BLUMEA 39 (1994) 289-294 Notes on the genus Amorphophallus(Araceae)— 5. Amorphophalluskonkanensis, a new species from India, and taxonomic reflections on Amorphophallussection Rhaphiophallus W.L.A. Hetterscheid S.R.Yadav & K.S. Patil Summary Anewspecies ofAmorphophallussect.Rhaphiophallus(Schott)Engl.from SWIndiais described. Themonophylyofandcharacterevolution in sect.Rhaphiophallusis discussed. DESCRIPTION Amorphophallus konkanensis Hett., Yadav & Patil, spec. nov. —Fig. 1 Ab A.mysorensis Barnes etFischer in floribus neutrissemicomplanatis, placentationebasali,et ab A. sylvatico (Roxb.)Kunth in dimensionibus maioribus,spathafusca,placentationebasali differt. —Typus:K.S. Patil4687-A (CAL nolo),India, Maneri,SindhudurgDistrict, 15-v-1992. Para- types: K.S.Patil4687-B(BLAT),4687-C(L),both samedataas holotype. Tuberousherb. Tuber globose or depressed globose, 3-8cm in diam., 3-4.5 cm high,producing short, rhizomatousoffsets.Lea/solitary; petiole smooth, 29-88cm long, 0.6-1.5 cm in diam., brown or greenish brown, mottledpinkish and with whitish stripes; lamina40-96cm in diam.,rhachises winged, except forthemost proximal parts; leaflets lanceolate,acuminate,4-19cmlong, 1-4cm indiam.Inflo- rescence long-peduncled; peduncle as petiole, 25-55cm long,0.3-1 cm in diam.; spatheerect, ovate,acute,notconstricted, limbpoorlydifferentiatedfrombase, 3.3- 8.5 cm long,2.3-7cm in diam.,outsidedirtypinkish withabrownishhue andfaint brownishspots, veinsdarkpurplish brown, insidemaroon,basewithindarkmaroon, longitudinally ridged. Spadix stipitate, up to twice as long as spathe, 9.5-16 cm long; stipe c.0.5 cm long, green;female zone cylindric, 0.8-1.7 cm long, flowers congested; malezone cylindric, 1.8-3.1 cm long, flowersslightly distant; stamino- dialzone between femaleand male zone 0.6-1.5cm long, staminodescongested; U The firstthree papers in thisserieswerepublishedin,respectively, Blumea 36 (1991)467-475 and 39,thepresentissue,pages 237-281 and 283-287;No. 4(byC.I.Peng) willbe publish- edin Bot.Bull. Acad. Sinica, 1995. 2) Chrysantenstraat28, 1214BM Hilversum,TheNetherlands. 3) DepartmentofBotany,GoaUniversity,TaleigaoPlateau,P.O.Bambolim Complex,Goa404 203, India. 4) DepartmentofBotany,Shivaji University,Kolhapur416004,Maharashtra State,India. 290 BLUMEA Vol. 39, No. 1/2, 1994 Fig. 1.—a, b: Amorphophalluskonkanensis Hett., Yadav & Patil; a: inflorescence,c. x0.7; b: lower halfofspadix, c.x 1.7. appendix elongate conic,apex blunt, 5.5-13cm long,0.5-1cm indiam., dirtyolive green.Ovaries depressed-globose, diamond-shaped incross section, 2-3 mm in diam.,0.8-1 mm high,pale green,nearthetop becoming purplish, 3- or4-locular, one basalovule perlocule; style short, cylindric orconic, 0.2-0.5 mm long, 1 mm in diam.(base), purplish; stigma large, subcircularor slightly irregular incross sec- tion,2mm in diam.,0.7-1 mmhigh, verruculate, paleyellow, 3-or 4-lobed,lobes very shallowly conical or rounded, obtuse, separated by shallow grooves. Male flowers consisting of4-6 stamens;stamens 1mm long;filaments0.2-0.3mmlong, entirely connate; antherstruncate, 0.7-0.8mm long, 1-1.5 mm in diam., irregular in cross section, thecae whitish, connectivebrownish, poresapical, elongate, reni- form or straight. Pollenpsilate. Staminodes ovate or rhomboidin cross section, W.L.A. Hetterscheid, S.R. Yadav & K.S. Patil: Notes onAmorphophallus-5 291 slightly convex, 3-6x 2.5-4mm, 1-3 mm high, whitishor with afaintpurplish hue. Berries pink at maturity, 2-4-seeded; seeds subglobose, 0.3-0.5 cm long, 0.2-0.35cm in diam.2n= 26. Etymology - Thespecies epithet referstotheKonkanregion inthe south-western partof MaharashtraState, wherethisspecies is found. Distribution India,SW MaharashtraState(Konkan region). - Habitat&Ecology - Commonthroughout theKonkanforestonlateritesoils, near bushes. Fl. April-May; fr. May-June. NOTES ON MONOPHYLY AND CHARACTER EVOLUTION Amorphophallus konkanensis is clearly amemberofAmorphophallus sect. Rhaphio- phallus (Schott) Engl. (Engler, 1911) (emend. Sivadasan, 1989), possessing the characteristic spathe-shape and staminodes. The general morphology and colour- pattern ofA. konkanensis are very similartoA. mysorensis Barnes& Fischer from the state of Mysore (India), fromwhich it differs markedly in the semi-flattened staminodes(vs. globose inA. mysorensis)and the basal placentation (vs. axillary halfway up thelength ofthe locule inA. mysorensis). Threeother species ofsect. Rhaphiophallus possess similarly flattenedstaminodesas inA. konkanensis, viz. A. bonaccordensisSivad. & Mohanan,A. hohenackeri(Schott) Engl, and,A. sylvaticus (Roxb.) Kunth, allfromsouthernIndia.Ofthese, A. sylvaticus hasanaxile placen- tationas inA. mysorensis and muchsmallerdimensions,whereasA. bonaccorden- sisand.A.hohenackeridifferinhaving muchbroaderleaflets, very long, thin stolons andthespadix equal toor shorterthanthespathe. SectionRhaphiophallus, ashereunderstood, contains eight species whichare sepa- rated as a groupby Sivadasan(1989) withinAmorphophallus onthe basis ofthe large, flattenedor globose neuter organsbetweenthemaleand femalezones. How- ever, contrary to Sivadasan's statement thatonly species ofsect.Rhaphiophallus possesses such neuter organs, thereare threemore published species with similar structures not belonging to this section, viz. A. krausei Engl. [syn. A. sutepensis Gagnep., A. ximengensis H. Li, from northernBurma, northern Thailand,China (Yunnan)], A. saraburiensis Gagnep. (central Thailand) andA. albus P.Y. Liu & . J.F.Chen(China, Szechuan). Anadditional two new species (A. atroviridis Hett., ThailandandA. salmoneusHett. [Philippines (Palawan)], published inthe present issue, also sharethis character. As a consequence,this charactercannot be useda priori to distinguish sect. Rhaphiophallus from theremainderofAmorphophallus, without considering the distributionofthis and othercharacters ofthe sectionin otherspecies of thegenus(see Table 1on thenext page). Theaxileposition halfway upthelength oftheplacenta oftheovulesinthreespecies ofsect.Rhaphiophallus is unique inthegenusbutthis doesnot suffice tosupportthe monophyly oftheentiresection. Thesameholds forthe globose staminodes,which arealso only foundinthreespecies of thesection.Thereremainsthestipitate spadix, a fairlyrare characterinthegenus. Itis alsofoundinanumberofHimalayan-Indo- chinese species, e.g.A. yunnanensis Engl. (syn. A. kerrii Gagnep.) from Thailand, 292 BLUMEA Vol. 39,No. 1/2, 1994 Table 1.CharacterdistributioninAmorphophallus sect.Rhaphiophallus compared to A. ‘groupyunnanensis’ andA. ‘group krausei’. \\ cchhaarraacctteerrss:: 11 22 33 44 55 ssppeecciieess \ AA..bboonnaaccccoorrddeennssiiss SSiivvaadd..&& MMoohhaannaann aa aa aa aa bb AA.. hhoohheennaacckkeerrii((SScchhootttt)) EEnnggll.. aa aa aa aa bb AA..kkoonnkkaanneennssiiss HHeetttt..,, YYaaddaavv&& PPaattiill aa aa aa aa bb AA.. lbonnggiiccoonnnneeccttiivvuuss BBooggnneerr bb ?? ?? ?? bb AA.. mmaarrggaarriittiiffeerr((RRooxxbb..)) KKuunntthh bb bb aa aa bb AA.. mmyyssoorreennssiiss BBaarrnneess && FFiisscchheerr aa bb bb aa bb AA..nniiccoollssoonniiaannuussSSiivvaadd.. aa bb bb aa bb AA.. ssyyllvvaattiiccuuss ((RRooxxbb..)) KKuunntthh aa aa bb aa bb AA.. ''ggrroouupp yyuunnnnaanneennssiiss’' aa -- bb bb bb AA.. ''ggrroouupp kkrraauusseeVi’ aa aa aa aa aa CChhaarraacctteerrss:: llaa:: aappppeennddiixx pprreesseenntt 44aa:: nneeuutteerroorrggaannsspprreesseenntt bb:: aappppeennddiixx aabbsseennttoorr ssttuubb bb:: nneeuutteerroorrggaannssaabbsseenntt 22aa:: nneeuutteerroorrggaannssffllaatttteenneedd 55aa:: ssppaaddiixxsseessssiillee bb:: nneeuutteerroorrggaannssgglloobboossee bb:: ssppaaddiixx ssttiippiittaattee 33aa:: ppllaacceennttaattiioonn bbaassaall bb:: ppllaacceennttaattiioonnaaxxiilleeeettcc.. ((sseeee tteexxtt)) Laos, China(Yunnan) and N Vietnam, A. corrugatusN.E.Brown from northern Burmaandnorthern Thailand, A. kachinensisEngl. & Gehrm. (syn.A. bannaensis H. Li) from north-eastIndia, northernBurma, northernThailand, China(Yunnan) andLaos.Ifthis characteris anapomorphy, linking thesespecies to sect.Rhaphio- phallus, thespecies of thelattersectionwouldstillcome out as a monophyletic group on thebasis ofthestaminodes. Allotherspecies withstaminodeshavesessile spa- dices, so,even inthatframe, sect.Rhaphiophallus wouldprobably stillseparateas a monophyletic group.Thespathe-shape ofalleight species is alsoquiterare inthe genusand occurs only in species whichshow nootherrelevantsimilarities to spe- ciesofsect.Rhaphiophallus andare thereforeconsiderednotto belong toit (e.g.A. elliottiiN.E. Brown,.A. interruptus Engl. & Gehrm.). On the basisofthese specu- lations and awaiting a phylogenetic analysis oftheentire genus(Hetterscheid, in prep.), itseems fairto considersect.Rhaphiophallus monophyletic. Thisprovides theopportunity to discuss comparatively some charactersofitsspecies. Theposition ofA. longiconnectivus Bogner (ined.) insect.Rhaphiophallus is uncertainatthemo- mentbecause ofdifficultiesininterpreting some charactersoftheonly knownspeci- men. Evenitsstatus asspecies in relationtoitsnearest morphological allyA.marga- ritifer (Roxb.) Kunthis under discussion. Apreliminary cladisticanalysisofsect. Rhaphiophallus using allpossible alternativestatesofsome charactersofA.longi- connectivus andentering itas a separatespecies, indicate itis thesister species of A. margaritifer. W.L.A. Hetterscheid, S.R. Yadav & K.S. Patil: Notes on Amorphophallus-5 293 A geographic trendhas beensuggested (pers. comm. J. Bogner, Munich, Germany) inthissection ofareductionoftheappendix inspecies going from Southto North Indiabutthis trendis brokenby theoccurrence ofA. konkanensisnorth ofthedis- tributionofA. mysorensis. This latterspecies shows a marked variationin the ap- pendix-length whichhasbeeninterpreted as beingintermediatebetweenthesituation in, on theone hand,A. sylvaticus andA. nicolsonianusSivad. (respectively S and SWIndia) possessing along appendix and, onthe otherhand, A. margaritiferand A.longiconnectivus (northern India) bothlacking an appendix or the appendix being reducedto amerestub. Thischaracter ofbothA. margaritiferandA. longiconnec- tivusmustbeconsideredaderivedcharacterwithintheframeworkofsect. Rhaphio- phallus, sinceitoccurs in speciesofamonophyletic 'subunit'ofsect. Rhaphiophal- lusmadeupofallspecies withglobose staminodes(a uniquecharacterinAraceae!). Thelackofanakedappendix maybetheresultoftwo alternative evolutionary path- ways, onebeing thedifferentiationoffertilemaleflowersontheappendix, theother being anactualsuppression oftheappendix. Twocluespoint to thelatteralternative. First, A. mysorensis shows markedvariationin appendix-length, whereby the ap- pendix maybe 'reduced' to amere shortstub. Atleast one specimen isknown of A. margaritifer, showing a similarstub. Second,in allspecies of thegroupwith globose staminodes which possess anappendix themalezonereaches to thetop of the spathe. InbothA. margaritiferandA. longiconnectivus this situationis similar. If, on a non-suppressed appendix male flowers would have differentiated, this wouldhaveresulted ina verylong malezone reaching farbeyond thespathe, as for the 'malezone +appendix' of theotherspecies. Thereforeit is concludedthatthe lack ofa nakedappendix in bothspecies discussed resulted from asuppression of the entireappendix. Theotherexplanation (male flowersdifferentiatedon the ap- pendix) is considered by Hetterscheid (in prep.) to be the cause ofthe lack ofa smoothappendix inA. coudercii(Bogner) Bogner (Laos & Vietnam) andmayalso explain thestaminodialappendix in.A. napalensis (Wall.)Bogner & Mayo (Bhutan, Nepal, NIndia) andA. sumawongii (Bogner) Bogner & Mayo (Thailand). Theinterpretation ofthesterilestructuresbetweenmaleandfemalezones as stami- nodesisbased onobservationsofserial homology betweenthemaleflowers andthe sterilestructures. Thedevelopment ofthis typeofstaminodemayhaveoccurred more thanonce inAmorphophallus. They arealso foundina groupofmainly Himalayan- Indochinesespecies (seeabove) and inA. atroviridis Hett., amemberofa groupof long-tubered species (see the description inthepresentissue,page245). The phylo- genetic interrelationships ofthesegroupsare unclearatthemoment,so anyspecula- tionontheamount ofhomoplasy intheoccurrence ofthesestaminodesin thegenus is futile.However, morphological investigation by the first authorinto allspecies possessing such staminodesindicatesthatthey are derivedin asimilarmannerfrom themaleflowers (assuming thatinall thesecases the presence ofstaminodesis a derivedcharacter). In anumberofthese species thefilamentsofat leastthe lower maleflowers are entirely connate andtheresulting column('cushion' in Sivadasan etal., 1994) is often enlarged in the lowermostflowers. Flowers morphologically intermediatebetweenthelatterandthestaminodesareoftenfoundandthey indicate thatthetransitionisbrought aboutby areductionofthethecae,leaving only thecol- 294 BLUMEA Vol. 39, No. 1/2, 1994 umnandtheconnective-tissue as theactualstaminode.Inthegroupoflong-tubered species this same derivationcan bebased by invoking topographic homology. Inall species ofthis groupthecolumnin thelowermostmaleflowers is extremely large, reducing theantherstomerestubson topofitandoftenpushing themapart, thereby resembling thestaminodialcondition.Thetransition to staminodesinthis groupis then 'completed' inoneofitsspecies, viz.A. atroviridis. Variationinthesculpture ofthepollen wallinsect.Rhaphiophallus is remarkablebe- ing psilate (A. konkanensis, A. margaritiferandA. mysorensis), finely verrucate (A. sylvaticus) orstriate(A. nicolsonianus). Palynological investigations (VanderHam & Hetterscheid, inprep.) show thatthis variationis muchwider thanin othermono- phyletic species groupsinAmorphophallus. ACKNOWLEDGEMENTS Thanks are duetoDr. Jin Murata(BotanicGarden, UniversityofTokyo) and V.T.Chougule(prin- cipal ofSmt.K.W. College, Sangli)forencouragement,to Dr. S.M. Patil (head,Departmentof Botany, Shivaji University) forprovidingnecessary facilities, Dr. M. Sivadasan (Department of Botany, Calicut University) for valuable comments, Dr. A.Hay (Sydney) andMr. P.C. Boyce (Kew)for correctingthe manuscript,Dr. J.F. Veldkamp (L)forproviding theLatin diagnosis, and Dr. R.W.J.M.van derHam (L)forprovidingthepollendata. REFERENCES Engler,A.1911.Araceae-Lasioideae. In:A.Engler(ed.),Das Pflanzenreich IV.23C,Heft48.Leip- zig. Sivadasan,M. 1989.Amorphophallusnicolsonianus (Araceae),anewspeciesfrom India,andanote onA. sect. Synantherias.Willdenowia 18: 435-440. Sivadasan,M.,N. Mohanan &G.Rajkumar. 1994.Amorphophallusbonaccordensis,anewspecies ofAraceae fromIndia. Blumea 39: 295-299 (thisissue).