THE NERIEAE (-)

A. J. M. LEEUWENBERG1

ABSTRACT The genera of tribe Nerieae of Apocynaceae are surveyed here and the relationships of the tribe within the family are evaluated. Recent monographic work in the tribe enabled the author to update taxonomie approaches since Pichon (1950) made the last survey. Original observations on the pollen morphology ofth egener a by S.Nilsson ,Swedis h Museum, Stockholm, are appended to this paper.

RÉSUMÉ L'auteur étudie lesgenre s de la tribu desNeriea e desApocynacée s et évalue lesrelation s del a tribu au sein de la famille. Un travail monographique récent sur la tribu a permit à l'auteur de mettre à jour lesapproche s taxonomiques depuis la dernière étude de Pichon (1950). Lesobservation s inédites par S. Nilsson du Muséum d'Histoire Naturelle Suédois à Stockholm sur la morphologie des pollens des genres sontjointe s à cet article.

The Apocynaceae have long been divided into it to generic rank and in his arrangement includ­ two subfamilies, Plumerioideae and Apocynoi- ed in the Echitinae. Further, because deae (Echitoideae). Pichon (1947) added a third, of its conspicuous resemblance to , the Cerberioideae, a segregate of Plumerioi­ it may well be that (Echiteae— deae—a situation which I have provisionally ac­ Amalocalycinae, according to Pichon) ought to cepted. These subfamilies were in turn divided be moved to the Nerieae. into tribes and subtribes. Comparative studies Pichon's system is artificial, because he used have shown that the subdivision of the Plume­ the shape and the indumentum of the area where rioideae is much more natural than that of the the connectives cohere with the head of the pistil Apocynoideae. as the only character to distinguish the four tribes Although the Apocynoideae is a coherent of his Apocynoideae. That the Nerieae is the grouping, it has so far proved very difficult to most satisfactorily defined tribe of the Apocy­ devise a satisfactory and practicable subdivision noideae only becomes evident after combina­ for it. Pichon (1950 ) had little confidence in early tions of characters are taken into account. classifications ofth e subfamily, although, like his Pichon divided the Nerieae into nine subtribes predecessors, he seems to have preferred a more (Table 1), but it is certain that not all of these or less artificial treatment. He recognized four can be maintained. Indeed, it is questionable tribes in Apocynoideae —Echiteae, Nerieae, whether the Nerieae ought to be divided into Apocyneae, and Ichnocarpeae—each of which subtribes at all. was further divided into a number of subtribes. Connections canb e seenbetwee n certain groups However, none of these can be maintained as ofgenera . Thus has flowers that resem­ delimited by Pichon. For example, there appears ble those oiNerium, especially in the corona and tob ea greate r degreeo fsimilarit y between in the long bristly appendages at the apices of and Rhodocalyx of the Echiteae and the anthers. and of the Ichnocarpeae than is ev­ , , and Stephanostema can ident from his classification. also be grouped together, as they have a corona Nevertheless, the Nerieae, dealt with in detail (deeply incised in the first two and almost entire here, are maintained largely as proposed by Pi­ in the last one), more or less bifurcate ramifi­ chon except for the following changes: Dewev- cation, and a similarly shaped corolla. There is, rella is transferred to Parsonsiineae (Echiteae). in addition, the resemblance shown by Beau­ is treated as a synonym of Aga- montia and to Wrightia, in particular to nosma. De Candolle (1844 ) regarded Amphineu­ those ofit sspecie swit h abroa d corolla and thick- rion as a section ofAganosma, but Pichon raised walled fruits. , with comparable

1 Department of , Agricultural University, Wageningen, The Netherlands.

MONOGR. SYST. BOT. MISSOURI BOT. GARD. 25: 157-160. 1988. 158 1985AETFA TPROCEEDINGS-MODER N SYSTEMATIC STUDIES

TABLE 1. Pichon's treatment of the Nerieae compared with the arrangement proposed here.

Pichon(1950) Leeuwenberg (1988) Nerieae Nerieae 1. Neriinae (as Amphineuriinae) Adenium (5;Afr. , Arab.; Plaizier, 1980) Amphineurion (= Aganosmä) (1; Medit. to ; Leeuwenberg, 1984) Potisia Strophanthus (38;Afr. , As.; Beentje, 1982) Dewevrella (to Parsonsiinae) Wrightia(23 ; Afr., As.; Ngan, 1965) lsonema Pleioceras (5; Afr.; Barink, 1983) Nerium Stephanostema (1; Afr.; Barink, 1983) 2. Beaumontiinae Beaumontia (9;As. ; Rudjiman, 1986) Vallaris Vallaris (3;As. ; Rudjiman, 1982) Parabecumontia( = Vallaris) Vallariopsis (1; As.;Rudjiman , 1986) Muantum (= Beaumontia) (15 ;As. ; Rudjiman, 1986) Beaumontia (ca. 15;Afr. , S.Amer. ;Va n der Ploeg, 1984, 1985; Fallen, 3. Strophantinae unpubl.) Christya( = Strophantus) Allowoodsonia (1; Pacif.; Fallen; unpubl.) Roupellina( = Strophantus) (2;Afr. ; Zwetsloot, 1981) Strophanthus (ca . 10;Afr. , Mad.; Markgraf, 1976) (1; Afr.; Zwetsloot, 1981) 4. Mascarenhasiinae (ca . 25;Afr. , Mad.; Pichon, 1954) Mascarenhasia lsonema (3; Afr.; Van der Ploeg, 1983) Echitella (= Mascarenhasia) (3 ;As. ; Tsiang & Li, 1977) 5. Alafiinae Alafia Aladenia (= Farquharia) 6. Kibataliinae Funtumia Kibatalia 7. Wrightiinae Wrightia Wallida( = Wrightiä) Scleranthera( = Wrightia) Pleioceras Stephanostema 8. Malouetiinae Malouetia Malouetiella( = Malouetia) (Pichon, 1952) 9. Adeniinae Adenium Vallariopsis not included. Allowoodsonia not mentioned.

fruits and flowers with a corona, also belongs these genera from each other. Within the group, here. Malouetia and Allowoodsonia are so close in all A third group comprises Kibatalia, Funtumia, their characters, including those of the seeds (the Mascarenhasia, Malouetia, and Allowoodsonia. only ones without a coma in the whole of the These five genera probably all conform to the Apocynoideae), that the two should probably be model of Koriba when in the flowering stage. united. Kibatalia differs from them only in the Funtumia, however, matches the model of Pré­ rostrum and coma of the seeds. vost in the juvenile stage. This has still to be Farquharia and Alafia, large lianes with cori­ investigated for the other four genera. The co­ aceous , exhibit similarity in their flowers riaceous leaves,smal l stunted inflorescences, and and fruits, and they therefore form a fourth group fleshy corollas can make it difficult to distinguish within the Nerieae. LEEUWENBERG-NERIEAE 159

Of the three genera not yet mentioned, Val- XXXIV. Les Malouetiinae d'Afrique. Bull. Jard. lariopsis, , and Pottsia, it is mainly Val- Bot. État 22: 115-135. . 1954. Classification des Apocynacées: lariopsis that shows connections with other gen­ XXXIX. Revision du Genre Alafia Thou. Bull. era. Thus, its flowers resemble those ofKibatalia, Jard. Bot. État 24: 129-222. Allowoodsonia, and Malouetia, and to a lesser PLAIZIER, A. C. 1980. A revision of Adenium Roem. extent those of Vallaris, Isonema, and Pottsia. & Schult, and of Diplorhynchus Welw. ex Fie. & Hiern (Apocynaceae). Meded. Landbouwhoge­ The four groups outlined above could, if nec­ school 80(12): 1-40. essary, be distinguished as the subtribes Neri- RUDJIMAN. 1982. A revision of Vallaris Burm. f. inae, Wrightiinae, Kibataliinae, and Alafiinae, (Apocynaceae). Meded. Landbouwhogeschool but they are difficult to define. 82(11): 1-17. Strophantus gratus has flowers reminiscent . 1986. A revision of Beaumontia Wall., Ki- batalia G. Don, and Vallariopsis Woods. (Apo­ of Adenium and Nerium in their corolla and co­ cynaceae). Agric. Univ. Wageningen Pap. 86(5): rona. The corona and syncarpous fruits of Neri­ 1-99. um have their counterparts in several Wrightia TSIANG, Y. & P. T. Li. 1977. Apocynaceae. In Flora species. The model of Leeuwenberg has been rec­ Republicae Popularis Sinicae 63:1-XH, 1-249. ognized in species of Nerium, Strophanthus boi- VAN DER PLOEG, J. 1983. A revision of Isonema R. Br. and Pycnobotrya Benth. (Apocynaceae). Med­ vinii, and Wrightia. The seeds of Beaumontia ed. Landbouwhogeschool 83(4): 1-20. resemble those of Isonema and Pottsia. In ad­ . 1984. Serieso frevision s ofApocynacea e XIV. dition, the flowers of Isonema bear lateral ap­ Notes on the African species of the Ma­ pendages on the corolla lobes and, in this respect, louetia A.DC. (Apocynaceae). Bull.Jard . Bot. État 54: 283-288. are like those of Mascarenhasia. The remarkable -. 1985. A revision of Cyclocotyla Stapf, Dew- long, slender, rostrate seeds of Strophanthus, with evrella De Wild., and the African species of Ma­ a coma on the rostrum, are also found in Fun- louetia A.DC. (Apocynaceae). Agric. Univ. Wa­ tumia and Kibatalia. On the other hand, the seeds geningen Pap. 85: 57-83. of Adenium and Farquharia have basal and api­ ZWETSLOOT, H. J. C. 1981. A revision of Farquharia Stapfand Funtumia Stapf (Apocynaceae). Meded. cal comas that are persistent, while in Strophan­ Landbouwhogeschool 81(16): 1-46. thus, Funtumia, and Isonema the second coma is deciduous.

APPENDIX LITERATURE CITED POLLEN DESCRIPTION OF BARINK, M. M. 1983. Series of revisions of Apocy- APOCYNACEAE-APOCYNOIDEAE-NERIEAE2 naceae XII. A revision ofPleioceras Baill., Stepha- nostema K. Schum. and Schizozygia Baill. (Apo- Pollen grains usually spheroidal, 3(-4)-porate, ex­ cynaceae). Meded. Landbouwhogeschool 83(7): 21- ceptionally polypantoporate (Vallariopsis); exine 53. smooth, perforated, often densely spaced towards the BEENTJE, H. J. 1982. A monograph on Strophanthus apertures; poresprovide d with thickened margin; inner DC. (Apocynaceae). Meded. Landbouwhoge­ exine consisting of globular or irregular elongated ele­ school 82(4): 1-191. ments that are free or partly fused. DE CANDOLLE, A. 1844. Apocynaceae. In Prodromus The exine fine structure varies among the taxa in­ Systematis Naturalis Regni Vegetabilis 8: 317-489. vestigated with five or possibly six types: Paris. (1) Exine stratified into tectum, a distinct central FALLEN, M. E. Unpubl. Revision ofMalouetia A. DC. granular-columellatestratu m and a basal stratum (sole, and Allowoodsonia Mgf. foot layer), endexine obsolete; the sole appears frag­ LEEUWENBERG, A. J. M. 1984. Series of revisions of mented towards the pores {). In Val­ Apocynaceae XIII. Notes on Nerium L. and Ta- laris solanacea the granular stratum, when discernible, bernaemontana L. Meded. Landbouwhogeschool is displaced proximally towards the laminated sole. 83(7): 57-60. (2) Exine with a relatively thick tectum, a distinct MARKGRAF, F. 1976. Apocynacées. In J.-F. Leroy granular stratum more proximally located than in pre­ (editor), Flore de Madagascar et des Comores, vious type, an unevenly thickened, partly discontin­ Famille 169. uous sole, fragmented near the apertures and possibly NGAN, P. T. 1965. A revision of the genus Wrightia subtended by a thin, indistinct endexine (Nerium (Apocynaceae). Ann. Missouri Bot. Gard. 52:114— oleander). 175. (3) Exine stratified into tectum, a thin indistinct PiCHON,M . 1947. Classification des Apocynacées: V. granular stratum subtended by a sole to which granules Cerbéroïdées. Notul. Syst. (Paris) 13: 212-229. or knoblike elements are attached; near the apertures —. 1950. Classification des Apocynacées: XXV. only granules (Pleioceras barteri). Echitoïdées et supplément aux Plumérioïdées. Mém. Mus. Natl. Hist. Nat., Sér. B, Bot. 1: 1-173. 2 S. Nilsson, Palynological Laboratory, Sweden Nat­ . 1952. Classification des Apocynacées: ural History Museum, Stockholm. 160 1985AETFA TPROCEEDINGS-MODER N SYSTEMATIC STUDIES

(4) Exine unstratified, with or without a central deviates from all others by being poly-panto-porate. commissural line;inne r faceconsist so fsmal lan d large Pollen morphology does not reveal any evidence as to granules (Farquharia elliptica, Funtumia africana, Ki- the position and interrelationships ofAllowoodsiana, batalia gitingensis, Mascarenhasia arborescens var. Beaumontia,Isonema, Malouetia (Malouetiella), Pott- boivinii, Strophanthusgratus,an d Wrightiapubescens). sia, and Stephanostema. Nor does it support the re­ (5) Exineconsist so fa noute rtecta l stratum towards moval ofAmphineuron an d Dewevrella from the Neri- whichfre e orfuse d rounded-elongated elementsappea r eae. piled up (Alafia lucida). Pollen morphologically, Adenlum and Nerium ap­ Theabov earbitrar ygroup sar enot sharply delimited pearrelated . Wrightiaan d Strophanthusappea r related from each other, in particular group 4 from groups 3 moret oKibatalia, Funtamia, Mascarenhasia, etc.tha n and 5, respectively. From the complete exine in Ade- to Pleioceras. Vallaris showscertai n similarity toAde­ nium there is a series of transitional exine structures nium; Vallariopsis is unique within the Nerieae. Far­ to the tectal-granular exine ofAlafia. quharia appears closer to Kibatalia, Funtumia, etc., Out of the remaining genera (pollen grains seen in than toAlafia. LM or SEM only) it isonl y Vallariopsis that markedly