The University of Chicago Carnivory in the Bromeliad Brocchinia reducta, with a Cost/Benefit Model for the General Restriction of Carnivorous Plants to Sunny, Moist, Nutrient-Poor Habitats Author(s): Thomas J. Givnish, Elizabeth L. Burkhardt, Ruth E. Happel and Jason D. Weintraub Source: The American Naturalist, Vol. 124, No. 4 (Oct., 1984), pp. 479-497 Published by: The University of Chicago Press for The American Society of Naturalists Stable URL: http://www.jstor.org/stable/2461590 . Accessed: 30/05/2013 13:53 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. . The University of Chicago Press, The American Society of Naturalists, The University of Chicago are collaborating with JSTOR to digitize, preserve and extend access to The American Naturalist. http://www.jstor.org This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions The AmericanNaturalist Vol. 124, No. 4 October 1984 CARNIVORY IN THE BROMELIAD BROCCHINIA REDUCTA, WITH A COST/BENEFIT MODEL FOR THE GENERAL RESTRICTION OF CARNIVOROUS PLANTS TO SUNNY, MOIST, NUTRIENT-POOR HABITATS THOMAS J. GIVNISH,* ELIZABETH L. BURKHARDT,* RUTH AND JASON D. WEINTRAUBt E. HAPPELt *Departmentof Organismicand EvolutionaryBiology, Harvard University,Cambridge, Massachusetts 02138; tDepartmentof Anthropology,Harvard University,Cambridge, Massachusetts 02138; tSection of Ecology and Systematics,Cornell University,Ithaca, New York 14850 SubmittedSeptember15, 1983; Accepted March 22, 1984 MembersoftheplantfamilyBromeliaceaeoccupya widerangeofhabitatsin the New World(Pittendrigh 1948;Tomlinson1969;Gilmartin 1973;Smithand on trees and shrubs, Downs 1974).Manyofthese,particularly epiphytic perches have substratesthatare exceedinglypoor in mineralnutrients (Benzingand Renfrow1971,1974;Benzing1973,1980).Perhapsbecauseofthis,and because leafbases manybromeliadsimpoundwaterin "tanks" amongtheiroverlapping have intowhichinsectsorotheranimalscan blunderanddrown,severalbotanists suggestedthatsomebromeliadsmaybe carnivorous. Here we describethe firstconfirmed case of carnivory amongbromeliadsin Brocchiniareducta,a terrestrial speciesnativetomoistsandsavannasandbogsin of the carnivorous theGuayanaHighlands.We discussthepossiblederivation inplants.A ofcarnivory habitwithin andprovidea rigorous definition Brocchinia, cost/benefit modelforthe evolutionof carnivory is presentedto analyzewhy sitesandis is restricted carnivory mainlyto plantsofsunny,moist,nutrient-poor rarein epiphytes and otherbromeliads. We beginbyconsidering previousclaims in theBromeliaceae. fortheexistenceofcarnivory HISTORICAL BACKGROUND Picado(1913)proposedthatmanytankbromeliads secreteproteolytic enzymes in theirleafrosettes.Howto digestanimalsthatperishin thefluidimpounded andthe ever,suchsecretionhas neverbeenconfirmed usingmoderntechniques, proteasesfoundmayresultsimplyfrombacterialactivity(Benzing1980).Rees and Roe (1980)suggestedthatthegiantAndeanPuya raimondii maybe carnivorous because smallbirdsthatnest amongits closelypacked leaves are often impaledon the leaf spines!It seems morelikely,however,thatPuya's spines Am.Nat. 1984.Vol. 124,pp. 479-497. ? 1984byThe University ofChicago.0003-0147/84/2404-0004$02.00. Allrights reserved. 479 This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions 480 THE AMERICANNATURALIST by Andean evolvedto deterconsumption of its singleterminalinflorescence bears,whichdestroythe flowerspikesof 90% of some populationsof closely or to deterleaf relatedPuya species(Wurdack1964,personalcommunication), (see Janzen consumption by elementsoftheextinctSouthAmericanmegafauna and Martin1982).McWilliams(1974)suggestedthatthenarrowtanksofcertain Brocchiniaand Catopsisspeciesrecalltheformofmanypitcherplantsand may Benzing(1980)hasechoedthisargument thusrepresent adaptations forcarnivory; morerecently.These authors,however,do not provideessentialdata on the capture,and absorption. actualpresenceof preyin tanksand theirattraction, has steeplyinclined Finally,Fish (1976) has shownthatCatopsisberteroniana wax thatinhibits leavescoatedwitha slippery theescapeofinsectsfromitstank. and reflective wax actually It is possible,however,thatthe leaf inclination and transpiration in itstreetop represent adaptations to reducelightinterception and inconsequential. habitat,witheffectson insectsbeingcoincidental demonOne conclusionto be drawnfromthesestudiesis thatanyconvincing of strationof carnivoryin bromeliadsrequiresa clear and rigorousdefinition in spiteofthegreatinterest accordedcarnivorous plants carnivory. Surprisingly, sincethetimeofDarwin(Darwin1875;Hepburnetal. 1920;Lloyd1942;Plummer and Kethley1964;Sorensonand Jackson1968;Chandlerand Anderson1976a, 1976b;Heslop-Harrison 1976,1978;Erickson1978;Greenet al. 1979;Slack 1979; Benzing1980;Dixon et al. 1980;Thompson1981;Luttge1983),thereis no such in theliterature. definition Definitionof Carnivory to be classifiedas We propose thata plantmustfulfilltwo requirements fromdead animalsjuxcarnivorous.First,it mustbe able to absorbnutrients obtainsomeincrement to fitness in termsof taposedto itssurfaces,and thereby pollenproduction, or seed set. Second,the increasedgrowth, chanceofsurvival, plantmusthave someunequivocaladaptationor resourceallocationwhosepridigestion ofprey. maryresultis theactiveattraction, capture,and/or frompurelydefensive The firstprovisois neededto differentiate carnivory or killpotentialanimalenemieswithoutleadingto adaptationsthatimmobilize substantialnutrient absorptionor consequentincreasesin plantsurvivaland reproduction (Lloyd1942;Benzing1980).The secondprovisois requiredbecause fromdead animals somenutrients manyplantscan passivelyprofit by absorbing inthesoilor on leafsurfaces(Benz(e.g., nematodes, earthworms) decomposing plantsshowsthat ing1980).A surveyofthe16generaofrecognizedcarnivorous forall threeprocessesofactivepreyattraction, capture,anddigestion adaptations on logicalor historical are notrequiredto qualifya plantas carnivorous grounds (table 1). For example, bladderworts(Biovularia, Polypompholyx,Utricularia) lack attractants and butterworts forprey(Lloyd (Pinguicula)oftenapparently 1942; Meyersand Strickler1979; Slack 1979). Pitcherplantsin the families Sarraceniaceae,Nepenthaceae,and Cephalotaccaceaeentrappreypassivelyin for thatdo, however,possessobviousmorphological specializations waterpitfalls thisfunction. Heliamphora)lack Finally,somepitcherplants(e.g.,Darlingtonia, This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTIONOF CARNIVOROUS PLANTS 481 TABLE 1 PRESENCE (+) OR ABSENCE (-) OF ADAPTATIONS FOR ACTIVE PREY ATTRACTION, CAPTURE, AND DIGESTION IN GENERA OF CARNIVOROUS PLANTS (see textfor sources) Prey Attraction* Genera (no. sp.): Prey Capture Prey Digestion Brocchinia (1) ........... Heliamphora (6) ......... Darlingtonia (1) ......... Sarracenia (9) ........... Cephalotus (1)........... Nepenthes (71) .......... + + + + + + passive passive passive passive passive passive Genlisea (35) ............ - passive + lobsterpot + + + + + + + + + + flypaper flypaper flypaper flypaper flypaper + + + + + + + + + + spring-trap spring-trap spring-trap spring-trap spring-trap Drosera (90)............. Drosophyllum(1) ........ Byblis (2) ............... Pinguicula(35) .......... (1) ....... Triphyophyllum Dionaea (1) ............. Aldrovanda (1) .......... Utricularia (280)......... Biovularia (1) ............ Polypompholyx (2) ....... + - + - + - +1+ + pitfall pitfall pitfall pitfall pitfall pitfall * A conservativeapproach has been taken in evaluatingattractantsthatmay be merelyincidental aspects of othertraits(e.g., glisteningdigestiveglands). digestiveglandsand apparently relyon bacteriaand otherorganisms to break downprey(Lloyd 1942;Adamsand Smith1977;Slack 1979). Thus,specializeddigestive glandsarenota sinequa nonforcarnivory, although theexcellentresearchof Heslop-Harrison and hercolleagues(Heslop-Harrison 1975, 1976, 1978; Green et al. 1979; Heslop-Harrisonand Heslop-Harrison1980) has demonstrated thepresenceofcharacteristic cells in mostcarnivosecretary rousgenera.Plantscapableofabsorbing nutrients fromdead animals,butwhich lackactivemeansofpreyattraction andpreydigestion, andpossessneither motile resultis imtrapsnorpassive structures likeone-waypassageswhoseprimary and mobilization ofanimalsnearplantsurfaces,mustbe consideredsaprophytes notcarnivorous plants.Manytankbromeliads probablyfallintothislattercategoryand mustbe studiedwithcare beforea givenspecies is claimedto be as has been emphasizedby Benzing(1980). carnivorous, CARNIVORY IN Brocchinia reducta In January Venezuelaas 1980,GivnishvisitedtheGranSabanaofsoutheastern partof an expeditionsponsoredby the Universidadde los Andes. The Gran Sabana is partoftheGuayanaHighlandsandis a massiveplateauroughly 1200to 1800minelevation(MayrandPhelps1967;Maguire1970;Steyermark 1982).It is underlainby bleachedsandstoneof theRoraimaFormation, and studdedwith numerousflat-topped mountainscalled tepuis,whose cliff-lined marginsreach This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions 482 THE AMERICAN NATURALIST 2000to 2800minheight.A sterile,highly acidicsandysoildevelopsdirectly over rotting sandstone,and supportsa vegetational mosaicdominatedby savannas, galleryforests,xeromorphic scrub,and bogs (Steyermark 1961,1966;BrewerCarias 1976). Annual precipitation generallyexceeds 2500 mm (Atlas de Venezuela 1969).Characteristics of thevegetation includea highproportion of woodyplantswithnarrow,leathery leaves; highincidenceofterrestrial orchids; and a richarrayof narrowly distributed speciesand unusualplantgroups(e.g., Rapateaceae,Eriocaulonaceae,Podostemonaceae,Velloziaceae) nativeto the GuayanaHighlands(Maguire1970;Steyermark 1961,1966,1976).Thereis an abundanceofterrestrial bromeliads intheprimitive and subfamily Pitcairnioideae, a highincidenceof recognizedcarnivorous plants.The latterincludetheSouth Americanpitcherplant(Heliamphora)and numeroussundews(Drosera),bladderworts(Utricularia),and lobsterpot plants(Genlisea)(Maguire1970;Steyermark1961,1966,1976). BrocchiniareductaBaker is a commonterrestrial bromeliadin the Gran Sabana, and is a conspicuouselementof manysavannasor bogs on wet sand above 1200 m. It is widelydistributed on tepuisin southernVenezuela and adjacentGuyana,and belongsto a pitcairnioid genusof 18 ecologicallydiverse to theGuayanaHighlands(Smithand Downs 1974).Plantsof speciesrestricted Brocchiniareductawerestudiedand collectedforanatomicaland physiological researchinJanuary of 1980,1981,and 1983,on a dampsandsavannanearkm148 of the El Dorado-Santa Elena Road in Estado Bolivar (N 5o48', W 61?25'). This specieshas severaltraitsthatsuggestit is a carnivorous plant. First,theleaves are bright yellow-green and heldnearlyvertically, so thatthe leafrosetteformsa conspicuousyellowish cylinder thatis taller(x = 32.8 + 14.5 herbaceousvegetation, andrecallstheaspectof cm,n = 20) thanthesurrounding in bogsand pocosinsofthesoutheastern Sarraceniaflavagrowing UnitedStates (fig.1). Second,theinnersurfaceofeach leafis coatedwitha finewaxypowder thatreadilyexfoliates(fig.2), servingto lubricatethe verticalsurfacesand increasing thedifficulty of escape fromthecentraltank.Camponotusantswere in a glassjar duringthe placednearthebottomofexcisedleaves resting upright Januaryfieldtrip.Of nine observedattemptsto ascend the innerwaxy leaf the surface,eightresultedinfailure,eventhoughtheantswereable to negotiate verticalglasswallsofthejar. Afterthewaxycuticlewas gentlyremovedusinga horsehairbrush(see Fish 1976),sevenof 10 attempts to ascend the innerleaf surfacesucceeded(P < .02, Fisherexacttest).The fluidimpounded withinthe tankis highly acidicwitha pH of2.8 to 3.0,themostextreme acidityyetfoundin bromeliads(cf. Laessle 1961;Benzing1980).However,no digestiveglandsare evidentin collectedmaterial,based on microscopicexamination of transverse, and tangential longitudinal, sectionsofleafbases. Third,Brocchiniatankscontaintheabundantremainsof dead antsand other thatdo notordinarily maketheirlivingin pools ofwater(fig.3; table arthropods chosenspecimensrevealed31 2). Analysisof thetankcontentsof 16 randomly familiesand 6 ordersofinsectsamongpresumptive prey,withmorethan90% of are knownto forage theindividuals beingants.The eightantgenerarepresented at nectaries(E. 0. Wilson,personalcommunication). frequently Commensal, This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTION OF CARNIVOROUS PLANTS 483 FIG 1.-Habitat andgrossaspectofBrocchinia reducta,GranSabana,1981.Notepitcherlikeleafrosettes. includepseudoscorpions nonpreyarthropods (dryspaces betweenleaves); spiand ders(centralopeningabovetank);mites(amongdeadleaves);andchironomid culicid(mosquito)larvae(tankfluid).Commensal mosquitolarvaeincludespecies and Runchomyta. of Wyeomyia odornotunlikethatproducedby Fourth,thetankfluidemitsa sweet,nectarlike thenectaries ofthecarnivorous Heliamphoraheterodoxa, whichis foundwithin a fewhundred metersofthestudysite.Sucha scentis highly unusual,ifnotunique, This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions 484 THE AMERICAN NATURALIST leafbase; widthof of cuticleon innersurfaceof lyophilized FIG.2.-SEM photograph 22 pum. Finestrandsofwax readilyexfoliate. photograph in our experiencewithscores of tankbromeliadsin the fieldand greenhouse. steyermarkii, TanksoffourotherBrocchiniaspecies(B. acuminata,micrantha, and tatei)inspectedemitno scentor a weakone; however,a similarodorcan be theirleafbases. releasedby crushing leafhairsofB. reductaare unusualin strucor absorbing Fifth,thetrichomes tureand can absorbaminoacids (andhenceanylikelyproductofthebreakdown materialin thetank)at a highrate.Trichomesplayan important of nitrogenous inepiphytic particularly inbromeliads, ofmineralnutrients roleintheabsorption and Tillandthe Bromelioideae speciesof the two moreadvancedsubfamilies, ofthethird sioideae(Benzing1976;Benzingetal. 1976;Benzing1980).Trichomes intowhichBrocchiniafalls,are Pitcairnioideae, subfamily and mostprimitive innutrient (Benzing1976;Benzingetal. absorption unimportant usuallyrelatively and regular are also usuallyless symmetric 1976).Theircellulararrangements forms(Tomlinson1969; tillandsioid thanthoseof bromelioidand (particularly) ofB. reducta,however,areexceptionally SmithandDowns 1974).The trichomes in thisrespect(fig.4). geometricand approachthose of certaintillandsioids appearto retaintheircytoplasm thecap cells atop thetrichomes Furthermore, kindlyprovidedby unusualtrait(Benzing1980).Autoradiographs intact,a highly rapidlyabsorbH3-labeledleucine,and BenzingshowthatB. reductatrichomes and is localizedinboththestalkcells(notunusualinbromelioids thatabsorption ofliving (fig.5). Portions unusual)oftrichomes andcap cells(highly tillandsioids) leafbases wereexcised,exposedto a 30-minpulse of H3-labeledleucine,and preparedusingthetechniquesofBenzing(1976)and Benzinget autoradiographs al. (1976). This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTION OF CARNIVOROUS PLANTS 485 ofleaves, reflective waxyinnersurface FIG.3.-Cross sectionofB. reductatank,showing abundantremainsofantsand otherinsects. This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions THE AMERICAN NATURALIST 486 TABLE 2 INSECT PREY FOUND IN SIXTEEN B. reducta TANKS (minimumnumberof individualsrequiredto accountforfragments notedforeachfamily; fraction ofindividuals represented by each ordershownin parentheses) Collembola(0.08%) ...........1 Entomobryiidae Coleoptera(2.5%) Anthicidae ......1 ........................ Chrysomelidae Curculionidae ... ...... Elateridae ...... ...... Lampyridae Ptilidae.......1 Scarabidae........ .........1 (Unidentified) Diptera(4.4%) Bibionidae ..... * ......... Dolichopodidae Drosophilidae.1 13 5 3 2 4 2 8 Membracidae. Hymenoptera (89.9%ants + 1.7%) Braconidae ........................ Chalcidoicea ...................... Formicidae ........................ Brachymyrmex Camponotus Otitidae ....1 Richardiidae ......... ........ Sarcophagidae 12 ........... (Unidentified) Hemiptera (0.3%) Reduviidae ........... 17 2 4 8 1,064 Crematogaster Hypoponera ......................... Halictidae 1 .................... Platygasteridae ......................... Sphecidae 1 1 ...................... (Unidentified) 4 Ichneumonidae.1 Phoridae.............................2 1 8 2 ~~~~~~~~~~Paratechina Pei Pseudomyrmex Solenopsis Muscidae3 Nematocera.........1 Sciaridae. Homoptera (1.1%) ........................ Aphidoidea .... Cicadillidae .8... ....................... Fulgoroidea Vespidae.1 3 we proposed,Brocchiniareductaqualifies Based on thecriteria as carnivorous. It emitsa nectarlike scenttoattract prey;has vertical, waxyleavestohelpcapture prey;and can absorbnutrients releasedfromdead preythroughits modified We have no directevidencethatabsorption ofnutrients increasesB. trichomes. butnotethatit growsin a nutrientreducta'sgrowth, survival,or reproduction, thatis hometo fourothergeneraofcarnivorous plantsand in poorenvironment whichcarnivory presumably yieldsan advantage. Pathway to Carnivoryin Brocchinia scenariofortheevolution ofcarnivory inBrocchinia We envisionthefollowing reducta.Excludingtheunexamined B. hechtioides, whichmayalso be carnivorous, B. reducta'sclosest relativesappear to be facultatively epiphytictank specieslikeB. tatei(SmithandDowns 1974).Epiphytic populations ofB. tateiin shady cloud forestshave a formtypicalof most tank bromeliads,witha basket- shapedrosetteofnearlyhorizontal greenleaves,welldesignedforlightcapturein a dimlylithabitatand forpassivecaptureof nutrients in theformof plantand animaldebris.Invasionof sunnysterilesavannasby B. reducta'sancestors presumably would have favoredthe evolutionof steeplyinclinedleaves with reflective andwaterloss (e.g., strongly waxycuticlesto reducelightinterception This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTION OF CARNIVOROUS PLANTS 487 fromleafbase, stainedwithtoluidine FIG.4.-Tangential(overhead)viewof trichome width50 pum. Trichomes becomeless denseand haveless prominent or blue,approximate morewornshieldstowardtheleaftip. see Ehleringer mineralpovertymay and Forseth1980;Givnish1984).Further, havefostered terrestrial populations ofB. tateiat chlorosis.In fact,sun-adapted, 1961,1966). highelevationsdo showthesetraits(Steyermark abouta tankwouldbe preadapBrightly colored,verticalwaxyleavesarranged attractor tationsfortheevolutionofcarnivory, inthattheycouldcoincidentally function. We believethe crucial entrapinsectswhileperforming theirprimary tocarnivory involvedleakageintothetankfluidofa sweet-smelling, adaptiveshift volatilecompoundstoredin theleafbases ofmanyBrocchiniaspecies.Thiskey of traitwouldattractinsectsto thetank,bringintoplaythepreadapted functions the leaves in preyattraction and promotethe evolutionof and entrapment, trichomes. nutrient-absorbing inB. reductais activepreyattraction Thus,thecrucialadaptation forcarnivory traitsare clearly via a sweet-smelling volatilecompound;otherleafandtrichome valuablein promoting carnivory buthave other,notreadilyseparablefunctions The dense waxy thatmay make theircontribution moreor less coincidental. in thatit is cuticlemayalso proveto be primarily an adaptationforcarnivory, foundmainlyon theunexposedinnerleafsurfaceswhereitwouldhavelittleeffect on leafheat load and transpiration. Whena planthas traitsthatfitit forprey but or attraction, capture, digestion, each of these traitshas substantialand functions forotherpurposes,withno clearallocationofenergysolely inseparable to call such plants to carnivory, we believe it would be moreparsimonious protocarnivorous ratherthancarnivorous.We proposeCatopsisberteroniana, This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions 488 THE AMERICAN NATURALIST *1Xek~~~~~~~~~~~~~~~~~~~~~~~1 - Ant ci This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTION OF CARNIVOROUS PLANTS 489 studiedby Fish (1976) and discussed above, as a possible instanceof protocarnivory and an independent,intermediatestep in the evolution of carnivoryin the tillandsioidbromeliads. Further research on this intriguingspecies is clearly needed, however. Brocchinia reducta is the least specialized of the known carnivorousplants, lacking recognizable digestiveglands and specialized nectaries,and possessing only a rudimentarycuticularlubricantaround its water pitfall.Nevertheless,a singletankusually holds more dead preythanseveral pitchersof nearbyHeliamphora. Presumably,thisis because Heliamphoratrapsconsistof singleleaves and persistonly as long as a leaf does, whereas Brocchinia traps are formedby leaf rosettesand persistas individualleaves die and are replaced. This relianceon leaf rosettes ratherthan leaves may hinder the evolution in Brocchinia and other bromeliadsofmoresophisticated,dorsiventralpitfallsliketheremarkablyconvergenttraps seen in otherpitcherplantfamilies. Brocchiniais unique amongflowering plantsin thatit is theonlygenus in which carnivoryis knownbut not universal.It thusaffordsan unparalleledopportunity to studythe evolutionof carnivoryin relationto othermodes of nutrientcapture, and we are currentlyinitiatinga comparativestudyof the ecology, morphology, nutritionalphysiology,photochemistry, and systematicsof species in the genus. Later papers will provide information on aspects of the morphology,ultrastructure,and volatile chemistryof B. reducta not covered in thispreliminary report. EVOLUTION OF CARNIVOROUS PLANTS Ecological Distribution Given the seemingadvantages of carnivory,whydo thereappear to be so few carnivorousbromeliads?Most bromeliadsare epiphytesgrowingin habitatswith littleor no soil, dependingfornutrientson rainwater,leachate fromsurrounding plants,and fallingdebris (Br-nzingand Renfrow1971, 1974; Benzing 1973). Carnivorywould appear to be a favorableadaptationunderthese conditions.Then why is carnivoryso rare among epiphytesgenerally(Benzing 1973; Thompson 1981)? Only 18 of at least 15,000epiphyticangiospermsare knownto be carnivorous (Madison 1977), an incidence60% lower thanthatseen forangiospermsas a whole, in whichroughly535 of 250,000 species are carnivorous(table 1). Both of theprecedingquestionscan be addressed ifwe can explainwhymostcarnivorous plants occur in habitatsthat are not only nutrient-poor, but sunnyand at least seasonallymoistas well, because mostepiphyticperches,althoughnutrient-poor, are eithershady or exposed to sunlightand regulardesiccation. AlthoughDarwin (1875) was thefirstto appreciatewhycarnivorousplantstend to be restrictedto nutrient-poor sites, therehas been littleexplicitrecognitionor explanation of the fact that most carnivorousplants grow in habitatsthat are sunnyand moistat least duringthe growingseason, mostlyin bogs, swamps, and aquatic habitats (Darwin 1875; Hepburn et al. 1920; Lloyd 1942; Plummerand Kethley 1964; Heslop-Harrison 1976; Erickson 1978; Green et al. 1979; Slack 1979; but see Heslop-Harrison1978; Thompson 1981; Luttge 1983). Such sites are This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions 490 THE AMERICAN NATURALIST occupied by Aldovandra,Biovularia,Byblis,Cephalotus,Darlingtonia,Dionaea, some Drosera, Genlisea, Heliamphora, Pinguicula, Polypompholyx,Sarracenia, and most Utricularia. There are but few exceptions to this rule (Lloyd 1942; Erickson 1978; Slack 1979). Drosophyllumlusitanicumremainsactive in arid sites duringthe dry Mediterraneansummer.A few shade-lovingDrosera live in the understoryof Queensland rain forests,but show signs of losing the carnivorous habitin havingfew glandulartentaclesper leaf. Nepenthes vines usually inhabit forestopenings on nutrient-poorsites (Richards 1936a, 1936b; Smythies 1964; Kurata 1976), but 6 of 71 species are epiphyticand a few grow under closed canopies. Most Utriculariaare aquatic or grow on moistopen ground,but 12 of 280 species are epiphyticin moist cloud forests(Taylor 1964; Madison 1977). CertainDrosera of semi-aridsouthwesternAustraliaare well known to occupy open, mineral-poor,upland sites but are not an exceptionbecause theyare active mainlyduringthe moistwinterand spring(Dixon and Pate 1978; Erickson 1978; Slack 1979). COST/BENEFIT MODEL To analyze why carnivorous plants are mainly restrictedto sunny, moist, nutrient-poor environments,we mustconsiderthe energeticbenefitsand costs of carnivoryin varioushabitats.Carnivoryshouldevolve ifthesebenefitsexceed the cost of a small investmentin carnivorousadaptations,because plantswithmutationsforsuch investmentsshould have an energeticadvantagein competingwith otherplants (see Givnish 1979, 1982). There are three potentialenergeticbenefitsassociated with carnivory.First, carnivorymay increase a plant's total rate of photosynthesisas a result of increased mineralabsorption,through(a) an increasedrateof photosynthesisper unit leaf mass or (b) an increase in the total leaf mass that can be supported. Medina (1970, 1971) showed that increasinga plant's supply of available soil nitrogencan enhance photosynthesisand enables it to develop higherconcentrationsof RuBP carboxylase-oxygenase,thephotosynthetic enzymeresponsiblefor CO2 capture; fertilizationstudies withothermineralnutrients(Naltr1975; Longstrethand Nobel 1980) have produced similarincreases in photosynthetic output. Weiss (1980) has confirmedthat nutrientinputthroughcarnivoryin Sarracenia flava can elevate the rate and seasonal durationof photosynthesisper unit leaf mass. Second, carnivorymay result in an increased level of nutrientsin seeds or increased seed production.Benzing (1976) and Benzing et al. (1976) have shown thatepiphyticbromeliadsdivertlarge fractionsof theirnitrogenand phosphorus to flowersand seeds, implyingthatnutrientsmay directly budgetspreferentially limitreproduction.Finally,carnivorymay serve to replace autotrophypartlywith heterotrophy as a source of chemical energy. All studiespointagainstthe last possibility:plantsobtainminerals,not carbon, fromcarnivory(Darwin 1875; F. Darwin 1878; Hepburn et al. 1920; Lloyd 1942; Plummerand Kethley 1964; Heslop-Harrison 1976). Dixon et al. (1980) found butthe uptakeof carbonfromC 14-labeledDrosophila fedto Drosera erythrorhiza, This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTION OF CARNIVOROUS PLANTS 491 formin whichit was absorbedwas not studied(see Chandlerand Anderson aminoacids 1976b).The carbonabsorbedseemsmorelikelyto be in N-bearing sinceDarwin(1875)and manysubsethanin readilyconvertedcarbohydrates, inDroseraoccursonlyin secretion quentinvestigators haveshownthatglandular material.Furthermore, Chandlerand Anderson(1976a) responseto nitrogenous and no haveshownthatDroserafedpreyina darkenedroomshowlittlegrowth, ofinorganic salts difference fromplantssuppliedinsteadwitha completemixture roleforcarbonheterotroimplying a negligible underthesamelightconditions, phy. partof The secondbenefit withrespecttoreproduction caneasilybe considered the firstbenefitto growth:if the mineraloutputof trapsproducedearlyin thatresultsin increased is divertedintophotosynthetic machinery development ratesofgrowthand moretrapsand leaves,thentheoutputofa fewlateradded model, Thus,forthesakeofthefollowing trapscan be allocatedto reproduction. is enhancedphotosynthesis. benefit ofcarnivory we shallassumethattheprimary condiHow shouldthebenefits andcostsofcarnivory varywithenvironmental tions?Considera plantwitha givenbiomassinleavesandroots.As theamountof inefficient energydevotedto carnivory(in termsof lures,photosynthetically increase traps,or digestiveenzymes)increases,thereshouldbe a corresponding rateofphotosynin theamountof nutrients absorbed.As a result,theeffective thesisper unitleaf mass-either in absolutetermsor relativeto the cost of nutrients and carbohydrates fornewleaves-should increase(fig. accumulating and theresulting as theamountofenergydevotedto carnivory 6). Furthermore, benefitexpectedshould mineralinputcontinueto increase,thephotosynthetic tendto plateauas factorsotherthannutrients limitphotosynthesis or theconversionofphotosynthate intonewleaves. Theextenttowhichphotosynthesis input canbe enhancedbyincreasedmineral rateof theeffective conditions. By definition, clearlydependson environmental areinshortsupplyandlimit toincreaseunlessnutrients photosynthesis is unlikely sites. photosynthesis, so thatthe greatestbenefitis expectedin mineral-poor Studiesby Sorensonand Jackson(1968) on Utriculariaand by Chandlerand thattheusual increasein growthof carAnderson(1976a) on Drosera confirm substrates largelydisappears nivorousplantssuppliedwithpreyon nutrient-poor iffactorslikelight inthesubstrate as nutrient increases.Furthermore, availability or waterare in shortsupply,thentheycan limitphotosynthesis (Bannister1976) andtheextenttowhichnutrients addedbycarnivory (orothermeans)can elevate photosynthesis. For example,Gulmonand Chu (1981)have shownthat,inDipphotosynthesis lacus auranticusexposed to variouslevels of soil fertilization, thanat highlight increasesmoreslowlywithleafN content at lowlightintensities informto intensities. Thehypothetical benefit 6 aresimilar curvesshowninfigure thoseactuallyobtainedbyGulmonandChu (1981)fortheresponseofphotosynthesisto increasedleaf N contentinducedby fertilization. Similartrendsare expectedifincreasedmineralsupplyincreasesnotthemineralcontentperleaf, buttherateof conversionofphotosynthate intonewleaves at constantmineral content.As nutrient availability increases,therateat whichnewleaves can be and or phosphorus, mineralslikenitrogen producedwilldependless on limiting This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions THE AMERICAN NATURALIST 492 C COST 0 P7 SUNNY. MOIST CD_| u H- z SHADY OR DRY Of)1 COST/INEFFICIENCY PHOTOSYNTHETIC ASSOCIATEDWITH CARNIVORY FIG. 6.-Photosynthetic andcostsassociatedwithdifferential levelsofinvestment benefits conditions in carnivorous site,as a function ofenvironmental adaptations in a nutrient-poor shouldbe rate(P) resulting fromaddednutrients (see text).Enhancement ofphotosynthetic to plateau,ina well-lit andmoistenvironment morerapid,andshowless tendency (PI) than in siteswherelightorwateraremorelikelytolimitphotosynthesis (P2). Dashedlines= net carnivbenefit andcost(C) ofobtaining nutrients through difference betweenphotosynthetic in carnivory ory. Carnivoryshouldevolve wheneverthe benefitof a smallinvestment curveslopesupwardnearC= 0. exceedsitsowncost,i.e., whenthenetprofit of carbonskeletons.The latterclearlydependson the moreon theavailability limitationof photosynthesisby lightor water availability,so the rate of conversion should also rise most quickly and plateau most slowly in well-lit,moist, sites. nutrient-poor Thus, in sterilehabitats that are sunnyand moist, carnivoryshould have its in gainsresulting fromaddedinvestments greatest impactandthephotosynthetic shouldrisequicklyandplateauslowly(fig.6). In sterilehabitatsthatare carnivory activity, lightor waterare shadyand/ordryduringtheperiodofphotosynthetic shouldbe smallerand morelikelyto limitgrowthand thegainsfromcarnivory and betweenphotosynthetic benefits plateaumorerapidly.Thus,thedifference in carnivory, costs are morelikelyto be positiveat low levels of investment and hence promoteits evolution,in nutrient-poor habitatsthatare also sunny Suchconditions andmoistduringtheperiodofphotosynthetic mayalso activity. in increaseinsectabundanceand thebenefitaccruingfroma giveninvestment carnivory. in generaland This modelpartlyexplainswhycarnivory is rarein epiphytes inparticular, becausemostepiphytes bromeliads occureitheron shadyperchesor on betterlitbranchessubjectto frequent desiccation.Nepenthesvinesprobably in avoidthisproblembymaintaining contactwithmorereliablestoresofmoisture the soil. Epiphytic Utriculariausually grow in wet cloud forestswhere water stress is rare even highin the canopy, or in pools impoundedat the expense of otherplants (oftenbromeliads)by the hosts' own inefficient, overlappingphotosyntheticstructures. This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTIONOF CARNIVOROUS PLANTS 493 The modelalso explainsmanyaspectsofseasonalheterophylly incarnivorous plants.Manycarnivoresfailto producetraps,or produceleaves witha higher proportion of photosynthetically efficient tissue,during"unfavorable"seasons whenfactorsotherthannutrients maylimitgrowth(Erickson1978;Slack 1979). SomeSarraceniaproducetraplessphyllodesduring latesummer droughts (Weiss 1980).Dionaea producesleaves withbroadphotosynthetic petiolesand small trapsinwinter.Cephalotusappearsto produceonlyphyllodes, nottraps,during thecool winterin southwestern Australia;itsreedswamphabitatremainsdamp through the springand drysummerwhenthepitchersare active(Lloyd 1942). inoverlydryorshadyconditions Nepenthesoftenfailsto producepitchers (Slack 1979).The seasonalflushofglandular leavesin Triphyophyllum justprecedesthe onset of the rainyseason and may correspondto the yearlyflushof insect fromglanavailability (Greenet al. 1979).The long-term shiftin Triphyophyllum dular,insectivorous seedlingsand juvenilesto eglandularadultvines remains enigmatic forlack ofdata on themicroclimates experienced byeach lifestage. Carnivoryand Myrmecophily is so strongly Finally,ifcarnivory selectedagainstin epiphytes, one mayask why myrmecophily is relativelycommon(Thompson1981). Myrmecophilous forantcoloniesinholloworswollennodes,petioles,orleaf plantsprovideshelter bases,andoftenprovidefoodintheformofextrafloral nectaries orBeltianbodies (Bequaert1922;Wheeler1942).In turn,theplantsmayreceiveprotection for herbivoresand competitors attackedby theirguests,and may receivean increasedsupplyofnutrients fromfoodwastes,dead nestmates, anddebrispacked bytheantsintoplantrecesses(Benzing1970;Janzen1974a;Huxley1978;Rickand Myrmecodiain the son 1979). Several epiphytes,like Hydnophytum Rubiaceaeand Tillandsiacaput-medusae in theBromeliaceae,are ant-fed myrand knownto receivenutrients mecophytes fromtheirguests(Thompson1981). This nutrient inputis probablysimilarto thatwhichcouldbe obtainedthrough carnivory,so why is ant-fedmyrmecophily relativelymorecommonin (and indeed,by current data nearlyrestricted to) epiphytes? reasonis thatmyrmecophily Perhapsone important otherthan yieldsbenefits nutrient input,notablydefenseagainstherbivores. The benefits of suchdefense are notlikelyto showthesametrendsbetweenhabitatsas thebenefits accruing fromcarnivory.In particular, the benefitsof defending leaves may be more in unproductive, important shady,or dryenvironments whereleaves are relativelymore costlyto replace thanin moreproductivesites (Janzen1974b). thathavebeencarefully However,in certainant-fed myrmecophytes examined, and Myrmecodia(Janzen1974a;Huxley1978),a defensive as in Hydnophytum role for the guest ants seems unlikelyeven thoughJanzen(1974a) reports willrespondaggressively to theirhost.Thus, Iridomyrmex to severedisturbance cautionmustbe exercisedin usingthisargument fortherelativeadvantagesof and carnivory. myrmecophily tendto be ant-fedratherthan Thompson(1981)has suggestedthatepiphytes carnivorous becausetheiraccess to wateris so limitedthattheycannotproduce This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions 494 THE AMERICANNATURALIST the glandularsecretionsassociatedwithcarnivory.Drosophyllum's abilityto maintain secretions thelongdryMediterranean summer glandular through (Slack 1979)undercuts thisargument, butitcan be incorporated in ourmodelbynoting thatdryconditionswouldincreasetheenergeticcost of secreting attractive or digestivefluids,therebyincreasingthe cost associatedwitha givenphotosyntheticbenefit.This woulddecreasethe slope of the benefitcurve(fig.6) and renderthe evolutionof carnivoryless likely.Since the permeable,nutrientare usuallycontainedwithina absorbingsurfacesof ant-fedmyrmecophytes cavity,Thompson(1981)suggeststhattherateofwaterloss andcostsassociated withmyrmecophily would be relativelyless, so thatmyrmecophily would be indrierhabitats.In addition, ant-fed favoredovercarnivory plantsshouldhavean advantagein treetopsitesand openhabitatsover"basket" epiphytes (e.g., tank bromeliads, staghorn ferns)thatpassivelytrapfalling debris,giventhepaucityof suchdebrisand associatedmineralsin thesesites(Janzen1974a). A criticaltestoftheseideas wouldbe to examinethenatureoftherelationship shrubsfoundin theunderbetweenantsand plantsin variousmyrmecophilous rainforests, suchas speciesofClidemia,Maieta,and Tococa storyofneotropical in a given in the Melastomataceae.If the principlebenefitof myrmecophily thenthephotosynthetic speciesis increasedmineralabsorption, gainsthusobthatarefed, tainedshouldbe lowerinshadyhabitats (see above).Myrmecophytes theantsproviding butnotdefended-whether thedefenseis director through nutrients essentialforthesynthesis ofdefensive compounds (D. Janzen,personal ofcarnivorous thusparallelthedistribution communication)-should plantsalong oflightintensity, gradients and be largelyrestricted to well-litareas and/orlose formyrmecophily intheshade.Onlymyrmecophytes thatareat least adaptations defended,directlyor indirectly, by theirguestsshouldbe foundin the shade. Hence,onetestofourcost/benefit modelwouldbe toobservewhether understory aredefended myrmecophytes bytheirguests;iftheyarenot,andareonlyant-fed, thepredictions of the-modelwouldbe contradicted. Brocchiniaacuminata,on whichwe havebegunresearch,mayprovidematerial fora directtestinthatitis in openbogsand savannas,and loses itsswollenleafbases and myrmecophilous guestsinshadylocations(T. Givnish,D. Benzing,andE. L. Burkhardt, personal observation). The cost/benefit modeldevelopedinthispaperdemonstrates thatthe explicitly advantageof carnivory dependson the suppliesof lightand wateras well as nutrients. It thusprovidesa unified basisforunderstanding whycarnivory is rare in epiphytes, whycarnivory is commonin sunny,moist,sterilesites,and why myrmecophily appearsto replacecarnivory as a meansof nutrient capturein epiphytes.It also providesa conceptualframework forphysiological studieson therelationships betweencarnivory, nutrient status,and photosynthesis. SUMMARY Brocchinia reducta is the firstdocumentedcase of carnivoryin the Bromeliaceae.Its erectleavesforma yellowish witha cuticular lubricant cylinder inner fluid a its that emits nectarlike and bear on fragrance, surface,impound This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTION OF CARNIVOROUS PLANTS 495 trichomescapable of absorbingamino acids fromthis fluidin which numerous insects, mainly ants, drown. Trichome absorptivityand aspects of trichome structureappear unique in theprimitivesubfamilyPitcairnioideae.We presentthe firstrigorousdefinitionof carnivoryin plants,and discuss its implicationsforthe identificationof cases of carnivoryand protocarnivoryin bromeliads. A cost/ benefitmodel forthe evolutionof carnivoryis developed to analyze whycarnivorous plantsare restrictedmainlyto sunny,moist,nutrient-poor sites and seasons, and why carnivoryis rare in epiphytesand otherbromeliads.The relativeadvantages of carnivoryand ant-fedmyrmecophily are discussed in termsofthismodel, and predictionsmade regardingthe natureof the ant-plantmutualismin understorymyrmecophytes. ACKNOWLEDGMENTS For assistance in various phases of this study,we wish to thankJ. Silva, J. Steyermark,and L. Arias in Venezuela, and D. Benzing, P. Tomlinson,E. 0. Wilson, H. Levi, W. Muchmore,A. Spiegelmann,C. Wood, N. Woodley, and T. Zavortinkin the United States. D. Fish and T. Weiss kindlyprovidedaccess to theirunpublishedmanuscripts;D. Janzen,an anonymousreviewer,the Ecology Discussion Group at the Universityof Washington,and H. Neufeld contributed usefulcomments.V. Elliottprovidededitorialassistance. Supportwas provided by grantsto the seniorauthorfromthe AtkinsGarden Fund and the MiltonFund of Harvard University.Valuable logisticalsupportin 1981 was provided by Lt. Col. N. J. Rivas Espinoza, Commandant of the Ciudadela at Luepa, Estado Bolivar. LITERATURE CITED Adams, R. M., and G. W. Smith.1977.An SEM surveyof thecarnivorouspitcherplantgenera.Am. J. Bot. 64:265-272. Atlas de Venezuela. 1969. Ministeriode Obras Publicas, Direcci6n de Cartograffa, Nacional Respublica de Venezuela, Caracas. Bannister,P. 1976. Introductionto plant physiologicalecology. Blackwell Scientific,Oxford. Benzing, D. H. 1970. An investigationof two bromeliadmyrmecophytes:Tillandsia butziiMez, T. caput-medusae E. Morrenand theirants. Bull. TorreyBot. Club 97:109-115. 1973. Mineral nutritionand related phenomenain Bromeliaceae and Orchidaceae. Q. Rev. Biol. 48:277-290. 1976. Bromeliadtrichomes:structure,function,and ecological significance.Selbyana 1:330348. 1980. Biology of the bromeliads. Mad River Press, Eureka, Calif. Benzing,D. H., K. Henderson,B. Kessel, and V. Sulak. 1976.The absorptivecapacities ofbromeliad trichomes.Am. J. Bot. 63:1009-1014. Benzing, D. H., and A. Renfrow. 1971. The biology of the epiphyticbromeliadTillandsia circinata Schlecht. I. The nutrientstatus of populationsin south Florida. Am. J. Bot. 58:867-873. 1974. The mineralnutritionof Bromeliaceae. Bot. Gaz. 35:281-288. Bequaert, J. 1922. Ants in theirdiverserelationsto theplantworld.Bull. Am. Mus. Nat. Hist. 45:333583. Brewer-Carias,C. 1976. La vegetaciondel mundoperdido. Fundaci6n Eugenio Mendoza, Caracas. Chandler,G. E., and J. W. Anderson. 1976a. Studies on the nutritionand growthof Drosera species withreferenceto the carnivoroushabit. New Phytol.76:129-141. This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions 496 THE AMERICAN NATURALIST . 1976b.Uptakeand metabolism of insectmetabolites by leaves and tentaclesof Drosera species.New Phytol.77:625-634. Darwin,C. 1875.Insectivorous plants.Appleton, London. Darwin,F. 1878.Experiments on thenutrition of Droserarotundifolia. J. Linn. Soc. Bot. Lond. 17:17-32. Dixon,K. W., andJ.S. Pate. 1978.Phenology, morphology, andreproductive biologyofthetuberous Lindl. Aust. J. Bot. 26:441-454. sundew, Drosera erythrorhiza Dixon,K. W., J.S. Pate,andW. J.Bailey.1980.Nitrogen nutrition ofthetuberous sundewDrosera faunabyitsglandular leaves. Lindl.withspecialreference tocatchofarthropod erythrorhiza Aust.J.Bot. 28:283-297. Ehleringer, J.A., and I. M. Forseth.1980.Solartracking byplants.Science210:1094-1098. Erickson,R. 1978.Plantsofpreyin Australia.University ofWestern AustraliaPress,Nedlands. Fish,D. 1976.Structure andcomposition oftheaquaticinvertebrate community inhabiting bromeliads of in southFloridaandthediscovery bromeliad. Ph.D. diss. University ofan insectivorous Florida,Gainesville. A. J. 1973.Transandean inEcuador.Ecology54:1389-1393. Gilmartin, distribution ofBromeliaceae T. J. 1979.On theadaptivesignificance Givnish, ofleafform.Pages375-407in0. T. Solbrig,S. Jain, andP. H. Raven,eds. Topicsinplantpopulation G. B. Johnson, biology.ColumbiaUniversityPress,New York. 1982.On theadaptivesignificance ofleafheight inforestherbs.Am.Nat. 120:353-381. 1984.Leaf and canopyadaptations in tropicalforests.Pages 51-84 in E. Medina,H. A. eds. Physiological Mooney,and C. Vdsques-Ydnes, ecologyof plantsof thewettropics. Junk,The Hague. andleafglandfeatures Green,S., T. L. Green,andY. Heslop-Harrison. 1979.Seasonalheterophylly in Triphyophyllum (Dioncophyllaceae), a newcarnivorous plantgenus.Bot. J. Linn.Soc. 78:99-116. on photosynthesis, leaf Gulmon,S. L., and C. C. Chu. 1981.The effectof lightand nitrogen and drymatterallocationin the chaparralshrub,Diplacus aurantiacus. characteristics, Oecologia49:207-212. and allied of nutrients Hepburn,J. S., F. M. Jones,and E. Q. St. John.1920.The absorption in thepitchers Inst.189:147-184. oftheSarraceniaceae. J.Franklin phenomena Y. 1975.Enzymereleasein carnivorous Heslop-Harrison, plants.Pages525-578inJ.T. Dingleand R. T. Dean, eds. Lysosomesinbiologyandpathology. Vol. 4. NorthHolland,Amsterdam. anddigestuptakeincarnivorous 1976.Enzymesecretion plants.Perspect. Exp. Biol.2:443-476. 1978.Carnivorous plants.Sci. Am.238:104-115. Heslop-Harrison, Y., and J. Heslop-Harrison. 1980.Chlorideion movement and enzymesecretion fromthedigestive glandsofPinguicula.Ann.Bot. 45:729-731. Huxley,C. R. 1978.Theant-plants Myrmecodia andHydnophytum (Rubiaceae)andtherelationships betweentheirmorphology, antoccupants, physiology andecology.NewPhytol.80:231-268. inSarawak:mutualism Janzen,D. H. 1974a.Epiphytic myrmecophytes through thefeeding ofplants 6:237-259. by ants.Biotropica 1974b.Tropicalblackwater rivers,animals,and mastfruiting in theDipterocarpaceae. Biotropica6:69-103. anachronisms: thefruits thegomphotheres ate. Janzen,D. H., and P. S. Martin.1982.Neotropical Science214:19-27. Kurata,S. 1976.NepenthesofMountKinabalu.SabahNationalParksPublication 2, Kota Kinabalu, Malaysia. Laessle,A. M. 1961.A micro-limnological studyofJamaican bromeliads. Ecology42:499-517. Lloyd,F. E. 1942.The carnivorous plants.ChronicaBotanica,Waltham. D. J.,andP. S. Nobel.Nutrient Longstreth, influences onleafphotosynthesis. PlantPhysiol. 65:541-543. U. 1983.Ecophysiology ofcarnivorous plants.Pages489-517in 0. L. Lange,P. S. Nobel, Lfuttge, C. B. Osmond,and H. Ziegler,eds. Encyclopedia ofplantphysiology. Vol. 12C. Springer, New York. E. L. 1974.Evolutionary McWilliams, ecology.Pages40-55 in L. B. SmithandR. J. Downs,eds. Bromeliaceae (Pitcairnioideae). FloraNeotropica, Monograph 14. Hafner,New York. This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions EVOLUTION OF CARNIVOROUS PLANTS 497 Madison, M. 1977. Vascular epiphytes:theirsystematicdistributionand salient features.Selbyana 2:1-13. Maguire, B. 1970. On the floraof the Guayana Highlands. Biotropica2:85-100. Mayr,E., and W. H. Phelps, Jr.1967. The originof the birdfaunaof the South Venezuelan highlands. Bull. Am. Mus. Nat. Hist. 136:273-327. capacity,and carboxydismutMedina, E. 1970. Relationshipsbetween nitrogenlevel, photosynthetic ase activityin Atriplexpatula leaves. Carnegie Inst. Wash. Yearb. 69:655-662. 1971. Effect of nitrogensupply and light intensityduringgrowthon the photosynthetic capacityand carboxydismutaseactivityofAtriplexpatula ssp. hastata. CarnegieInst. Wash. Yearb. 70:551-559. Meyers,D. G., and J. R. Strickler.1979. Captureenhancementin a carnivorousaquatic plant:function of antennae and bristlesin Utriculariavulgaris. Science 203:1022-1025. Ndtr,L. 1975. Influenceof mineralnutrition on photosynthesisand the use of assimilates.Pages 537555 in J. W. Cooper, ed. Photosynthesisand productivityin different environments.CambridgeUniversityPress, Cambridge. Picado, C. 1913. Les Bromeliacees epiphytesconsidereescomme milieubiologique. Bull. Sci. France Belg. Ser. 7, 47:216-360. Pittendrigh, C. S. 1948. The bromeliad-Anopheles-malaria complexin Trinidad.I. The bromeliadflora. Evolution 2:58-89. Plummer,G. L., and J. B. Kethley. 1964. Foliar absorptionof amino acids, peptides, and other nutrientsin the pitcherplant,Sarracenia flava. Bot. Gaz. 125:245-260. Rees, W. E., and N. A. Roe. 1980. Puya raimondii(Pitcairnioideae,Bromeliaceae) and birds: a hypothesison nutrientrelationships.Can. J. Bot. 58:1262-1268. Richards,P. W. 1936a. Ecological observationson the rainforestof Mount Dulit, Sarawak. Part I. J. Ecol. 24:1-37. 1936b. Ecological observationson the rain forestof Mount Dulit, Sarawak. Part II. J. Ecol. 24:340-360. Rickson, F. R. 1979. Absorptionof animal tissue breakdownproductsintoa plant stem-the feeding of a plant by ants. Am. J. Bot. 66:87-90. Slack, A. 1979. Carnivorousplants. M.I.T. Press, Cambridge. Smith,L. B., and R. J. Downs, eds. 1974. Bromeliaceae (Pitcairnioideae).Flora Neotropica, Monograph 14. Hafner,New York. Smythies,B. E. 1964. The distributionand ecology of pitcherplants (Nepenthes) in Sarawak. In UNESCO Humid Tropics Symposium,UNESCO, Paris. Sorenson, D., and W. T. Jackson. 1968. Utilizationof parameciumby Utriculariagibba. Planta 83:166-170. Steyermark,J. A. 1961. Brocchinia. Bromeliad Soc. Bull. 1:35-41. 1966. Flora del Ptari-tepui.Acta Bot. Venez. 1:30-104. 1976. Future outlook forthreatenedand endangeredspecies in Venezuela. Pages 128-135 in G. T. Prance and T. S. Elias, eds. Extinctionis forever.New York Botanical Garden, New York. 1982. Relationshipsof some Venezuelan forestrefugeswithlowlandtropicalfloras.Pages 182220 in G. T. Prance, ed. Biological diversification in the tropics.Columbia UniversityPress, New York. Taylor, P. 1964. The genus UtriculariaL. (Lentibulariaceae) in Africa (south of the Sahara) and Madagascar. Kew Bull. 18:1-245. Thompson,J. N. 1981. Reversed animal-plantinteractions:the evolutionof insectivorousand ant-fed plants. Biol. J. Linn. Soc. 16:147-155. Tomlinson,P. B. 1969. Commelinales-Zingiberales. In C. R. Metcalfe,ed. Anatomyof the monocotyledons. Vol. III. Oxford UniversityPress, Oxford. Weiss, T. E., Jr.1980. The effectsoffireand nutrientavailabilityon thepitcherplantSarraceniaflava L. Ph.D. diss. Universityof Georgia, Athens. Wheeler, W. M. 1942. Studies of neotropical ant-plantsand theirants. Bull. Mus. Comp. Zool. 90(1): 1-262. Wurdack,J. J. 1964. A Peruvian bromeliadtrove. Bromeliad Soc. Bull. 14:110-116. This content downloaded from 128.104.1.219 on Thu, 30 May 2013 13:53:02 PM All use subject to JSTOR Terms and Conditions