40 AROIDEANA, Vol. 35 The End of N.E. Br.

Wilbert L. A. Hetterscheid Von Gimborn Arboretum Velperengh 13 3941 BZ Doorn, The Netherlands [email protected] (*31)318418187 Cyrille Claudel Biocentre Klein Flottbek and Botanical Garden Hamburg University Ohnhorststrasse 18 22609 Hamburg, Germany

ABSTRACT soon be transferred to , this new species, clearly a member of The aroid Pseudodracontium N.E. ‘‘Pseudodracontium’’, was published as a Br. is reduced to Amorphophallus Bl. ex species of Amorphophallus. Decne. New names and a new key to taxa Pseudodracontium has always been of the former Pseudodracontium are pre- thought to be very closely related to sented. A discussion on its phylogenetic Amorphophallus and both genera solely position within Amorphophallus is given. make up the tribe Thomsonieae. Thomso- nieae is now regarded to be the sister group KEY WORDS of Caladieae (incl. Zomicarpeae; Cabrera et al., 2008; Cusimano et al., 2011). Amorphophallus, Pseudodracontium,tax- onomy, Thomsonieae, , phylogeny. PSEUDODRACONTIUM AND AMORPHOPHALLUS INTRODUCTION Hetterscheid (1994) already suggested that The genus Pseudodracontium was es- Pseudodracontium may well be a part of tablished by N.E. Brown in 1882. He Amorphophallus instead of its sister genus. introduced one new species, P. anomalum This was based on morphological observa- N.E. Br., and transferred Amorphophallus tions which resulted in a small suite of lacourii Lind. & Andre´toPseudodracon- apomorphies for Amorphophallus + Pseudo- tium. The first mentioned species was , whereas without Pseudodra- chosen as the lectotype species by Nicolson contium there was no single character to be (1967). Subsequent years saw the addition found for a monophyletic Amorphophallus of several new names in Pseudodracon- (Hetterscheid, 1994; Serebryanyi, 1995). tium. These were dealt with taxonomically The monophyly of Pseudodracontium in a revision of the genus by Serebryanyi itself was never disputed and strongly (1995) which left a total of 6 accepted supported by a suite of morphological species, 4 of which were newly introduced (near-) autapomorphies (Serebryanyi, 1995; in that revision. After the revision only one Van der Ham & Van Heuven, 2001; Van new name was published in relation to der Ham et al., 1998, 2000, 2005). The most Pseudodracontium, namely Amorphophal- obvious ones are: unilocular thecae, slender lus glaucophyllus Hett. & Serebryanyi filaments (fully free or more or less connate (Hetterscheid, 2006): Because it was al- to form a slender column, Fig. 1b), appendix ready then strongly suspected by the with recognizable staminodial structure authors that Pseudodracontium would (Fig. 1c), pollen with ‘‘polar caps’’ (smooth

Aroideana arod-35-00-05.3d 3/4/12 10:58:43 40 W. L. A. HETTERSCHEID, C. CLAUDEL, 2012 41

Plate 1. 1a. Amorphophallus lacourii group. Habit. 1b. Amorphophallus lacourii group. Male flowers. 1c. Amorphophallus pseudoharmandii Appendix. 1d. Amorphophallus macrophyllus Pollen with polar caps.

or irregularly structured polar areas, erably less complex than the lateral ones, fig. 1d). and there is a general morphocline from We have always been puzzled by the more complex to rather simply divided retention of seemingly juvenile leaf archi- laminas (or the reverse). In some speci- tecture in many adult of Pseudodra- mens the leaf remains fully juvenile with contium taxa (fig 1a). In almost all taxa the only three, very large, leaflets in adult central main segment of the leaf is consid- plants, as is seen in seedlings, also in many

Aroideana arod-35-00-05.3d 3/4/12 10:58:44 41 42 AROIDEANA, Vol. 35 other Amorphophallus species. In nearly all Molecular analyses of Amorphophallus + seedlings ever observed by us (several Pseudodracontium published to date hundred) the first seedling leaf lacks the (Grob et al., 2002, 2004; Sedayu et al., central segment (leaflet) entirely. 2010) strongly support the sister-group The polar caps of the pollen grains appear relationship of Pseudodracontium to the to be a retained non-adult phase of pollen A. longituberosus group within Amorpho- formation. Looking at the , we phallus and thus confirms the earlier observe a few characters that may be seen as morphology-based hypotheses. More re- ‘‘plesiomorphic’’ for Amorphophallus at cent molecular analysis (Randomized Ax- large, like the appendix wall with fully elerated Maximum Likelihood [RAxML]) by developed and the long, free or the second author using the markers ITS1, partly fused filaments. We hypothesize that FLint2, rbcL and matK and applied to ca. the morphogenic pathways of several phe- 130 species of Amorphophallus (incl. Pseu- notypic characters in Pseudodracontium is dodracontium) again strongly supports the deregulated. It may well be that the chaotic aforementioned grouping of Pseudodra- molecular and morphological patterns dis- contium in Amorphophallus (Claudel et al., play an evolutionary phase Pseudodracon- in prep., Hetterscheid & Claudel, in prep.) tium is going through whereby (past and but expands it with a small strongly ongoing) hybridisation (see also paragraph supported clade of 3 (possibly only 2) 3.1) may well have been the onset of the species (A. saraburiensis Gagn., A. scuta- deregulation. Heterochrony seems to be part tus Hett. & T.C. Chapman and A. tenuistylis of this deregulation, leading to adult plants Hett.). This group of the Amorphophallus with partly juvenile morphologies. species mentioned + Pseudodracontium is Pseudodracontium as a group within part of a larger and strongly supported Amorphophallus is considered to be closely clade of the ‘‘Continental Asia-II’’-clade of related to A. longituberosus (Engl.) Engl. & Sedayu et al. (2010). Gehrm. and its immediate allied species A. In the light of all evidence discussed albispathus Hett., A. coudercii (Bogn.) above it is here decided to reduce the Bogn. and A. tenuispadix Hett. This group genus Pseudodracontium to the synonymy turned up again and again in morphology- of Amorphophallus. The necessary new based attempts to unravel the phylogeny of combinations and one new name are Amorphophallus. In these analyses Pseudo- presented below. The tribe Thomsonieae dracontium species were always associat- will become monotypic. ed with the aforementioned group as a sister group. Even in chemical analyses of THE ‘‘SPECIES’’-PROBLEM IN the scent of species in this alliance (Kite & Hetterscheid, 1997; Kite et al., 1998), traces ‘‘PSEUDODRACONTIUM’’ AND A of a compound (4-methoxyphenetyl alco- PROVISIONAL NEW KEY TO THE TAXA hol, or ‘‘anise oil’’) unique to it, were found Species Number in Pseudodracontium in P. lacourii (Lind. & Andre´) N.E. Br. and P. fallax Serebryanyi. This seemed to Serebryanyi (1995) recognized 6 species strengthen the suggested relationship. The in Pseudodracontium and provided an trouble with all full-morphology-based identification key. In subsequent years, phylogeny reconstructions of Amorpho- many more collections of Pseudodracon- phallus was that no statistically relevant tium species were made by many botanists support for this grouping could be found, and studied by the first author morpholog- not even applying the Implied Weighing ically. It turned out that many species- procedure of the TNT program of Goloboff defining character combinations proposed et al. (2004; Hovenkamp & Hetterscheid, by Serebryanyi can no longer be used. 2008). The species group of A. longituber- Many character combinations that were osus reappeared in all analyses but never used to define particular species prove to associated with Pseudodracontium. be more widespread and recombined with

Aroideana arod-35-00-05.3d 3/4/12 10:58:51 42 W. L. A. HETTERSCHEID, C. CLAUDEL, 2012 43

other characters to form an ever more Amorphophallus kuznetsovii (Sereb- complex of combinations blurring almost ryanyi) Hett. & C. Claudel, comb. all conventional species borders in the nov. Basionym: Pseudodracontium ‘‘genus’’. kuznetsovii Serebryanyi. Blumea Molecular analyses by the second author 40(1)(1995): 226, fig. 2. 1995 - Type: (Claudel at al., in prep.) also show huge from a cultivivated in Hortus overlap in characters between ‘‘taxa’’ as Botanicus Leiden, the Netherlands, 15 well as suggesting extensive hybridization July 1992, Hetterscheid H.AM.165-T (particularly in ITS1 analyses). (holotypus, MHA, spiritcoll.) - orig. As a result there is a strong conviction coll. Vietnam, Xuen Moc Reserve, that the number of species of Pseudodra- easternmost part of the Dong Nai contium is overstated. In the forthcoming province, Kuznetsov s.n., 1991. Amorphophallus treatment in the Flora of Amorphophallus lanceolatus (Sereb- Thailand (Hetterscheid, in press) a key to ryanyi) Hett. & C. Claudel, comb. the taxa of ‘‘Pseudodracontium’’ is present- nov. Basionym: Pseudodracontium ed, which is duplicated here (see below) lanceolatum Serebryanyi, Blumea using the new combinations and new name 40(1)(1995): 230, fig 4a, b. - Type: in Amorphophallus. from a plant cultivivated in Hortus Botanicus Leiden, the Netherlands, 15 Five New Combinations and One New July 1992, Hetterscheid H.AM.179-T Name in Amorphophallus (holotypus, MHA, spiritcoll.) - orig. coll. Vietnam, Xuen Moc Reserve, As a result of the transfer of all Pseudo- dracontium species to Amorphophallus,five easternmost part of the Dong Nai new nomenclatural combinations and one province, Kuznetsov s.n., 1991. new name in Amorphophallus are presented Amorphophallus latifolius (Serebrya- here (names considered heterotypic syno- nyi) Hett. & C. Claudel, comb. nov. nyms by Serebryanyi in 1995 are not Basionym: Pseudodracontium latifo- recombined; for invalidly published names lium Serebryanyi, Blumea 40(1)(1995): and full synonymy, see there). The new 231, fig. 4c, d. - Type: from a plant nomenclatural combinations do not mirror cultivivated in Hortus Botanicus Lei- taxonomic opinion of the authors on the den, the Netherlands, 7 August 1991, biological reality of the taxa carrying these Hetterscheid H.AM.167-T (holotypus, names. The of this group of MHA, spiritcoll.) - orig. coll. Thailand, ‘‘species’’ is not finished yet (Hetterscheid, Kanchanaburi prov., Thong Pha in prep.). The introduction of a species group Phum, Kwai River valley, alt. 100 m., incl. A. lacourii aspresentedinthekeyinthe steep hillslope covered in bamboo, 2 next paragraph (3.3) mirrors this uncertainty July 1985, Dransfield JD 6219 p.p. and must be considered a preliminary Amorphophallus macrophyllus (Gagn. opinion and not a definitive one as yet. ex Serebryanyi) Hett. & C. Claudel, comb. nov. Basionym: Pseudodra- Amorphophallus fallax (Serebryanyi) contium macrophyllum Gagn. ex Ser- Hett. & C. Claudel, comb. nov. ebryanyi, Blumea 40(1)(1995): 232, Basionym: Pseudodracontium fallax fig. 5. - Type: Thailand, Kanchanaburi Serebryanyi, Blumea 40(1)(1995): 221, prov., Wang Kanai, 200 m. alt., in fig 1. - Type: Vietnam, Vungtau-Con crevices of limestone rocks, 15 May Dao special district, limestone hills c. 1927, Kerr 12866 (holotypus, K). 10 km E of Vung Tau port, near Mount Amorphophallus pseudoharmandii N. Chau Vien, 300 m. alt., SE slope, Hett. & C. Claudel, nom. nov. - Syn.: 100 m. below Jesus Christ momument, Pseudodracontium harmandii Engl., in thickets, 28 May 1989, Serebryanyi Bot. Jahrb. 25 (1898): 15. - Type: 8908 (holotypus, MHA, spiritcoll.). Cambodia, Compon Chnang, June

Aroideana arod-35-00-05.3d 3/4/12 10:58:51 43 44 AROIDEANA, Vol. 35

1875, Godefroy s.n. ‘‘in Exped. Dr. two or a few more staminodes Harmand’’ (holotypus, P). fused resulting in short ridges, these often slightly sinuous, The new name proposed here is made sometimes appendix with deeper necessary because of the existence of the longitudinal cracks independent name Amorphophallus harmandii Engl. & from the staminodial structure . . Gehrm, in Pflanzenr. (IV, 23C) (1911) 83 - ...... A. macrophyllus Type: Cambodia, Compon Chnang, 6 June 2b. Stigma depressed or hemispheric, 1875, Godefroy 144 ‘‘in exp. Harmand’’ 0.4–1.5 mm, with or without a (p.p., only the ) (holotypus, central depression; style very short P). To complicate matters both the holo- or distinct; flowering mostly along- type and isotype (B) of A. harmandii are a side developing or after leaf combination of inflorescences of A. har- development; appendix sometimes mandii and leaves of A. pseudoharmandii, reduced, staminodial structure which apparently got mixed. The holotype of appendix papillate, papillate- of A. pseudoharmandii is a full flowering brain-like or brain-like...... 3 specimen. 3a. Staminodial structure of appen- dix largely papillate, tops of KEY TO THE TAXA OF THE FORMER staminodes orbicular or very GENUS PSEUDODRACONTIUM, short elongate (with fusion of 2 NOW AMORPHOPHALLUS or 3 neighboring staminodes) and then often slightly sinuous; 1a. Leaf blade bright pale bluish gray stigma 1–2 mm diam ...... (not resulting from a wax layer); ...... A. pseudoharmandii appendix with a ridged-grooved 3b. Staminodial structure of appendix staminodial pattern, tops of stami- for the larger part brain-like, with nodes distinctly separated; stigma several staminodes fused into depressed to slightly hemispheric, longer, strongly sinuous brain- 1.5 mm in diam., bright yellow, fold-like ridges, more rarely with strongly echinate-scabrate; style progressed fusion to form a shal- 0.4 mm long ..... A. glaucophyllus lowly corrugate or smooth surface 1b. Leaf blade green, with or without (often in reduced appendices), or variegation, or with a greenish blue papillate but with scattered, small wax layer; appendix with a ridged- groups of a few fused staminodes, grooved, papillate, or more or less forming short convolutions; stig- brain-like staminodial pattern, or ma0.4–1mmdiam....A. lacourii rarely almost to entirely smooth (this group (incl. A. lacourii, A. kuz- usually in reduced appendices); sta- netsovii, A. fallax, A. lanceolatus minodes congested or the tops sep- and A. latifolius). arated; stigma off-white ...... 2 2a. Plants often flowering before leaf ACKNOWLEDGMENTS development; stigma disciform, 2 mm diam., with a distinct We first and foremost like to thank Michail central depression, surface nearly Serebryanyi (Russia) for the exchange of smooth or slightly corrugated, or opinions on the matters discussed in this minutely echinate; style almost paper. There wasn’t agreement on all issues nil; appendix always fully devel- but it certainly improved our view on several oped, never reduced, stamino- matters.WealsoliketothankMrs.Mary dial structure of the appendix a Sizemore (USA) who provided us with much mixture of terete, rod-like stami- pictorial information on ‘‘Pseudodracon- nodes with orbicular tips and tium’’speciesinthewildandforthesame

Aroideana arod-35-00-05.3d 3/4/12 10:58:51 44 W. L. A. HETTERSCHEID, C. CLAUDEL, 2012 45

reason Dr. Nguyen van Du (Vietnam). We Hetterscheid, W. L. A. (in press). Amor- also acknowledge the professional way in phophallus in: which personnel of the Leiden Botanical P. C. Boyce, D. Sookchaloem, W. L. A. Garden (Netherlands) and Hamburg Botan- Hetterscheid, G. Gusman, N. Jacob- ical Garden (Germany) have taken care of sen, T. Idei & V. D. Nguyen (in press) the living collection in which many ‘‘Pseudo- (eds.), Araceae. Flora of Thailand. dracontium’’ plants are present. Forest Herbarium, Bangkok. Hovenkamp, P. & W. L. A. Hetterscheid. LITERATURE CITED 2008. Searching reliable characters with Implied Weighing. Poster pre- Brown, N. E. 1882. Four new genera of sented 27th Annual meeting of the . J. Bot. n.s. 11: 193–197. Willi Hennig Society, 27–31 October Cabrera, L. I., G. A. Salazar, M. W. Chase, 2008, Tucuman. S. J. Mayo, J. Bogner & P. Da´vila. 2008. Kite, G. C. & W. L. A. Hetterscheid. 1997. Phylogenetic relationships of aroids Inflorescence odors of Amorphophal- and duckweeds (Araceae) inferred lus and Pseudodracontium (Araceae). from coding and noncoding plastid Phytochemistry 46(1): 71–75. DNA. Amer. J. Bot. 95(9): 1153–1165. Kite, G. C., W. L. A. Hetterscheid, M. J. Cusimano, N., J. Bogner, S. J. Mayo, P. C. Lewis, P. C. Boyce, J. Ollerton, E. Boyce, S. Y. Wong, M. Hesse, W. L. A. Cocklin, A. Diaz & M. J. Simmonds. Hetterscheid, R. C. Keating & J. C. 1998. Inflorescence odors and polli- French. 2011. Relationships within the nators of and Amorphophal- Araceae: comparison of morphologi- lus (Araceae). in: S. J. Owen & P. J. cal patternswith molecular phyloge- Rudall (eds.), Reproductive Biology, nies. Amer. J. Bot. 98(4): 1–15. pp. 295–315. Royal Botanic Gardens, Goloboff, P. A., J. S. Farris & K. C. Nixon. Kew. 2004. TNT. Tree analysis using new Nicolson, D. H. 1967. Selection of lecto- technology vs. 1.0. Distributed by the type species for genera of the family authors. Araceae. Taxon 16: 514–519. Grob, G. B. J., B. Gravendeel & M. C. M. Sedayu, A., M. C. M. Eurlings, B. Graven- Eurlings. 2004. Potential phylogenetic deel & W. L. A. Hetterscheid. 2010. utility of the nuclear Floricaula/Leafy Morphological character evolution of second intron: comparison with three Amorphophallus (Araceae) based on a chloroplast DNA regions in Amorpho- combined phylogenetic analysis of phallus (Araceae). Mol. Phyl. and trnL, rbcL and LEAFY second intron Evol. 30: 13–23. sequences. Bot. Stud. 51: 473–490. Grob, G. B. J., B. Gravendeel, M. C. M. Serebryanyi, M. M. 1995. A taxonomic Eurlings & W. L. A. Hetterscheid. 2002. revision of Pseudodracontium (Ara- Phylogeny of the tribe Thomsoniae ceae - Aroideae - Thomsonieae). (Araceae) based on chloroplast matK Blumea 40: 217–235. and trnL intron sequences. Syst. Bot. Van der Ham, R. W. J. M., G. Grob, W. L. A. 27(3): 453–467. Hetterscheid, W. Star & B. J. van Hetterscheid, W. L. A. 1994a. Preliminary Heuven. 2005. Notes on the genus taxonomy and morphology of Amor- Amorphophallus (Araceae) - 13. Evo- phophallus Blume ex Decaisne (Ara- lution of pollen ornamentation and ceae), in M. M. Serebryanyi (ed.), ultrastructure in Amorphophallus and Proc. Moscow Aroid Conf. 1992 Pseudodracontium. Grana 44: 252– pp. 35––48. Moscow. 265. Hetterscheid, W. L. A. 2006. Notes on the Van der Ham, R. W. J. M., W. L. A. Genus Amorphophallus (Araceae) 15. Hetterscheid & B. J. van Heuven. New Species from SE Asia. Aroideana 1998. Notes on the genus Amorpho- 29: 53–79. phallus (Araceae) - 8. Pollen morphol-

Aroideana arod-35-00-05.3d 3/4/12 10:58:51 45 46 AROIDEANA, Vol. 35

ogy of Amorphophallus and Pseudo- Morphology and Biology, pp. 241–248. dracontium. Rev. Palaeobot. & Paly- Royal Botanic Gard, Kew. nol. 103: 95–142. Van der Ham, R. W. J. M. & B. J. van Van der Ham, R. W. J. M., W. L. A. Heuven. 2001. Pollen morphology of Hetterscheid, B. J. van Heuven & Amorphophallus (Araceae). In: D. K. W. Star. 2000. Exine architecture in Goodman & R. T. Clarke (eds.), echinate pollen of Amorphophallus Proceedings of the IX International (Araceae) in relation to taxonomy. Palynological Congress, Houston, Tex- In: M. M. Harley, C. M. Morton & S. as, USA, 1996 pp. 585–591 Amer. Ass. Blackmore (eds.), Pollen and Spores: Strat. Palyn. Found.

Aroideana arod-35-00-05.3d 3/4/12 10:58:51 46