148 AROIDEANA, Vol. 29 New Evidence of Pollination in Gearum brasiliense (-Spathicarpeae)

Eduardo G. Gon~a1ves Universidade CatOlica de Brasilia Predio Sao Gaspar Bertoni, sala M-206, QS7, Lote 1, EPfC CEP 72030·170 Taguatinga-DF-Brazll Artur Campos DaJia Maia Universidade Federal de Pernambuco Centro de Ci~ncias Biol6gicas, Departamento de Zoologia. Av. Prof. Moraes Rego, s/n---Cidade Universitiiria CEP 50670-901 Recife, p~Brasn

ABSTRACT sidered as belonging to their own tribe Dieffenbachieae are now considered New evidence is presented to suggest Spathicarpeae too (French et al., 1995; that the monotypic Gearnm (Ara­ Keating, 2004; Gon~alves et al., in prep.). ceae) may be truly cantharophilous, not Very recently, two new genera (Croatiella myophilous as suggested before. The E. G. Gon~. and Incarnm E. G. Gon~.) credible pollinators in Gearnm brasiliense were added to the tribe (Gon~alves, 2005). are large scarab beetles of the species Cy­ Bogner & Gon~alves (1999) pointed out clocepbala celata, which were collected that the pollinators in Gearnm brasiliense inside floral chambers of inflorescences could be flies of the genus Beebeomyia between the female and the male phases. (Richardiidae). These conclusions were Along with the direct observations of in­ based on Bogner's findings of larvae and sects within inflorescences, general floral pupae of these insects lying at the bottom morphology and construction are used as of Gearnm inflorescences. In fact, these indirect evidences of a cantharophily pol­ flies are quite common in inflorescences lination syndrome in this aroid genus. of Gearnm and other genera of Spathicar­ peae such as Taccarnm (Maia and collab­ KEYWORDS orators, in prep.) and Dieffenbacbia Beetle pollination, Spathicarpeae, Gear­ (Young, 1986), but they were never reput­ um, cantharophily, Cyclocepbala ed as effective pollinators. The suggested fly pollination in Gearnm brasiliense could be judged an idiosyncratic feature INTRODUCTION when considering the number of morpho­ The monotypic aroid genus Gearnm logical similarities between the inflores­ N.E.Br. belongs to the tribe Spathicarpeae cences of Gearnm and other confirmed and occurs in Central Brazil. It is the only cantharophilous groups, such as Dieffen­ genus that can be considered endemic to bacbia (Young, 1986, Beath, 1999), Xan­ the Cerrado biome (Gon~alves, 2004) and tbosoma (Garda-Robledo et al., 2004), it is rather common locally in southern To­ Montricbardia (Gibernau et al., 2003) and cantins State. Pbilodendron (Gottsberger & Amaral, The tribe Spathicarpeae traditionally 1984; Gibernau et al., 1999). comprises eight genera (Mayo et al., 1997) In November of 2005, the finding of a but the genera Dieffenbacbia Schott and single Cyclocepbala species in two inflo­ Bognera Mayo & Nicolson, formerly con- rescences in the field reinforced the hy- E. G. GO <;:ALVES, A. c. DALlA MAlA, 2006 149

Fig. 1. Beetle captured inside the fl o ral chamber of GearU1?1 brasiliense in Tocan­ tins state . (Photo: E. Gons;:a lves)

pothesis that the main pollinato rs in G. bras'iliense could be scarab beetles and that the Richardi idae fli es are onl y inflo­ rescence predators. Arguments for this hy­ pothesis are presented below. Fig. 3. Detail of inflo rescences in Gear­ um bnlsiliense, showing an intact inflores­ cences in fu ll anthesis, female phase (l e ft), an exposed spad ix (middle) and an un­ opened inflo rescence (right). In the ex­ posed spadix, note the female flowe rs sur­ rounded by white staminodes, the ste ril e male po rtion (middle) and the fe rtile male portion (apex).

NEW EVIDENCE Between November 2nd and 3rd , 2005, two visits were conducted to the same site in the vicini ty of the village o f Arraias, southe rn Tocantins state , Central Brazil. The fl owering populati o n was located in an o pen area in a hyperseasonal savanna, mode rately disturbed by cattl e activity. There were at least 50 Geant1n bnlsiliense infl orescences in d ifferent develo pment stages. Only o ne individual had already unfurled (partiall y) its pedate leaf blade. Other ca ntharophilous species fo und at the sa me area were the aroid Xantbosoma Fig. 2. Inflorescence o f Gearum brasi­ striattpes (Kunth) Madison and the Annon­ liense in the beginning of female phase. aceae, Annona wanningiana Mell o-Sil va Note that the inflo rescence arises singly & Pirani. from the peat. A second cl osed inflores­ A moderate thermogenesis (temperature cence can be seen be hind. not measured) was evidenced at ea rl y 150 AROIDEANA, VoL 29 dusk during the female phase, along with are usually strongly thermogenic (Gotts­ the emission of a camphor-like smell by berger & Amaral, 1984; Gottsberger, 1990). the male flowers. Thermogenesis was not Thermogenesis is usually known to occur detected in male phase, but observations at dusk in these species. in male phase were restricted to the morn­ 4. Strong odor emission-Crepuscular/ ing when heating is not usually observed nocturnal emission of odoriferous volatiles in other cantharophilous species. are known for all large scarab pollination Four beetles were found inside the flo­ syndromes and seems to be the key at­ ral chamber of one inflorescence during traction mechanism for these beetles the male phase. Another smaller inflores­ (Gottsberger & Amaral, 1984; Gottsberger, cence at the same phase was also found 1990; Gottsberger & Silberbauer -Gottsber­ with two beetles of the same species. The ger, 1991). These scents are usually asso­ beetles were all identified as Cyclocepba/a ciated with the female phase of anthesis, celata Dechambre 1980 (Scarabaeidae, when they are volatilized at higher spadix Dynastinae, Cyclocephalini), a common temperatures (Gibernau & Barabe, 2002). species in Brazil (Maia and collaborators, 5. Nocturnal anthesis (or more precisely ined. data). In both inflorescences, the crepuscular anthesis)-Flowering in spe­ beetles ate all the staminodes, including cies adapted to be pollinated by Dynasti­ those in synandrodia, produced between nae beetles has to be coordinated with the fertile male flowers and female flowers. time when the beetles are mostly active No damage was observed in female flow­ (Gottsberger, 1999). Pollen release and ers but the fertile male flowers were oc­ thermogenic peaks are usually known to casionally chewed. occur at dusk in these species. The only aspect to be considered as DISCUSSION lacking in the previously described Gear­ um brasiliense system is that the spathe is No controlled test was conducted in or­ not white or whitish green in all individ­ der to confirm positively if Cyclocepbala uals, as is common to other large scarab celata is in fact the main pollinator of pollination systems. Most specimens have Gearum brasiliense. However, consider­ a spathe that is white with dark stripes. ing the sum of morphological aspects, the The inverse also occurs commonly, i.e. cantharophily involving large scarab bee­ dark spathes with whitish stripes. The tles can be easily detected by the follow­ same patterns are also shared with other ing "syndromic features" as suggested by species known to be truly cantharophilous many authors (e.g. Gottsberger, 1999) such as Taccarum caudatum Rusby (G. 1. Large pollination chamber-Scarab Gerlach, pers. observ.) and Taccarum ulei Dynastinae beetles are usually robust and Eng!. & K. Krause (Maia and collaborators, commonly remain for at least 24 hours in prep.). within the inflorescences, so a large polli­ In the tribe Spathicarpeae, only the ge­ nation chamber (spathe tube) is generally nus Dielfenbacbia is well studied regard­ required. ing its pollination (Young, 1986; Beath, 2. Massive quantities of feeding materi­ 1999), and long known to be essentially al-It is widely known that Dynastinae are visited by scarab beetles of the subfamily eager beetles and eat many floral parts Dynastinae (Cyclocepbala and Erioscelis). during their stay in the inflorescences Probable pollinators were also observed in CGotsberger, 1999; Gibernau et al., 1999). the field for the monospecific genus Bog­ Providing edible parts that could feed the nera and they proved to be a still uniden­ beetles and avoid damage to the female tified Dynastinae (E. Gon\;alves, pers. ob­ flowers or fertile male flowers is a vital serv.). Within the core geophytic Spathi­ strategy. carpeae, only Taccarum is well known 3. Easily detected thermogenesis­ concerning its pollination (Maia and col­ pollinated by large scarab beetles laborators, in prep.). Visitation of Droso- E. G. GONC;ALVES, A. C. DALIA MAlA, 2006 151

phyllidae flies were observed in Spathi­ and fruit predation in Xanthosoma carpa and Spathantheum, in the field for daguense (Araceae).]. Trop. Beol. 20: the former and in the cultivation for the 459-469. latter (E. Gon\;alves, pers. observ.). More­ Gibernau, M" D. Barabe, P. Cerdan & A. over it is not expected that pollination in Dejean. 1999. Beetle pollination of the genus Asterostigma or in the recently Philodendron solimoesense (Araceae) described Croatiella or Incarum (Gon\;al­ in French Guiana. Inti.]. . Sci. ves, 2005) should be performed by large 150:1135-1143. scarab beetles, because the floral chamber ---& D. Barabe. 2002 Pollination ecol­ (spathe tube) is too slender in these gen­ ogy of Philodendron squamiferum era. (Araceae). Canad.]. Bot. 80:1-5. Gearum is considered basal to a lineage ---, D. Barabe, D. Labat, P. Cerdan & within the tribe Spathicarpeae (Gon\;alves A. Dejean. 2003. Reproductive biolo­ et al., in prep.). This new evidence rein­ gy of Montrichardia arborescens (Ar­ force the hypothesis of Gon\;alves (2002) aceae) in French Guiana.]. Trop. Ecol that the pollination in basal members of 19:103-107. the tribe Spathicarpeae may be performed Gon\;alves, E. G. 2002. Sistematica e ev­ mainly by Dynastinae beetles and that the olu\;ao da tribo Spathicarpeae (Ara­ use of different strategies of pollinations in ceae). Unpublished Ph.D. Thesis, modern lineages probably evolved in re­ Universidade de Sao Paulo. sponse to the adaptive radiation to more ---. 2004. Araceae from Central Brazil: seasonaVsubtropical environments. Comments on their diversity and bio­ geography. Ann. Missouri Bot. Gard. ACKNOWLEDGMENTS 91:457-463. ---. 2005. Two new Andean genera Field trip was supported by Instituto for the tribe Spathicarpeae (Araceae). Plantarum de Estudos da Flora Ltda. Stud­ Willdenowia. 35:319-326. ies in the tribe Spathicarpeae by one of us Gottsberger, G. 1990. Flowers and beetles (EGG) was sponsored by FAPESP (99/ in the South-American tropics. Botan­ 02921-7) and Funda\;ao Botiinica Margaret ica Acta, 103(4):360-365. Mee. We are grateful to Marc Gibernau for ---. 1999. Pollination and evolution in important suggestions on the manuscript. neotropical Annonaceae. Plant Spe­ cies Biology. 14:143-152. lITERATURE CITED ---& A. Amaral Jr. 1984. Pollination strategies in Brazilian Philodendron Bogner, J. & E. G. Gon\;alves. 1999. The species. Berich. Deustch. Bot. Ger. 97: genus Gearum (Araceae: Tribe Spath­ 391-410. icarpeae). Aroideana. 22:20-29. ---& I Silberbauer-Gottsberger. 1991 Beath, D N. 1999. Dynastine scarab beetle Olfactory and visual attraction of Er­ pollination in longis­ ioscelis emarginata (Cyclocephalini, patha (Araceae) on Barro Colorado Dynastinae) to the inflorescences of Island (Panama) compared with La Philodendron selloum (Araceae). Bio­ Selva biological station (Costa Rica). tropica 23:23-28. Aroideana.22:63-7l. Keating, R. C., 2004. Systematic occurrence French, J. C., M. G. Chung & Y. K.]ur, of raphide crystals in Araceae. Ann. 1995. Chloroplast DNA phylogeny of Missouri Bot. Gard. Ann. Missouri the Ariflorae, pp. 255-275. In P. J. Ru­ Bot. Gard.91:495-504. dall, P. J. Crib, D. F. Cutler & J. Hum­ Mayo, S. J., J. Bogner & P. C. Boyce, 1997. phries (eds.), : Sys­ The Genera ofAraceae. Royal Botanic tematics and Evolution, Royal Botan­ Gardens, Kew. ic Gardens, Kew. Young, H. J. 1986. Beetle pollination of Garda-Robledo, c., G. Kattan, C. Murcia & Dieffenbachia longispatha (Araceae). P. Quintero. 2004. Beetle pollination Amer.]. Bot. 73:931-944.