsareconceptuallysimilartospeciesarearelationshipsinislandbiogeography.Theempiricalpatternsarecomparabletootherspeciesenergyandspeciesarearelationships(MacAr-thur&Wilson,1967;Wright,1983;Wrightetal.,1993;Cur-etal.,2004;Hurlbert,2006),whereatypical-valueof0.25wouldrequirefourordersofmagnitudeincreaseinNPPtoincreasespeciesrichnessbyoneorderofmagnitude(i.e.a10,000-foldincreaseinproductivityfora10-foldincreaseindiversity).IconcludethatthereisalatitudinalgradientofproductivityduetothetemperaturedependenceofNPP(seebelow),anditplaysasmallbutsignicantroleintheLDG.TheLDGinalphadiversitydepends,therefore,notsomuchontherateofresourcesupplyasonhowtheseresourcesareapportionedamongindividualsandspecies.Indeed,ithaslongbeenknownthattropicalcommunitiestypicallycontaindisproportionatelymorespecializedandrarespeciesthanassemblagesathigherlatitudes(Klopfer&MacArthur,1961;Hubbell,1979,2008).Thisisperhapsbestdepictedbycom-paringtheshapesofrankedspecies-abundancedistributions(SADs):whenthenumberofindividualsisscaledlogarithmi-cally,highlatitudecommunitiesaretypicallysteepandapproximatelylinear,whereastropicalcommunitiesaremuchatterandstronglycurvilinear(Fig.2,top).ThetemperaturedependenceofalphadiversityandtheshapeofSADsimplythatlocalspeciesrichnessisregulatedbytemperature-dependentecologicalandevolutionarypro-cesses,mediatedthroughthekineticsofmetabolismasRo-hdesuggestedin1992.Thekineticsofecologicalinteractionsandcoevolutionaryadaptationresultinspeciesoccupyingmorespecializednichesinthetropics.Localspeciesrichnessisindeedstronglytemperature-dependent.Arrheniusplotsofdiversityasafunctionofinversetemperaturetypicallyhavenegativeslopes,with0.8(e.g.Allenetal.,2002,2006;Wangetal.,2009;Fig.1).Metabolictheoryhasthepotentialtomodelandaccountfortheserelationships.Therehavebeenattemptstodoso(e.g.Allenetal.,2002;:Brownetal.,2004;Allen&Gillooly,2006;Wangetal.,2009),butIagreewithDavidStorch(2012)that‘ametabolictheoryofbiodiversityisaworkinprogress’.Theobviousquestionsare:whatarethetempera-ture-dependentprocesses,andhowdotheygenerateandmaintainstandingstocksoflocalspeciesrichness?IbelievethattheanswerliesprimarilyintheeffectoftemperatureoninterspecicecologicalinteractionsandRedQueencoevolu-tion.‘Diversitybegetsdiversity’togeneratetheLDG,becauseecologicalandevolutionaryratesincreaseexponentiallywithtemperatureandarehighestinthetropics.Butweneedthe-oreticalandsimulationmodelsandrelevantempiricalstudiestoshowhowthisoccurs.TogofromtypicalloglineartemperatetocurvilineartropicalSADsrequiresspecies-specicdensity-dependentprocessesthat(1)suppressbroad-nicheddominantspecies,freeingupresources,and(2)facilitatethepersistenceofspe-cializedspecies,allowingthemtoincreasewhenrare,colo-nizefromthemetacommunity,andspeciate.OnephenomenonthathasthiseffectistheJanzenConnellpro-cess(Janzen,1970;Connell,1971).Theideaisthat‘enemies’,somecombinationofpredators,herbivores,parasitesanddiseases,havenegativedensity-dependentimpactsontheir Table1Preliminarydataformeanannualtemperature,annualnetprimaryproductivity(NPP),numberofindividualtreesandnumberoftreespeciesatsixforestsitesofourMacrosystemsproject(veLong-TermEcologicalResearchsitesandtheSmithsonian’sBarroColoradoIsland).NPPdataarefromKasparietal.(2000)exceptforLuquillo,whichisfromhttp://daac.ornl.gov/NPP/guides/npp_doc.html.Datafornumberofindividualtrees0.25cmd.b.h.andnumberoftreespeciesinstandardizedmodied‘Gentryplots’totalling5000minareawerecollectedbyV.Buzzard,C.Sides,A.HendersonandB.J.Enquist.SiteLatitudeLongitudeTemperature((gC/mNumberofofspeciesNiwotRidge(Colorado,USA)40N105418020316H.J.AndrewsForest(Oregon,USA)44N122W9200232217HarvardForest(Massachusetts,USA)42N72W7450138425CoweetaForest(Georgia,USA)35N83W13550173052LuquilloForest(PuertoRico)18N65W231033212896BarroColoradoIsland(Panama)9N79W2713104971263JournalofBiogeography,8–222013TheAuthorsJouralofBiogeographyPublishedbyJohnWiley&SonsLtdJ.H.Brown preyorhoststhroughbothshort-termecologicalinteractionsandlonger-termcoevolutionaryadaptations.TheJanzenConnellprocesshasincreasinglygainedempiricalsupport(e.g.MacArthur,1972;Clark&Clark,1984;Sax,2001;Mit-etal.,2007;Schemskeetal.,2009;Comitaetal.2010;Ricklefs,2010;Swamy&Terborgh,2010;Johnsonetal.,2012;Terborgh,2012).Itisfarfromclear,however,howtemperature,byincreas-ingtherateofinteractionsandcoevolution,generatesandmaintainshigherspeciesrichness.Recentmeta-analyseshavecollectedandanalysedpublisheddataontemperaturedepen-denceofratesofecologicalinteractions,includingcompeti-tion,predation,herbivoryandparasitism(Brownetal.2004;Delletal.,2011;Englundetal.,2011).Theresultsarevariable,samplesizesaregenerallysmall,andstatisticalreso-lutioniscorrespondinglylimited.Amorecomprehensivecompilationandanalysisrevealsaclearcentraltendency,withvaluesclusteringaroundthevalueof0.65aspre-dictedbymetabolictheory(W.R.Burnside,S.T.Hammond&J.H.Brown,unpublished).Additionaltheoreticalandempiricalstudiesareneededtoshowhowwarmertempera-tures,byspeedingupratesofinteractionandcoevolution,causehigherspeciesdiversityinthetropics.SpatialrelationshipsThevastmajorityofempiricalstudiesoftheLDGhavefocusedonalphadiversityorlocalspeciesrichness.Muchlessattentionhasbeendevotedtothespatialcontextofdiversity,explicitlytobetadiversityorspatialturnoverinspeciescomposition.Higherbetadiversityinthetropics,however,appearstobeanotherpervasivefeatureoftheLDGetal.,2002;Qianetal.,2004,2009;Rodrguez&Arita,2004;Qian&Ricklefs,2007;Wangetal.,2009;Kraftetal.,2011).Thisphenomenonisexpressedonarangeofspatialscales.Withdecreasinglatitudeandincreasingtem-perature:(1)similarityinspeciescompositionoflocalcom-munitiesdecreasesmorerapidlywithincreasingdistancebetweensamples(distance-decayrelationships:e.g.Nekola&White,1999);(2)thenumberofspeciesincreasesmorerap-idlywithincreasingsamplearea(speciesarearelationships:Fig.2,middle;e.g.Wangetal.,2009);and(3)speciesoccupysmallergeographicalrangesandanarrowerrangeofabioticenvironmentalconditions(Rapoport’srule:e.g.Rapo-port,1982;Stevens,1989;Brownetal.,1996).Thesepatternsofbetadiversityareconsistentwiththeeffectoftemperatureonecologicalinteractionsandcoevolu-tionaryprocessesasoutlinedabove.Inaspatialcontext,‘diversitybegetsdiversity’becausebioticresistancefromotherspeciesrestrictsspeciestotheirspecializednichesandlimitstheircapacitiestodisperseandinvadeotherenviron-ments,whichalsohaveadiversecomplementofwell-adaptedspecies.Thepatternsofbetadiversityareuniquelyexplainedbybioticlimitingfactors;otherwisetherelativelybenignconditionsinthetropicsshouldallowspeciestobewidespread(e.g.Terborgh,1973;Gaston&Chown,1999;Colwell,2011).Again,itremainstobeexplainedhowtemperaturegener-atesandmaintainsthesepatternsofbetadiversity.Criticalhereistheeffectoftemperatureondispersal.Isuggestthatingeneral,temperaturetendstoincreasetheratesbutdecreasethedistancesofdispersal.Ontheonehand,thelargenumberofveryrarespeciesintropicalcommunitiessuggeststhatrelativelyhighratesofcolonizationofspeciesfromsomelarger,moreextensivemetacommunityareessen-tialtoreplenishlocalrichnessafterstochasticextinctions.Ontheotherhand,thesmallrangesanddifcultyofinvadingdistantsitesduetobioticresistancesuggestthatlong-dis-tancedispersalisrareandprobablymaladaptiveformosttropicalspecies.Metabolictheoryaddressesthetemperaturedependenceofmetabolicrate,lifehistorytraitsanddispersal Figure1Temperaturedependenceofproductivity,numberoftreesandnumberoftreespeciesonthesixforeststudysitesofourMacrosystemsproject(Table1).DataarepresentedasArrheniusplots,withthenaturallogarithmofrateasafunctionofinversetemperature,1/(socoldertemperaturesaretotheright).Datawerettedbyordinaryleastsquaresregression:valuesofgivetheslopes,themeasureoftemperaturedependence,andvaluesofgivetheproportionofvariationexplained.Above:rateofnetprimaryproduction(NPP);below:numberofindividualtreesandnumberoftreespecies:datafromV.Buzzard,C.Sides,A.HendersonandB.J.Enquist.NotethatthetemperaturedependenceforspeciesrichnesswassubstantiallyhigherthanforNPP,whichwassubstantiallyhigherthanforthenumberofindividuals.JournalofBiogeography,8–222013TheAuthorsJouralofBiogeographyPublishedbyJohnWiley&SonsLtdWhyaretheresomanyspeciesinthetropics? distanceinectothermicorganisms.Theeffectoftemperatureondevelopmentandlifespanpredictsshorterdispersaldis-tancesinatleastsometropicalorganisms.Oneexampleisplanktoniclarvae,theprimarydispersalstageofbenthicmar-inealgae,invertebratesandsh.Inwarmerenvironmentslarvaedevelopmorerapidlyandconsequentlydisperseovershorterdistances(O’Connoretal.,2007).Andnally,higherratesofecologicalinteractionswithenemiesandtheresultingmortalityshouldadditionallylimitdispersaldistancesandestablishmentofpropagulesinthetropics.SpeciesdynamicsUltimatelythegenerationandmaintenanceofdiversitydependsonspeciesdynamics,onhowratesofcolonization,speciationandextinctionvarywiththenumberofspecies.Whenspeciationrateexceedsextinctionrate,thenumberofspecieswilltendtoincreaseexponentially;whenextinctionrateexceedsspeciationrate,thenumberofspecieswilltendtodeclineexponentiallytowardszero.Becausesuchexponen-tialtrajectoriescannotbecontinuedforverylong,speciesdiversitymustcometoanapproximateequilibriumorsteadystatebetweenratesoforiginationandextinction.So,itisreallythis‘carryingcapacityforspecies’,ratherthanspecia-tionandextinctionratesperse,thatmaintainsspeciesdiver-sityoverbothecologicalandevolutionarytime.Itisconstructivetodevelopasimplegraphicalmodel,sim-ilartoMacArthur&Wilson’s(1967)equilibriummodelofislandbiogeography(Fig.2,bottom).Here,bothspeciationandextinctionratesaredepictedasincreasingfunctionsofspeciesrichness,buttheextinctioncurveissteeperthanthespeciationcurvesothattheycrosstogiveastableequilib-rium.Ihavedrawnthegraphwithasingleextinctioncurveforbothtemperateandtropicalenvironments,butwithahigherrateofspeciationforanygivenrichness,,inthetropi-calenvironment,reectinghow‘diversitybegetsdiversity’duetotheratesofbioticinteractionsandcoevolutionaryprocesses(seealsoEmerson&Kolm,2005).Thismodelpredictsahigher‘carryingcapacity’forspecies,,inthetropics,main-tainedbyahigherturnoverrateofspecies,,i.e.higherratesofbothspeciationandextinctionatthisequilibrium.Thereismuchroomtoelaborateandimproveonthis ProducNiche relaonsal relaonssunlightwaternutrientsnet primary producon (NPP) heterotrophic productotal energy, abundance, compeherbivoryparasiphysical stressdisturbanceregional community(beta diversity)local community(alpha diversity)geographic gradient(gamma diversity)“zero sum” ecosystem capacityfecunditymovementestablishment:propagulesurvivalon growthgeographic rangecolonizadivergencecoevoluon sizeonsc interac Figure2Atentativesyntheticframeworkforthecausalmechanismsthatgenerateandmaintainthelatitudinalgradientofspeciesdiversity.Causeeffectrelationshipsareindicatedbyarrows.Mechanismsthatareatleastinparttemperaturedependent,andhenceconsistentwiththekineticsofmetabolictheory,areinred.Othermechanismsareingreen.Someemergentoutcomesareshowninthegraphs.Thepresentationisarrangedinapproximateorderofincreasingspatialscaleandevolutionarytime,startingwithlocalecologicalprocessesatthetopandendingwithregionalspeciesdynamicsatthebottom.Butthemechanismsoperatebothbottom-upandtop-down,asindicatedbythedouble-headedverticalarrowsontheright,whichindicateimportantfeedbacksamongprocessesandacrossscales.Seetextforadditionalexplanation.,carryingcapacityforspecies;turnoverrateofJournalofBiogeography,8–222013TheAuthorsJouralofBiogeographyPublishedbyJohnWiley&SonsLtdJ.H.Brown model:e.g.changingtheshapesandpositionsofthecurves,makingtheextinctionratesdifferentinthetropicandtem-perateenvironments,andaddingacolonizationratecurvetodepictspeciesmigratingoutofthetropicsintothetemperatezone,andsoon.Nevertheless,thissimplemodelcanbeuse-fulinguidingourthinkingaboutthefundamentalsofspe-ciesdynamicsandthetemperaturedependenceoftheunderlyingrateprocesses.ThemodelinFig.2isatleastqualitativelyconsistentwithmetabolictheoryandempiricalobservations.Thereisincreasingevidencethatratesofevolution,fromnucleotideandnucleicacidsubstitutiontospeciationandphyleticdiversication,arehigherinwarmerenvironmentsand,whenexamined,exhibitalatitudinalgradient(Gilloolyetal.2005;Allenetal.,2006;Estabrooketal.,2007;Gillooly&Allen,2007;Mittelbachetal.,2007;Gillmanetal.,2010;etal.,2010,2011;Machacetal.,2012).Inparticular,themodelinthebottompanelofFig.2predictsthatatequilibriumbothspeciationandextinctionratesaretemper-aturedependent.Thisissupportedbyturnoveroffossilmorphospeciesofplanktonicorganismsasafunctionofpal-aeolatitudeandsea-surfacetemperatures(Allenetal.,2006;Allen&Gillooly,2006).Itremainstobeexplainedjusthowhigherratesofevolu-tionanddiversicationgenerateandmaintainhigherspe-ciesdiversityinwarmerenvironmentsandattropicallatitudes.ThemodelofAllenetal.(2002)thatpredictedauniversaltemperaturedependenceofbiodiversityhasnotbeensupported.Asinglevalueofisunlikelyontheoreti-calgrounds(Storch,2012),andempiricalstudieshaveshownthatvarieswithspatialscale(betadiversity:seeabove),taxonandgeographicalregion(Wangetal.,2009;etal.,2007).ItistemptingtosuggestthatthetemperaturedependenceofbiodiversityandtheLDGarenotsimplyduetohigherratesofevolutioninwarmerenvi-ronments,buttohigherratesofRedQueencoevolutionduetomoreandfasterbioticinteractions.Giventheevi-dencepresentedabove,Iamcondentthat‘theRedQueenrunsfasterwhensheishot’,butadditionaltheoreticalandempiricalworkwillberequiredtoelucidatethemecha-Ihavenarratedthesynthesisabovefromthebottom-up,startingwithhowproductivityandnicherelationshipsaffectlocal(alpha)diversityandworkinguptospatialturnover(betadiversity)andnallytogeographicalscalespeciesdynamics(gammadiversity).Idonotmeantoimply,how-ever,thatthecausalrelationshipsallowinthisdirection.Infact,asdepictedbythearrowsontherightsideofFig.2,therearetop-downfeedbacksatalltheselevels.Thelarge-scalepatternsofcolonization,speciationandextinctionfeeddowntoinuenceregionaldiversityandthecompositionofmetacommunities,andthesefeeddowntoaffectthediversityandcompositionoflocalcommunities.Someofthesecross-scalefeedbacksareimplicitinthepresentationaboveandFig.2,butafewwarrantsomeelaboration.One,describedbyDarwin(1859),hasbeencalledtheDo-MacArthurphenomenon(DMP):theequatoriallimitsofspeciesgeographicalrangesareusuallyduetobio-ticinteractions,whereasthepolarlimitsareduetostressfulabioticconditions.ThereisconsiderableempiricalsupportfortheDMP(e.g.MacArthur,1972;Root,1988;Stephens&Wiens,2003;Lomolinoetal.,2010;Sundayetal.,2011,2012;andothercasessummarizedinLomolinoetal.,2010).Perhapsthemostconvincingevidencecomesfromthelati-tudinallimitsofgeographicalrangesofexoticspeciesoncontinentsandislandswheretheyarenativeandwheretheyhavebeenintroduced(e.g.Sax,2001;Wiens&Graham,2005).TheDMPoffersawaythatthehistorical‘outofthetropics’dynamicsandphylogeneticnicheconservatismoflineagescanbereconciledwiththefactthattheLDGisanancientpattern,reectingasteady-staterelationshiptocli-matedatingbackhundredsofmillionsofyears(e.g.Stehlietal.,1969;Crane&Lidgard,1989;Crame,2001).Needlesstosay,theDMPisconsistentwiththeabovesuggestionsthattheLDGisdueinlargeparttothetemperaturedepen-denceofbioticinteractionsandRedQueencoevolution.Anotherinterestingcross-scalefeedbackistheinuenceofthegeographyofspeciationonlocalandregionaldiversity.Ithaslongbeenrecognizedthatdiversityishighestnotjustinthetropics,butintopographicallydiverseregions:mountainterrainonland,suchastheslopesoftheAndesandHimala-yas,andislandarchipelagosinthesea,suchastheIndo-WestPacicandCaribbean.Janzen(1967)calledattentiontotheformerinawonderfulpaperentitled‘Whymountainpassesarehigherinthetropics’.Hepointedoutthatbecauseofsea-sonalvariationinclimate,agivenchangeinelevationposesagreaterbarriertodispersalinthetropicsthanathigherlati-tudes(Fig.3,top).Bydispersinginaparticularseason(winterorsummer)anelevationallyrestrictedtemperatespeciescanpotentiallycrossmountainpassesatlowerorhigherelevationswithoutencounteringtemperaturesoutsidetherangethatitnormallyexperiencesduringanannualcycle.Anelevationallyrestrictedtropicalspecies,bycontrast,cannotcrossoverpassesatsubstantiallyhigherorlowerelevationswithoutbeingexposedtomoreextremetemperaturesthanitnormallyexpe-riences.Janzensuggestedthatadaptationstosuchlimitedsea-sonalclimaticvariationresultintropicalorganismswithnarrowthermalnichesandrestricteddistributions.McCain(2009)hasrecentlyshownthatelevationalrangesarenarrowerinthetropicsthanathigherlatitudes,apatternconsistentwithbothJanzen’shypothesisandRapoport’srule(seeabove).AlthoughJanzenwascautiousaboutextrapolatinghisideatoexplainhigherspeciationratesandspeciesdiversityinthetro-pics,itiseasytoseehowtemperature-limiteddispersalwouldhavethiseffect.Isuggestthatasomewhatsimilarphenomenonmayfacili-tatespeciationinmarineorganismsintropicalarchipelagos(Fig.3,bottom).Speciesrichnessofbenthicmarineorgan-ismsishighestintheislandsoftheIndo-WestPacicandJournalofBiogeography,8–222013TheAuthorsJouralofBiogeographyPublishedbyJohnWiley&SonsLtdWhyaretheresomanyspeciesinthetropics?