Phenology and pollination biology of Ceiba pentandra (Bombacaceae) in the wet forest of southeastern Costa Rica

2008·3.8 MB·English

by  Rojas-Sandoval, Julissa

© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at
Phenology and pollination biology of Ceiba
pentandra (Bombacaceae) in the wet forest of south-
eastern Costa Rica
Fenología y biología de la polinización de Ceiba pentandra
(Bombacaceae) en el bosque húmedo del Sudeste de Costa Rica
Julissa ROJAS-SANDOVAL, Katharina BUDDE, Mauricio FERNÁNDEZ,
Eduardo CHACÓN, Mauricio QUESADA & Jorge A. LOBO
Abstract:PhenologicalpatternsandpollinationbiologyofCeibapentandrawerestudiedinnaturalpopulationslocatedinwet
forestsofSECostaRica.Althoughthetimingofleaffall,flowering,andfruitingactivitywererestrictedtothedryseasonandoc-
curredinasmalltimeperiod,thisspeciesshowedgreatvariabilityattemporalandindividuallevelinthefrequencyandintensi-
tyoffloweringandfruitingevents.Weobservedhighreproductiveactivityforwetforesttreesin2003thatcoincidedwithanad-
vanceinfloweringtimingtoDecember.Pollenofthisspeciesdemonstratedhighgerminationratesuntil14hoursaftergrainswere
removedfromflowersandinself-pollinatedflowers,pollentubeswereabletogrowoverthestyleinasimilarproportiontoflow-
erswithoutcrossingpollination.However,inflowerswithself-pollination,fruitsetwasnilandthisresultsuggeststheexistence
oflateincompatibilitybarriersforthisspecies.Inaddition,differencesinfruitproductionforflowerswithcross-pollinationand
naturalpollinationtreatmentssuggestthatfruitsetinC.pentandraispartiallylimitedforthearrivalofexogenouspollen.Inthe
wetforestoftheSEofCostaRica,C.pentandraflowersarevisitedbynectarivorous,frugivorousandgeneralistbatspecies.How-
ever,differencesindiversityandabundanceoffloralvisitorswereobservedinregardstootherforesttypes(i.e.,dryforest).In-
deed,inthisregion,speciessuchasPhyllostomusdiscolor,whichisveryabundantinGuanacaste,werenotobserved,andneither
wereagreatnumberofbatspeciesflyingaroundfloweringtreesasiscommonfortreesofthisspecieslocatedinthedryforestsof
CostaRica.
Keywords:Bombacaceae,Ceibapentandra,CostaRica,plantreproduction,phenology,pollinationbiology,bats.
Resumen:LafenologíaydiversosaspectosdelabiologíadelapolinizaciónCeibapentandrafueronestudiadosenpoblacionesen
elbosquehúmedodelSEdeCostaRica.Aunquelacaídadehojas,floraciónyfructificaciónocurrenenunperíodolimitadode
tiempoenlaépocaseca,existeunaampliavariacióntemporaleindividualenlafrecuenciadefloraciónyfructificación.Seob-
servóunaño(2003)dealtareproducciónpoblacionalquecoincideconunadelantodelafloraciónalmesdeDiciembre.Elpolen
deCeibapentandramuestraaltacapacidaddegerminaciónhasta14horasdespuésdesuremoción,yenfloresautofecundadases
capazdedesarrollarsehastaelfinaldelestiloenlamismaproporciónqueencrucesoutcross.Sinembargoeléxitodefrutosself
esnulo,loquemuestrabarrerastardíasdeincompatibilidad.Elincrementodelfruitsetconpolinizacióncruzadaconrespectoa
polinizaciónnaturalmuestraqueC.pentandraesunaespecielimitadaporpolenexógeno.LasfloresdeC.pentandradelbosque
húmedodelSEdeCostaRicasonvisitadasporvariasespeciesdemurciélagos,peronofuéobservadalaespecieP.discolorquees
muyabundanteenGuanacaste.Tampocoseobservarongrandesconcentracionesdemurciélagoscomoescomúnenelbosqueseco
deCostaRica.
Palabrasclave:Bombacaceae,Ceibapentandra,CostaRica,reproduccióndeplantas,fenología,biologíadelapolinización,mur-
ciélagos.
Stapfia88,
zugleichKatalogeder
oberösterreichischen
Landesmuseen
NeueSerie80(2008):
539-545
© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at
IntheSEofCostaRica,C.pentandraisanemergent
tree with tall individuals (> 60 m) that are part of the
landscapeinlowlandopenareas,aswellasinmountain-
ousareas(Figure1).Althoughmanyindividualsremain
in agricultural areas mainly due to the low value of its
wood and because they are used to provide shade for
livestock; the reproductive population size is gradually
reducedbynaturalmortality,selectiveloggingandcon-
straints on natural regeneration (ROJAS-SANDOVAL
2004).
Themaingoalofthisstudywastoprovidedataon
phenologicalpatternsofCeibapentandrapopulationslo-
catedinwetforestsoftheSEPacificCoastofCostaRi-
ca, as well as general aspect on pollination biology
mainlyrelatedtothecapacityofthisspeciestoauto-fer-
Fig.1:CeibapentandratreesatLaPalma,OsaPeninsula.Afloweringtree tiliseandtheirdependenceonpollinatoragents.Specif-
withoutleaves(left)isobservedbesideatreeinfullleaf(right)withrecently
ically we present information on: (1) flowering and
replacedfoliage.Thistreedidnotproduceflowersthatyear.Thissceneisan
exampleofthevariationinthereproductionfrequencyamongtreesforthis fruiting frequencies for C. pentandra populations locat-
species. ed in SE Costa Rica, (2) the effect of pollen source (i.
e.,selfingvs.outcrossing)inthefruitsetoftreeslocat-
Introduction
edinthesepopulations,(3)pollenviability,(4)differ-
Ceibapentandraisapantropicalcanopytreeoftrop- encesinpollenloadsandpollentubenumbersinflow-
ical dry and wet forests, distributed in the neotropics ers with self-pollination, outcross and natural pollina-
from Mexico to the Amazonian basin and in the tiontreatments,and(5)batspeciesvisitingC.pentan-
Palaeotropics throughout western Africa (HARTSHORN dra flowers in some locations of the SE forest of Costa
1983).Thisspeciesisatall(>40m)emergentdecidu- Rica.
oustreethatmayreachadiameter(d.b.h.)ofmorethan
200 cm. C. pentandra has hermaphroditic flowers with
Material and methods
fivestamensaroundaprotrudingstyle,andpinkpetals
(CASCANTE-MARÍN 1997). Flowers show nocturnal an- Studysites
thesis and produce large amounts of nectar, and in We conducted our study in the tropical wet forest
neotropical forests phyllostomid bats are reported as located in the South Pacific Coast of Costa Rica
pollinators(GRIBELetal.1999).Fruitsareelliptic,with (8°32‘N,83°18‘W).Thedryseasoninthissiteextends
seeds surrounded by a pale yellow silk cotton, used for from December to April but it is not as intense as in
winddispersal.
tropicaldryforest.(GÒMEZ&HERRERA1985)Theav-
ThefloweringandfruitingepisodesofC.pentandra erageannualrainfallhereis3500mm.AllC.pentandra
are limited to the dry season in neotropical forests individuals used in this study were reproductive adults
(FRANKIEetal.1974,GRIBELetal.1999).Forinstance, withadiameteratbreastheightgreaterthanonemetre.
on the Pacific Coast of Central America, trees located Weidentifiedtwopopulationsofisolatedtrees(Boruca-
at different latitudes within this region flower simulta- Chacaritan=39andLaPalma-PuertoJimenezn=30)
neouslyinJanuary(LOBOetal.2003).Intheneotrop- and two populations consisted of trees in continuous
ics,thereproductivephenologyofthisspecieshasbeen forest(GolfoDulceForestryReserven=8,andSirena
describedassupra-annualwithgreatvariationinrepro- BiologicalStationn=16).
ductive intensity between individuals, populations and
Phenologicalobservations
years(FRANKIEetal.1974,GRIBELetal.1999,LOBOet
al. 2003). However, synchronised population patterns Inordertodeterminethesequenceofphenological
havebeenobservedinyearsofgreaterreproductivein- phases of trees, we recorded the phenology of marked
tensity (GRIBEL et al. 1999, LOBO et al. 2003). This treeseverytwoweeksfromDecembertoAprilforthree
speciespractisesmixedmating(MURAWSKI&HAMRICK years(2001,2002and2003),startingonNovember1st
1992,LOBOetal.2005)andself-pollentubesgrownor- each year. Trees were observed with binoculars. Leaf,
mallyinthestyles(GRIBELetal.1999).However,man- flowerandfruitphenologiesweredeterminedbasedon
ualself-pollinationsintreesfromtheAmazonianbasin the percentage of cover of the crown suggested by
resultsinlowlevelsoffruitset(GRIBELetal.1999). FOURNIER(1974).
540
© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at
Fig.2:Phenologicalpatterns
ofleaves.....flowers———
andfruits——— forCeiba
pentandratreesinthe
tropicalwetforestofSECosta
Ricaoverthreeyears(2001,
2002and2003).
Pollengerminationandpollentubegrowth and styles were stained with aniline blue and observed
underafluorescencemicroscopetodeterminatetheto-
Pollengerminationexperimentswereperformedfol-
lowing the methodology proposed by BREWBAKER & tal of germinated pollen grains on stigma area and
KWACK (1963). Closed flowers from two trees were pollentubesgrowingoverthestyle.
baggedinlateafternoonandpollenwasremovedimme-
Pollinationexperiments
diately after flower anthesis and stored in Eppendorf
tubes(0.5ml)atroomtemperature.Pollengrainswere To determine the breeding system of C. pentandra,
placed in culture plates containing germination media two different pollination treatments (self-pollination
andwetfilterpaperat0,1,2,14,18,22,26,36,48,60 and cross-pollination) were performed on 1158 flowers
and 72 hours after grainswere removed from the flow- from five different trees. Flowers were marked and
ers. Culture plates were kept at room temperature and baggedbeforefloweranthesistopreventpollinatorvis-
thetotalofgerminatedpollengrainswerecountedun- itation. The treatments were: (1) Cross-pollination:
der a light microscope after six hours. The presence of when bagged flowers opened, they where emasculated
pollentubesonstigmaandstylewasanalysedinflowers and hand-pollinated with pollen from at least two to
fromsixdifferenttrees,inwhichweperformedpollina- three different donors (n = 610). (2) Self-pollination:
tionexperiments(i.e.self-pollination,cross-pollination the bagged flowers were hand-pollinated with pollen
and natural pollination). Early in the morning, we re- from the same flower (n = 548); self and exogenous
moved stigma and style from flowers and preserved pollen were collected in plastic cups one hour after
them in 70% ethanol. Once in the laboratory, stigmas flower anthesis and hand pollinations were performed
541
© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at
saturating the stigma with pollen (approximately 500 was possible to observe that 21% of trees never flow-
pollengrainsperstigma)usingasoftbrush.Aftereach ered,39%floweredonce,28%floweredtwiceand12%
treatment, marked flowers were bagged again and fol- flowered during three consecutive years (Table 1). In
loweduntilfruitmaturationorfruitabortionoccurs.In addition, for the same 68 trees, 31% never produced
addition, to analyse the reproductive success for this mature fruit, 48% fruited one year, 18% fruited two
species under natural conditions, we selected and years and only 3% were able to produce mature fruit
marked 798 flowers in 5 different branches in one C. duringthreeconsecutiveyears(Table1).
pentandratree.Theseflowersweremonitoreduntilfruit
Inregardstoreproductivesuccess,profuseflowering
productionorfruitabortion.
episodes in many C. pentandra trees were followed by
littleornofruitproduction.Inthesecases,theflower-
Floralvisitors
ingprocesswasfollowedbytheabortionofflowerbuds
Inordertodeterminatethediversity,abundanceand
anddevelopingfruits.Thisphenomenonoccurredboth
activityoffloralvisitorstofloweringC.pentandratrees,
inisolatedtreesandtreeslocatedinareasofcontinuous
we placed two 12-meter mist nets in a natural corridor
naturalforest.Totalabortionofflowerbudsoccurredin
near a flowering tree. We performed this methodology
15% of the population in 2001, 22% in 2002 and was
forthreedifferenttrees.Eachcapturedbatwasidentified
mostevidentin2003,when40%oftreesthatflowered
tospeciesusingthekeyofTIMM&LAVAL(1998).Pollen
abortedalltheflowerbuds.
sampleswerecollectedfrombodiesandfacesforallfru-
givorous and nectarivorous bat species captured using Fivehundredandfiftyfiveflowerbudsand280de-
cubesofjellystainedwithfucsine(BEATTIE1971). velopingfruitsabortedfromC.pentandratreeswerecol-
lectedfromthesoilandinspectedlookingforinsectlar-
vae.Fromthissample,53%offlowerbudsand14%of
Results
developingfruitscontainedcurculionidbeetlelarvaein
Phenologicalpatternsandreproductivesuccess differentdevelopmentalstages.ThebeetlespeciesLon-
AllC.pentandratreesstudiedlosttheirleavesdur- chophorusfusiformis(SubfamilyAnthonominae)wason-
ing the dry season, in December and January. Leaf fall ly found in closed flower buds and L. santarosae was
activityreacheditspeakinthebeginningofJanuary.By foundindevelopingfruits(FERNÁNDEZetal.2008).
the beginning of March, most trees had flushed large
Pollengerminationandpollinationexperiments
quantitiesofnewleavesandbyApril,alltreeswerein
fullleaf(Fig.2). Pollen germination experiments demonstrated that
pollen grains were able to continue germinating, al-
Floweringwasalsorestrictedtothedryseason(Fig.
thoughinlowquantities,until60hoursaftertheywere
2).Fortheyears2001and2002,floweringactivitystart-
removedfromflowers(Fig.3).Itwaspossibletoobserve
edinthebeginningofJanuary.In2003,however,trees
thatduringthefirst14hoursafterpollengrainswerere-
bloomedprematurelyandflowerswerepresentonsome
movedfromflowers,thegerminationratewasrelatively
trees in early December. The flowering peaks for 2001
high with values around 60% to 70%. However, when
and2002occurredattheendofJanuary,butfor2003oc-
pollen removal time increased, we observed a decrease
curredduringthefirsttwoweeksofJanuary(Fig.3).This
inthegerminationratethatreachedvaluesnearzeroaf-
year most trees flowered profusely, but it was not fol-
ter72hours(Fig.3).
lowed by profuse fruiting. Seventy percent of the trees
thatbloomedthatyearabortedalltheflowers.Thedu- The abundance of pollen grains in stigma and
rationoffloweringwasapproximately70days.Inregards pollentubesattheendofthestyleinflowersfrompol-
tothefruitingperiod,in2001and2002,theproduction lination experiments (self, cross and natural pollina-
offruitbeganduringthefinaltwoweeksofJanuary,and tions),isshowninFigure4.Flowersgivennaturaland
fruiting peaked in March. For 2003, fruiting began in cross pollination treatments showed high and similar
Januaryasaresultofearlyfloweringactivity(Fig.2). mean numbers of pollen grains germinated in stigma
area. Also, when we analysed the number of pollen
The percentage of individuals that flowered and
tubesattheendofthestyle,wefoundthattheirabun-
fruitedvarieddependingontheyear.Forinstance,the
danceinflowersfromself-pollination,crosspollination
lowestreproductivesuccesswasobservedin2002,when
andnaturalpollinationtreatmentswasalsoverysimilar
only19%ofthetreesproducedmaturefruit.Thegreat-
intermsofmeanandstandarddeviation(Fig.4).
est reproductive success was recorded in 2003, when
40%oftreesproducedmaturefruit.Inaddition,irregu- Finally,fruitsetobtainedforpollinationtreatments
larity in flowering and fruiting cycles was observed for showed that 25% of flowers on which we performed
individualtrees.For68treesmonitoredover3years,it cross-pollinationsetfruit.Contrarily,flowerswhichun-
542
© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at
derwentself-pollinationdidnotproducefruit.Fruitset
of10%wasobtainedforflowerswithnaturalpollination
treatment.
Floralvisitors
Thefollowingbatspecieswerecapturedinmistnest
carrying C. pentandra pollen: Glossophaga soricina (n =
12,Fig.4),Urodermabilobatum(n=9),Artibeusphaeo-
tis (n = 2) andCarolia perspicillata (n = 2). The great
number of bats flying around flowering trees which is
commonly observed for trees of this species located in
thedryforestsofGuanacaste,NWCostaRicawasnot
reflectedinourstudy.
Discussion
WeobservedthatphenologicalcyclesofCeibapen-
Fig.3:PollengerminationrateforCeibapentandra.Datawerecollectedfrom
tandraareirregularandshowedgreatvariabilityinflow- twotreesandgerminationrateswereobtainedatdifferenttimesafterflower
ering and fruiting intensity among individuals and re- anthesis.
productive episodes, in accordance with previous de-
scriptionsforthisspeciesintheliterature(BAKER1983, Inthisspecies,leafabscissiondatewasrelativelyin-
FRANKIE et al. 1974, MURAWSKI & HAMRICK 1992,
variablefromoneyeartothenext,whileflowerinitia-
NEWSTROM et al. 1994). Based on our data, we found
tiondatewasmorevariable.Theonsetofleafabscission
that some individuals reproduced every year, while
variedwithinaweek,whiletheonsetoffloweringvar-
many others showed a variable duration between each
ied within three weeks across the three years of study.
floweringandfruitingepisode.MostC.pentandratrees
Therefore, periods between leaf loss and flowering ac-
lost and replaced their leaves every year. The produc-
tivityarevariable,andmightbedeterminedbythepres-
tionofflowersandfruit,however,werenotasregularas
leaf replacement. Although complete leaf abscission is ence of complementary abiotic signs that trigger the
necessary for the start of flowering activity, the loss of maturation of flower meristems. We found that timing
leaves does not necessarily predict the production of and intensity of flowering activity were associated in
flowersonC.pentandratrees(ROJAS-SANDOVAL2004). oneofthethreeyearsofstudy.In2003,whentherewas
Fig.4:Numberof
pollengrainsor
pollentubesin
flowersfromdifferent
Ceibapentandra
trees.Eachcolumn
representsthemean
ofpollengrains
(stigma)ornumberof
pollentubes(bottom
style)in20flowers
fromdifferent
pollinationtreatments
(self-pollination,
outcrossingornatural
pollination).
Horizontalaxis
numbersindicatethe
identityoftrees
wherepollination
treatmentswere
performed.
543
© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at
ofself-pollenandgrowthofpollentubes,asissuggested
byGRIBELetal.(1999).However,ourresultsshowthat
fruit set for flowers with self-pollination treatment is
very low or absent, whereas fruit set for flowers with
cross-pollinationtreatmentisabout25%.Theseresults
aresimilartoresultsobtainedbyGRIBELetal.(1999)for
C. pentandra trees in Central Amazonia. This species
seems to have a self pollen rejection mechanism at
ovary level or a high mortality of self-pollinated em-
bryos,andthetwopossibilitiesaredifficulttoseparate.
Ontheotherhand,ourresultsuggeststhatfruitproduc-
tioninC.pentandraispartiallylimitedforthearrivalof
exogenous pollen, due to differences in fruit set for
cross-pollinatedflowersandnaturallypollinatedflowers
(10%), and the copious quantity of pollen grains ob-
servedinflowerswithnaturalpollinationtreatment.
Ourstudyshowsthatpollenforthisspeciesisviable
forseveralnightsandthisissuemaypromoteoutcross-
Fig.5:GlossophagasoricinabatcapturedinamistnetlocatednearaCeiba
ingbypollencarry-over,sincepollenwouldremainvi-
pentandratreeinPalmarNorte,SEofCostaRica.Thepresenceofabundant
ableinbats’furforfuturevisitstoothertrees.However,
yellowpollentypicalofthistreespeciesisobservedinface,chestandwing.
itisnecessarytostudytheeffectofgroomingbehaviour
anadvanceinthedateoftheonsetofflowering,there onthepossibilitythatpollenpresentinbats’furduring
onenightbeavailableforfloralfecundationsonsubse-
werealsohighlevelsofflowerproduction.
quentnights.
Althoughthetimingofeventsishighlyregularand
In the dry forest of Guanacaste (Costa Rica), bat
synchronised at the population level for C. pentandra,
species that have been observed visiting C. pentandra
individual phenology is a very irregular process. Al-
treesare:Glossophagasoricina,Glossophagaleachii,Phyl-
thoughtreesweresubjecttoverysimilarenvironmental
lostomus discolor, Phyllostomus hastatus, Artibeus ja-
conditionsofhumidityandlight,theirphenologicalbe-
maicensis, Artibeus phaeotis, Artibeus intermedius and
haviours were often completely different (Figure 1).
These observations suggest that the phenological be- Sturniralilium(HEITHAUSetal.1975,LOBOetal.2005).
Likewise, the list of bat species that visit C. pentandra
haviourofC.pentandramayberegulatedbyendogenous
flowersinthewetforestofSECostaRicaincludenec-
orinternalcycles,whichvaryamongindividualswithin
tarivorous,frugivorousandgeneralistspeciesthatbene-
the same population of trees. Years with exceptionally
fitfromthisresourceduringthedryseason.However,in
highfloweringactivitydidnottranslateintoyearswith
theOsaPeninsula,wedidnotobserveagreatnumberof
great reproductive output, because there was massive
batsoflargesize(i.e.,P.discolor)flyingaroundtreesand
abortionofflowerbudsanddevelopingfruit.
visiting flowers almost all night long, as is common to
Ourresultsshowedthatahighproportionofabort- observefortreesofthisspecieslocatedinthedryforests
ed flower buds and developing fruits were attacked by ofGuanacaste(LOBOetal.2005).Theseresultsmaybe
different species of curculionid beetles. Indeed, high explainedbydifferencesinrelativehumidityforthedry
ratesofabortioncausedbyinsectsmaynotonlyhelpus seasonbetweenGuanacasteandtheOsaPeninsula,that
toexplaintheeventsinwhichtreeslosealltheirflower convertafloweringepisodeofC.pentandraintoamore
budsanddevelopingfruitwithinoneyear,butcouldal- attractiveeventinthedryforestthaninthewetforest.
so be a selective pressure for the evolution of massive Inadditiontobatspecies,weobservedkinkajous(Potus
anddiscontinuousfloweringovertheyears.Wepropose flavus)visitingandfeedingpollenandnectarinC.pen-
thattheproductionofasurplusofflowerbudsandde- trandra trees located at the Corcovado National Park
velopingfruitsinC.pentandracanbeanadaptationto andnearPalmarNorteTown(ROJAS-SANDOVAL2004).
satiatepredatorsofreproductiveorgans(JANZEN1971).
Thesimilarnumberofpollentubesinflowerswith
self, cross and natural-pollination treatments suggests
that for this species, incompatibility barriers are not
present at stigma-style levels. As a result, there is no
mechanismattheselevelsthatcaninhibitgermination
544
© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at
References
BAKER H.G. (1983): Ceiba pentandra. — In: JANZEN D.H. (ed.),
Costa Rican Natural History. University of Chicago Press:
212-214.
BEATTIE A.J. (1971): A technique for the study of insect-borne
pollen.—PanPacificEntomologist47:82.
BREWBAKERJ.L.&B.H.KWACK(1963):Theessentialroleofcalcium
ioninpollengerminationandpollentubegrowth.—Am.
J.Bot.50:859-865.
CASCANTE-MARIN A. (1997): La Familia Bombacaceae (Malvales)
enCostaRica.—Brenesia48:17-36.
FERNÁNDEZM.,LOBOJ.,CHACÓNE.&M.QUESADA(2008):Curculion-
idbeetlesinabortedflowerbudsandimmaturefruitsof
Ceibapentandra.—PlantEcology194:1-4.
FOURNIERL.A.(1974):Métodocuantitativoparalamediciónde
característicasfenológicasenárboles.—Turrialba24:422-
423.
FRANKIEG.W.,BAKERH.G.&P.A.OPLER(1974):Comparativephe-
nologicalstudiesoftreesintropicalwetanddryforestsin
thelowlandsofCostaRica.—J.Ecol.62:881-919.
GOMEZL.D.&W.HERRERA(1985):VegetaciónyClimadeCostaRi-
ca.—EUNED.SanJose,CostaRica.
GRIBELR.,GIBBSP.E.&A.L.QUEIROZ(1999):Floweringphenology
and pollination biology of Ceiba pentandra (Bomba-
caceae)inCentralAmazonia.—J.Trop.Ecol.15:247-263.
Addressesofauthors:
HARTSHORN G.S. (1983): Plants: Introduction. — In: JANZEN D.H.
(ed.), Costa Rican Natural History. University of Chicago JulissaROJAS-SANDOVAL
Press:118-157. DepartmentofBiology
HEITHAUSE.R.,FLEMINGT.H.&P.A.OPLER(1975):Foragingpatterns UniversityofPuertoRico
andresourceutilizationinsevenspeciesofbatsinasea- RíoPiedrasCampus
sonaltropicalforest.—Ecology56:841-854. P.O.Box23360
JANZEN D.H. (1971). Seed predation by animals. — Ann. Rev. SanJuan
Ecol.Syst.2:465-492. PuertoRico
LOBOJ.A.,QUESADAM.&K.E.STONER(2005):Effectsofpollination USA
bybatsonthematingsystemofCeibapentandra(Bomba-
E-mail:[email protected]
caceae)populationsintwotropicallifezonesinCostaRica.
—Am.J.Bot.92:370-376. KatharinaBUDDE
LOBOJ.A.,QUESADAM.,STONERK.E.,FUCHSE.J.,HERRERIAS-DIEGOY., MauricioFERNÁNDEZ
ROJASJ.&G.SABORIO(2003):Factorsaffectingphenological JorgeA.LOBO
patternsofBombacaceoustreesinseasonalforestsinsea-
EscueladeBiología
sonalforestsinCostaRicaandMexico.—Am.J.Bot.90:
1054-1063. UniversidaddeCostaRica
SanJosé,CostaRica
MURAWSKID.A.&J.L.HAMRICK(1992):Matingsystemandphenol-
ogyofCeibapentandra(Bombacaceae)inCentralPanama. E-mail:[email protected]
—J.Hered.83:401-404. [email protected]
NEWSTROML.E.,FRANKIEG.W.&H.G.BAKER(1994):Anewclassifi- [email protected]
cationforplantphenologybasedonfloweringpatternsin
lowlandtropicalrainforesttreesatLaSelva,CostaRica.— EduardoCHACÓN
Biotropica26:141-159. CiudaddelNeotrópicoHúmedo
ROJAS-SANDOVALJ.(2004):BiologíaReproductivayFenologíade Golfito
Ceibapentandra(Bombacaceae)endoszonasdevidade UniversidaddeCostaRica
CostaRica.—TesisdeMaestría.UniversidaddeCostaRica. CostaRica
TIMMR.M.&R.K.LAVAL(1998):AfieldkeytothebatsofCosta E-mail:[email protected]
Rica. — Occasional Publication Series 22: 1-32. Center of
Latin American Studies, The University of Kansas, MauricioQUESADA
Lawrence,Kansas,USA. CentrodeInvestigaciónesenEcosistemas
UniversidadNacionalAutónomadeMéxico
México
E-mail:[email protected]
545