Notes on the Flora of Madagascar, 14-21 Martin W

Home , Angraecopsis, Angraecum eburneum, Brownleea, Diospyros, Diospyros seychellarum, Dypsis madagascariensis

MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page397

Notes on the flora of Madagascar, 14-21 Martin W. Callmander, Peter B. Phillipson & Laurent Gautier (ed.)

Abstract Résumé CALLMANDER, M. W., P. B. PHILLIPSON & L. GAUTIER (ed.) (2011). CALLMANDER, M. W., P. B. PHILLIPSON & L. GAUTIER (ed.) (2011). Notes on the flora of Madagascar, 14-21. Candollea 66: 397-423. In English, Notes sur la flore de Madagascar, 14-21. Candollea 66: 397-423. En anglais, English and French abstracts. résumés anglais et français. Ongoing research on Madagascar’s flora is revealing numer- Les recherches en cours sur la flore de Madagascar révèlent de ous taxonomic novelties and nomenclatural inconsistencies, nombreuses nouveautés taxonomiques, des problèmes de nomen- and providing new data on species distribution. This is the third clature et de nouvelles données sur la distribution des espèces. set of notes in a series that aims to provide the botanical com- Par cette série, nous souhaitons donner à la communauté bota- munity working on the flora of Madagascar an opportunity to nique internationale travaillant sur Madagascar la possibilité de publish short communications on these topics, and comprises publier de courtes contributions traitant de ces aspects. Cette troi- eight notes. sième livraison comprend huit notes. – Note 14. Lectotypification of two species of Bignoniaceae – Note 14. Lectotypification de deux espèces de Bignoniaceae from Madagascar, by Martin W. Callmander & Peter B. Phillip- de Madagascar, par Martin W. Callmander & Peter B. Phillip- son. Lectotypes are designated for two species of Bignoniaceae son. Des lectotypes sont désignés pour deux espèces de Bigno- from Madagascar, Rhodocolea boivinii (Baill.) H. Perrier niaceae de Madagascar, Rhodocolea boivinii (Baill.) H. Perrier and Stereospermum variabile H. Perrier. Additional observa- et Stereospermum variabile H. Perrier. Des observations addi- tions and conservation threat analyses are also provided for the tionnelles ainsi qu’une évaluation du degré de menace sont aussi two species. fournies pour chacune des deux espèces. – Note 15. Two new combinations in Memecylon L. (Melas- – Note 15. Deux nouvelles combinaisons dans Memecylon L. tomataceae) of Madagascar, by R. Douglas Stone & Martin (Melastomataceae) à Madagascar, par R. Douglas Stone & W. Callmander. This note proposes new combinations at Martin W. Callmander. Cette note propose des nouvelles com- species rank for the taxa originally described as Memecylon binaisons au rang spécifique pour des taxa décrits originellement pedunculatum var. centrale Jacq.-Fél. and Memecylon eglan- comme Memecylon pedunculatum var. centrale Jacq.-Fél. dulosum var. bezavonense Jacq.-Fél. A lectotype is also des- et Memecylon eglandulosum var. bezavonense Jacq.-Fél. Un ignated for the latter. lectotype est aussi désigné pour le second. – Note 16. Pandanaceae specimens relocated from the – Note 16. Transfert des spécimens de Pandanaceae de l’Uni- University of Neuchâtel (NEU), with two new synonymies, versité de Neuchâtel (NEU), avec deux nouvelles synonymies, by Martin W. Callmander, Michel O. Laivao & Peter B. Phillip- par Martin W. Callmander, Michel O. Laivao & Peter B. Phil- son. In 2010, the Pandanaceae specimens held at the Univer- lipson. Les spécimens de Pandanaceae déposés à l’herbier de sity of Neuchâtel Herbarium (NEU) were permanently trans- l’Université de Neuchâtel (NEU) ont été transférés de façon ferred to the herbarium of the Conservatoire et Jardin permanente en 2010 à l’herbier des Conservatoire et Jardin

Addresses of the editors: MWC: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. and Conservatoire et Jardin botaniques de la Ville de Genève, ch. de l’Impératrice 1, CP 60, 1292 Chambésy. Switzerland. E-mail: [email protected] PBP: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. and Muséum national d’Histoire naturelle, Département Systématique et Evolution, UMR 7205, OSEB, case postale 39, rue Cuvier 57, 75231 Paris, cedex 05, France. E-mail: [email protected] LG: Conservatoire et Jardin botaniques de la Ville de Genève and Université de Genève, Laboratoire de botanique systématique et biodiversité, ch. de l’Impératrice 1, CP 60, 1292 Chambésy. Switzerland. E-mail: [email protected]

Online ISSN: 2235-3658 Candollea 66(2): 397- 423 (2011) © CONSERVATOIRE ET JARDIN BOTANIQUES DE GENÈVE 2011 MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page398

398 – Candollea 66, 2011

botaniques de la Ville de Genève (G). These included 11 type botaniques de la Ville de Genève (G). Il est donc nécessaire specimens, and it is therefore necessary to amend the proto- d’amender les protologues de ces espèces afin de donner la logues for these species, to reflect these transfers. This note nouvelle répartition des 11 spécimens types inclus dans ce trans- provides a complete list of specimens concerned and details fert. Cette note fourni une liste complète des spécimens concer- the necessary changes. Two species names based on this mate- nés et détaille les changements nécessaires. Deux noms d’espèces rial are discussed, and are now formally treated as synonyms. basés sur ce matériel sont discutés et formellement traités comme – Note 17. The genus Vernoniopsis Humbert (Asteraceae) in synonymes. Madagascar, by Martin W. Callmander & Peter B. Phillipson. – Note 17. Le genre Vernoniopsis Humbert (Asteraceae) à Mada- The authors present a brief review of the endemic Malagasy gascar, par Martin W. Callmander & Peter B. Phillipson. Les genus Vernoniopsis Humbert (Asteraceae). Two distinct species auteurs présentent un bref compte rendu du genre endémique are recognised: Vernoniopsis caudata (Drake) Humbert and a malgache Vernoniopsis Humbert (Asteraceae). Deux espèces dis- second species, based on Vernoniopsis caudata subsp. loko- tinctes sont reconnues: Vernoniopsis caudata (Drake) Humbert hense Humbert, for which we designate a lectotype and pro- et une deuxième espèce, basée sur Vernoniopsis caudata subsp. vide the new combination Vernoniopsis lokohense (Humbert) lokohense Humbert, pour laquelle les auteurs désignent un lec- Callm. & Phillipson. totype et fournissent la nouvelle combinasion Vernoniopsis loko- – Note 18. Two new records of Orchidaceae from Madagascar, hense (Humbert) Callm. & Phillipson. with an updated list of species shared with continental Africa, – Note 18. Deux nouvelles occurences d’Orchidaceae à Mada- by Phillip Cribb, Laurent Gautier, Sonia Trigui & Louis gascar, ainsi qu’une liste mise à jour des espèces communes avec Nusbaumer. Didymoplexis verrucosa J. Stewart & Hennessy l’Afrique continentale, par Phillip Cribb, Laurent Gautier, Sonia and Hetaeria heterosepala (Rchb. F.) Summerh., two orchid Trigui & Louis Nusbaumer. Didymoplexis verrucosa J. Stewart species hitherto only known from Africa, are newly reported & Hennessy et Hetaeria heterosepala (Rchb. f) Summerh., deux from Madagascar, the latter also representing a new generic espèces connues jusque là uniquement d’Afrique, ont été trou- record for the island. A list of species shared between Mada- vées à Madagascar, la seconde appartenant à un genre nouveau gascar and Continental Africa is provided and discussed. pour l’Ile. Une liste des espèces partagées entre Madagascar et – Note 19. Malagasy Garcinia L. (Clusiaceae) revisited: new l’Afrique continentale est présentée et discutée. combinations and lectotypifications for Ochrocarpos goudotianus and O. evonymoides and the lectotypification of Garcinia – Note 19. Contribution à la connaissance des Garcinia L. comorensis, by Zachary S. Rogers, Patrick W. Sweeney & malgaches (Clusiaceae): nouvelles combinaisons et lectoty- Laurent Gautier. New combinations are provided for two Mala- pifications pour Ochrocarpos goudotianus et O. evonymoides gasy species of Clusiaceae: Garcinia evonymoides (Planch. & et lectotypification de Garcinia comorensis, par Zachary S. Triana) P. Sweeney & Z. S. Rogers, Garcinia goudotiana(Planch. Rogers, Patrick W. Sweeney & Laurent Gautier. Des nouvelles & Triana) P. Sweeney & Z. S. Rogers. The names Ochrocarpos combinaisons sont proposées pour deux espèces malgaches de multiflorus O. Hoffm. and Garcinia disepala Vesque are treated Clusiaceae: Garcinia evonymoides (Planch. & Triana) P. Swee- as synonyms of Garcinia evonymoides Planch. & Triana, and ney & Z. S. Rogers, Garcinia goudotiana (Planch. & Triana) Garcinia melleri Baker and Garcinia comorensis Drake as syn- P. Sweeney & Z. S. Rogers. Les noms Ochrocarpos multiflorus onyms of Garcinia goudotiana Planch. & Triana. Lectotypes are O. Hoffm. et Garcinia disepala Vesque sont traités comme syno- designated for three names (Garcinia comorensis, Ochrocarpos nymes de Garcinia evonymoides Planch. & Triana; et Garcinia evonymoides and Ochrocarpos goudotianus). melleri Baker et Garcinia comorensis Drake comme synonymes – Note 20. The legacy of Henri Jumelle in Marseille: an over- de Garcinia goudotiana Planch. & Triana. Des lectotypes sont looked collection of Palms from Madagascar, by Martin W. désignés pour trois noms (Garcinia comorensis, Ochrocarpos Callmander, Bruno Vila, John Dransfield & Henk Beentje. evonymoides and Ochrocarpos goudotianus). Henri Jumelle (1866-1935) described 328 new taxa from – Note 20. L’héritage de Henri Jumelle à Marseille: une collec- Madagascar, including numerous Arecaceae, based mainly on tion oubliée de palmiers de Madagascar, par Martin W. Call- specimens collected by Henri Perrier de la Bâthie. A recent mander, Bruno Vila, John Dransfield & Henk Beentje. Henri visit to the herbarium of the University of Provence in Mar- Jumelle (1866-1935), a décrit 328 nouveaux taxons de Mada- seille (MARS) established that a nearly complete set gascar, comprenant en particulier de nombreuses Arecaceae, (217 collections) of Perrier de la Bâthie’s Arecaceae material, principalement basées sur des spécimens récoltés par Henri which has been generally overlooked, is present at MARS. Perrier de la Bâthie. Une récente visite à l’herbier de l’Univer- – Note 21. A new name for the endemic species of Diospyros sité de Provence à Marseille (MARS) a permis de retrouver une L. (Ebenaceae) in the Seychelles, by George E. Schatz & collection presque complète (217 échantillons) du matériel Porter P. Lowry II. Discovery in the Paris herbarium of a d’Arecaceae de Perrier de la Bâthie. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page399

Notes on the flora of Madagascar, 14-21 – 399

mislabeled duplicate of Pervillé 36 results in an older name – Note 21. Un nom nouveau pour l’espèce endémique de Dios- for the endemic Seychelles ebony. The authors propose a new pyros L. (Ebenaceae) des Seychelles, par George E. Schatz & combination for this taxon: Diospyros boiviniana (Baill.) Porter P. Lowry II. La découverte dans l’herbier de Paris d’un G. E. Schatz & Lowry. double de Pervillé 36 étiqueté de façon erronée a pour consé- quence la mise à jour d’un nom antérieur pour l’espèce d’ébène endémique des Seychelles. Les auteurs proposent une nouvelle Key-words combinaison pour ce taxon: Diospyros boiviniana (Baill.) G. E. Schatz & Lowry. Madagascar – Africa – Seychelles – BIGNONIACEAE – MELASTOMATACEAE – PANDANACEAE – ASTERACEAE – ORCHIDACEAE – CLUSIACEAE – ARECACEAE –EBE- NACEAE – Rhodocolea – Stereospermum – Memecylon – Martellidendron – Pandanus – Vernoniopsis – Didymoplexis – Hetaeria – Garcinia – Ochrocarpos – Dypsis – Ravenea – Diospyros – Nomenclature – Taxonomy – IUCN Red List – Jumelle MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page400

14. CALLMANDER, Martin W. & Peter B. PHILLIPSON: Lectotypification of two species of Bignoniaceae from Madagascar

Introduction Louis H. Boivin, we choose Hildebrandt 3224 as the lectotype because both the number and the quality of duplicates avail- A recent review of the Malagasy Bignoniaceae for the able are superior. The species occurs in north-west Madagas- Catalogue of the Vascular Plant of Madagascar (MADAGASCAR car within the Sub-humid bioclimatic region in low elevation CATALOGUE, 2011) has enabled us to refine species delimitations forest (the Sambirano Domain sensu HUMBERT, 1951, 1955) and nomenclature, and to discover new species in most of the (Fig. 1). It is characterized by its large sub-coriaceous impar- genera (CALLMANDER & al., 2011a; in press a; CALLMANDER & ipinnate leaves, its long terminal paniculate inflorescences PHILLIPSON, in press b, c). The family is currently represented in bearing flowers with large pink corollas with white margins Madagascar by c. 70 species, all endemic, but we estimate that and marked yellow in the throat. It is related to a small group the total will reach close to 100 once our revisionary work is of distinctive species of Rhodocolea Baill. that includes complete (CALLMANDER & al., 2011b). Certain species require R. telfairiae (Hook.) H. Perrier and two or three as yet unde- lectotypification, among these are Rhodocolea boivinii (Baill.) scribed species that also possess terminal panicles but differ H. Perrier and Stereospermum variabile H. Perrier, which are in other floral and vegetative characters. the subjects of this note. We designate lectotypes, provide some Conservation status. – With an EOO of 1695 km2, an AOO additional observations and present conservation threat analy- of 63 km2 and 5 subpopulation, one encompassed in a pro- ses (IUCN, 2001) for both species. Lists of all the correspon- tected area (Lokobé), R. boivinii is assigned a preliminary sta- ding collections and distribution maps for these and related tus of Vulnerable (VU B1ab[i, iii]; B2ab[i, iii]) (calculation species can be found in the MADAGASCAR CATALOGUE (2011). following CALLMANDER & al., 2007). Stereospermum variabile H. Perrier in Ann. Mus. Colon. Mar- Lectotypification seille ser. 5, 6: 13. 1938. Rhodocolea boivinii (Baill.) H. Perrier in Ann. Mus. Colon. Lectotypus (designated here): MADAGASCAR. Prov. Marseille ser. 5, 6: 23. 1938. Mahajunga: Morondava, s.d, fl. & fr., Perrier de la Bâthie 10296 (P [P00093992]!; isolecto-: P [P00093990, P0009 ϵ Colea boivinii Baill. in J. Linn. Soc., Bot. 25: 308. 1888. 3991]!). Observations. – Stereospermum variabile is based on five Lectotypus (designated here): MADAGASCAR. Prov. syntypes (Louvel 163, Grevé 27, 41, 50, and Perrier de la Antsiranana: Nosy-Be, IX.1879, fl., Hildebrandt 3224 Bâthie 10296). Only Perrier de la Bâthie’s collection bears (P [P00648798]!; isolecto-: BM!, G [G00191496, G001953 both flowers and mature fruits and therefore, being the most 26, G00191455]!, M [M-0105565] image seen, P [P00648 representative of the species, it is designated as the lectotype. 799]!, W image seen). The species occurs in south-western Madagascar in forest and ϭ Colea glaucescens Palacky, Cat. Pl. Madagasc. 3: 54. secondary vegetation on sandy soils, between the Onilahy 1907 [nom. nud.]. and Tsiribihina Rivers and essentially within the Sub-arid bio- Observations. – Rhodocolea boivinii is based on two syn- climatic region (Fig. 1) and is characterized by its narrow types (Boivin 2105/2, Hildebrandt 3224). Both are gatherings pubescent silvery leaflets borne on long petiolules. Like most of flowering specimens. Even though the species is named after Malagasy species of Stereospermum Cham. it has flowers with

Notes on the flora of Madagascar, 14-21 – 401

Acknowledgements Our research was supported by grants from the U.S. National Science Foundation (0743355) and the Andrew W. Mellon Foundation.

References CALLMANDER, M. W. & P. B. P HILLIPSON (in press b). Four new species of the endemic genus Rhodocolea (Bignoniaceae) from Madagascar. Adansonia 33. CALLMANDER, M. W. & P. B. P HILLIPSON (in press c). Two new species in the genus Colea (Bignoniaceae) from Madagascar. Adansonia 33. CALLMANDER, M. W., P. B. PHILLIPSON, M. RAZANAJATOVO & L. NUS- BAUMER (2011a). The genus Ophiocolea H. Perrier (Bignoniaceae) in northern Madagascar with description of four new species and two lectotypifications. Candollea 66: 133-145. CALLMANDER, M. W., P. B. PHILLIPSON & G. E. SCHATZ (in press a). Two new species of Stereospermum (Bignoniaceae) from Mada- gascar. Novon. CALLMANDER, M. W., G. E. SCHATZ, P. P. LOWRY II, M. O. LAIVAO, J. RAHARIMAMPIONONA, S. ANDRIAMBOLOLONERA, T. RAMINOSOA & T. CONSIGLIO (2007). Application of IUCN Red List criteria and assessment of priority areas for plant conservation in Mada- gascar: rare and threatened Pandanaceae indicate new sites in need of protection. Oryx 41: 168-176. CALLMANDER, M. W., P. B. PHILLIPSON, G. E. SCHATZ, S. ANDRIAMBO- LOLONERA, M. RABARIMANARIVO, N. RAKOTONIRINA, J. RAHARI- MAMPIONONA, C. CHATELAIN, L. GAUTIER & P. P. LOWRY II (2011b). The endemic and non-endemic vascular flora of Madagascar updated. Pl. Ecol. Evol. 144: 121-125. Fig. 1. – Distribution of Rhodocolea boivinii (Baill.) H. Perrier (squares) and HUMBERT, H. (1951). Les territoires phytogéographiques du nord de Stereospermum variabile H. Perrier (stars) in Madagascar mapped on the Madagascar. Compt. Rend. Sommaire Séances Soc. Biogéogr. 246: phytogeographical map of HUMBERT (1955). 176-184. HUMBERT, H. (1955). Les territoires phytogéographiques de Mada- gascar. Leur cartographie. Année Biol. ser. 3, 31: 439-448. pure white corollas. It is partially sympatric with S. euphori- IUCN (2001). IUCN Red List Categories and Criteria (Version 3.1). oides DC. (which also occurs further to the North and further IUCN Species Survival Commission, IUCN, Gland & Cambridge. inland) and S. nematocarpum DC. (which is widespread throughout the South-West and South). Perrier de la Bâthie MADAGASCAR CATALOGUE (2011). Catalogue of the vascular plants of Madagascar [http://www.efloras.org/madagascar]. remarked on the probable existence of natural hybrids between S. variabile and these species in their areas of overlap, and we concur with this point of view. Conservation status. – With an EOO of 45,946 km 2, an AOO of 171 km2 and 15 subpopulation, one encompassed in three protected area (Isalo, Kirindy Mitea and Zombitsy), S. variabile is assigned a preliminary status of Least Concern (LC) (calculation following CALLMANDER & al., 2007). MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page402

15. STONE, R. Douglas & Martin W. CALLMANDER: Two new combinations in Memecylon L. (Melastomataceae) of Madagascar

Introduction Observations. – Memecylon centrale appears to be a narrow endemic of forest vestiges in the Bongolava mountains, Memecylon L. (Melastomataceae) is widely distributed in situated along the western edge of Madagascar’s Central the Old World tropics and consists of more than 300 species Plateau. It shares with M. pedunculatum Jacq.-Fél. the char- of generally small trees or shrubs, most of them occurring in acter of young branchlets with brownish-black bark soon exfo- the understory of evergreen forest (RENNER & al., 2007-2011). liating in strips to reveal whitish inner bark, as well as the char- For Madagascar, the most recent treatment of the genus rec- acter of the fruit, which is sessile above the peduncle. Its leaves ognized 78 species (JACQUES-FÉLIX, 1985a, 1985b), and much however are narrowly lanceolate (dimensions 3.3-4.4 ϫ 0.9- of the recently collected material appears to represent species 1.2 cm) and weakly acuminate at the apex, while in M. pedun- new to science (STONE, 2006). Pending completion of a com- culatum the leaves are ovate-lanceolate (dimensions 4.3-6.1 ϫ prehensive taxonomic revision of Malagasy Memecylon, the 1.3-2.4 cm) and distinctly caudate-acuminate (the acumen lin- purpose of this note is to propose new combinations at species ear, c. 1.5 cm long). Furthermore, the area inhabited by M. cen- rank for the taxa originally described as M. pedunculatum var. trale is separated by c. 600 km from the region where M. centrale Jacq.-Fél. and M. eglandulosum var. bezavonense pedunculatum occurs, in and around Marojejy National Park Jacq.-Fél. A lectotype is also designated for the latter. Each in northeastern Madagascar (Antsiranana Province). Species species is provided with observations and conservation status status for M. centrale is justified by its geographic isolation using the current IUCN Red List Criteria (2001). from M. pedunculatum and the evident lack of morphological intermediates between these taxa. Taxonomy and nomenclature Additional specimens. – MADAGASCAR. Prov. Antananarivo: Tsiroanomandidy, Ambaravaranala, 18°25’S 45°40’E, alt. 1190-1200 m, forêt Memecylon centrale (Jacq.-Fél.) R. D. Stone, stat. nov. sur pente et forêt galerie, 25.VI.1997, fr., Rakotomalaza & al. 1356 (CAS, MO, P, TAN). ϵ Memecylon pedunculatum var. centrale Jacq.-Fél. in Bull. Mus. Natl. Hist. Nat., B, Adansonia 6: 451. 1985. Memecylon bezavonense (Jacq.-Fél.) R. D. Stone, stat. nov. Typus: MADAGASCAR. Prov. Antananarivo: massif ϵ Memecylon eglandulosum var. bezavonense Jacq.-Fél. d’Anorimbato, à l’Ouest d’Ambaravaranala (Tsiroanoman- in Bull. Mus. Natl. Hist. Nat., B, Adansonia 7: 10. didy), aux confins de la haute Mahavavy et de la haute 1985. Bebao, [18°26’S 45°37’E], 15-16.V.1970, fr., Service Lectotypus (designated here): MADAGASCAR. Prov. Forestier 29170 (holo-: P [P00062672]!; iso-: K [K000 Antsiranana: massif du Bezavona (entre la Fanambana et 313587]!, MO [MO-313781, MO-313782]!, P [P00062673, la Manambery), pentes inférieures de la rive droite de P00062674]!, TEF [TEF000321]!). l’Andilana, [13°32’S 49°54’E], 20.III.1967, fl., Service Conservation status. – With an AOO of 18 km2 and 2 sub- Forestier 27546 (P [P00057775]!; isolecto-: P [P00057 populations, neither of which are situated within the protected 776]!, MO [MO-345892, MO-391665]!, TEF!). area network, M. centrale is assigned a preliminary status Conservation status. – With an EOO of > 100 km 2, an of Endangered (EN B2ab[iii]) following the IUCN Red List AOO of 18 km2 and 2 subpopulations, neither of which are sit- Categories and Criteria (IUCN, 2001) (calculation following uated within the protected area network, M. bezavonense is CALLMANDER & al., 2007). assigned a preliminary status of Endangered (EN B1ab[iii];

Addresses of the authors: RDS: School of Biological and Conservation Sciences, University of KwaZulu-Natal, Pietermaritzburg campus, Private Bag X01, Scottsville 3209, Republic of South Africa. Email: [email protected] MWC: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. and Conservatoire et Jardin botaniques de la Ville de Genève, ch. de l’Impératrice 1, CP 60, 1292 Chambésy, Switzerland. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page403

Notes on the flora of Madagascar, 14-21 – 403

B2ab[iii]) following the IUCN Red List Categories and Crite- collections have revealed that the fruits of M. bezavonense are ria (IUCN, 2001) (calculation following CALLMANDER & al., characteristically asymmetrical and have a distinct calycinal 2007). crown, quite unlike the fruits of M. eglandulosum which are Observations. – This species is evidently restricted to dry, larger, globose and have a stylopodium projecting past the trun- semi-deciduous forests in northeastern Madagascar. Its des cri - cate rim of the calyx (Fig. 1). Species status for M. bezavonense ption as a variety of M. eglandulosum H. Perrier was seemingly is further justified by its smaller, oblanceolate leaves (dimen- based on floral characters, which were said to be identical sions 4-5 ϫ 1.2-1.8 cm) in comparison to the larger, ± elliptic between the two taxa (JACQUES-FÉLIX, 1985b: 12). Additional leaves of M. eglandulosum (dimensions 5-7.5 ϫ 2-3 cm). Additional specimens. – MADAGASCAR. Prov. Antsiranana: type loc., 13.XII.1966, st., Service Forestier 27239-bis (P, TEF); Antsiranana II, Andrafi- abe, Saharenana, lieu de récolte Andranomadio-Sahafary, 3 km au Sud-Est du village de Saharenana, 12°36’26”S 49°26’21”E, alt. 227 m, forêt sèche semi- caducifoliée, 10.XI.2004, fr., Rakotondrajaona & al. 342 (CAS, MO, P, TAN). S. loc.: s.d., fr., Service Forestier 29413 (TEF).

Acknowledgements Support from the California Academy of Sciences (John J. Rose Postdoctoral Fellowship) is gratefully acknowledged. Curators of the following herbaria are thanked for allowing access to specimens: BR, G, K, MO, P, TEF. A

References

CALLMANDER, M. W., G. E. SCHATZ, P. P. LOWRY II, M. O. LAIVAO, J. RAHARIMAMPIONONA, S. ANDRIAMBOLOLONERA, T. RAMINOSOA & T. CONSIGLIO (2007). Application of IUCN Red List criteria and assessment of Priority Areas for Plant Conservation in Mada- gascar: rare and threatened Pandanaceae indicate new sites in need of protection. Oryx 41: 168-176. IUCN (2001). IUCN Red List Categories and Criteria (version 3.1). IUCN Species Survival Comission. IUCN, Gland & Cambridge. JACQUES-FÉLIX, H. (1985a). Les Memecyleae (Melastomataceae) de Madagascar (1re partie). Bull. Mus. Natl. Hist. Nat., B, Adanso- nia 6: 383-451. JACQUES-FÉLIX, H. (1985b). Les Memecyleae (Melastomataceae) de Madagascar (2e partie). Bull. Mus. Natl. Hist. Nat., B, Adansonia 7: 3-58. RENNER, S. S., D. TRIEBEL, F. ALMEDA, R. D. STONE, C. ULLOA, F. A. MICHAELANGELI, R. GOLDENBERG & H. MENDOZA (ed.) (2007- 2011). MEL names – a database with names of Melastomataceae. Botanische Staatssammlung München. [http://www.melastomataceae.net/ MELnames/]. B STONE, R. D. (2006). New species of Memecylon L. and Warneckea Gilg (Melastomataceae) from Madagascar and Mayotte. Adanso- nia 28: 337-358.

Fig. 1. – Comparison of fruit morphology (scale bar = 1 cm). A. Fruit of Meme- cylon bezavonense (Jacq.-Fél.) R. D. Stone [Rakotondrajaona & al. 342, CAS] B. Fruit of Memecylon eglandulosum H. Perrier [Madiomanana & al. 196, NU]. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page404

16. CALLMANDER Martin W., Michel O. LAIVAO & Peter B. PHILLIPSON: Pandanaceae specimens relocated from the University of Neuchâtel (NEU), with two new synonymies

Introduction where MWC is now based. The gift comprised some 200 collections of Pandanaceae from Madagascar collected between In the context of a global revision of the family Pandanaceae 1996 and 1999 by MOL, MWC and their colleagues, of which in Madagascar for the Flore de Madagascar et des Comores, eleven specimens are holotypes of Pandanus species described MOL, MWC and their colleagues, have described several species between 1997 and 2007 (HUYNH, 1997, 1999; LAIVAO & al., of Pandanus Parkinson from the island. The initial stages of this 2000; CALLMANDER, 2001; CALLMANDER & al., 2001; CALL- research on the systematics of Pandanaceae in Madagascar were MANDER & LAIVAO, 2002; CALLMANDER & al., 2003a, 2003b; conducted at the University of Neuchâtel (Switzerland) while LAIVAO & al., 2006, 2007). While the holotypes have been incor- MOL and MWC were doing their PhD dissertations under Prof. porated into the general collection at G, several duplicates spec- Philippe Küpfer’s direction. The research on this paleotropical imens (iso types) were already present in the G herbarium and family was linked to the presence of Dr. Kim-Lang Huynh, a consequently these were given to Paris (P) or Missouri (MO) renowned Pandanaceae specialist, who was an Associate Pro- (Table 1). This note provides a complete listing of the locations fessor at the University of Neuchâtel between 1976 and 2000. In of the type specimens for the 11 species described between 1999- 2008, when Prof. Küpfer retired, the entire Pandanaceae collec- 2007 by the authors and their colleagues at NEU. Two species tion at the herbarium of the University (NEU), which is prima- dating back to this period are now regarded as synonyms: Pan- rily focused on the European flora, was permanently transferred danus masoalensis Laivao & Callm. (ϵ Martellidendron to the herbarium of the Conservatoire et Jardin botaniques de la masoalense (Laivao & Callm.) Callm. & Chassot) and Pandanus Ville de Genève (G) which has an active research programme manongarivensis Huynh. We provide a discussion of the delim- and specimen acquisition policy focused on tropical regions, and itation of these species, along with original line drawings.

Table 1. –Pandanaceae holotypes transferred from NEU to G, with current repository of isotypes and currently accepted names. M. = Martellidendron, P. = Pandanus.

Collector(s) and number Herbaria (protologue) Herbaria (current) Type of: Currently accepted name Callmander & Raveloson 114 holo-: NEU iso-: G, P, TAN holo-: G iso-: MO, P, TAN P. gallinarum M. gallinarum Callmander & al. 8 holo-: NEU iso-: TAN holo-: G iso-: TAN P. masoalensis M. karaka Laivao 2 holo-: NEU iso-: TAN holo-: G iso-: TAN P. nosibicus M. karaka Callmander & al. 37 holo-: NEU iso-: G, TAN holo-: G iso-: P, TAN P. callmanderiana P. callmanderiana Callmander & Laivao 57 holo-: NEU iso-: TAN holo-: G iso-: TAN P. humbertii P. humbertii Callmander & al. 33 holo-: NEU iso-: G, P, TAN holo-: G iso-: MO, P, TAN P. kimlangii P. kimlangii Callmander & Wohlhauser 144 holo-: NEU iso-: G, TAN holo-: G iso-: P, TAN P. kuepferi P. kuepferi Callmander & Laivao 59 holo-: NEU iso-: G, P, TAN holo-: G iso-: MO, P, TAN P. marojejicus P. marojejicus Callmander & Wohlhauser 141 holo-: NEU iso-: G, P, TAN holo-: G iso-: P, TAN P. maromokotrensis P. maromokotrensis Gautier & Chatelain 2648 holo-: NEU iso-: G, TAN holo-: G iso-: P, TAN P. manongarivensis P. sambiranensis Callmander & Wohlhauser 145 holo-: NEU iso-: G, P, TAN holo-: G iso-: MO, P, TAN P. tazoanii P. tazoanii

Addresses of the authors: MWC: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. and Conservatoire et Jardin botaniques de la Ville de Genève, ch. de l'Impératrice 1, CP 60, 1292 Chambésy, Switzerland. E-mail: [email protected] MOL: Université de Neuchâtel, Laboratoire de Botanique Evolutive, rue Emile-Argand 11, case postale 158, 2009 Neuchâtel, Switzerland. Current: Ministère de l’Environnement et des Forêts, BP 571, Ampandrianomby, Antananarivo 101, Madagascar. PBP: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. and Muséum national d’Histoire naturelle, Département Systématique et Evolution, UMR 7205, case postale 39, rue Cuvier 57, 75231 Paris, cedex 05, France. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page405

Notes on the flora of Madagascar, 14-21 – 405

Systematics cave pileus of M. masoalense (vs. more flattened in Martellidendron gallinarum (Callm.) Callm. in Taxon 52: M. karaka) (Fig. 1) and its occurrence at higher altitudes 756. 2003. (ca 800 m) on the Masoala Peninsula and Marojejy (vs. altitudes of 0-10 m and occurring only near Maroantsetra). How- Pandanus gallinarum Callm. in Bot. J. Linn. Soc. 137: ϵ ever, several new collections from the Mananara Reserve (at 371. 2001. 200-400 m) are morphologically intermediate and suggest that Typus: MADAGASCAR. Prov. Toamasina: PN de Mana- the limits of M. karaka should be extended to include nara-Nord, 9 km east of Vavary, Beketra, 16º22’51”S M. masoalense, which we now treat as a synonym. 49º40’16”E, 300 m, 3.XI.1999, Callmander & Raveloson 114 (holo-: G [G00303526]!; iso-: MO!, P [P00459652]!, Pandanus callmanderiana Laivao & Buerki in Adansonia 28: TAN!). 278. 2006. Martellidendron karaka (Martelli) Callm. in Taxon 52: 756. Typus: MADAGASCAR. Prov. Toamasina: Tampolo FC, 2003 (Fig. 1). 8 km au Nord de Fénérive-Est, 5 m, 17°17’09’’S 49°24’ ϵ Pandanus karaka Martelli in Webbia 4: 417. 1914. 22’’E, 30.VII.1997, Callmander, Laivao & Wohlhau ser 37 Typus: MADAGASCAR. Prov. Toamasina: Ivoilina, 1907, (holo-: G [G00303537]!; iso-: P!, TAN!). Rollot 3 (holo-: FI [FI000997]!). Pandanus humbertii Laivao, Callm. & Buerki in Adansonia ϭ Martellidendron masoalensis (Laivao & Callm.) Callm. 29: 42. 2007. & Chassot in Taxon 52: 756. 2003. ϵ Pandanus maso - Typus: MADAGASCAR. Prov. Antsiranana: Marojejy PN, alensis Laivao & Callm. in Bot. Helv. 110: 43. 2000. 2e campement, 14°25’52’’S 49°45’41’’E, 780 m, 6.XI.1998, Typus: MADAGASCAR. Prov. Toamasina: Péninsule Callmander & Laivao 57 (holo-: G [G00303538]!; iso-: de Masoala, le long du chemin vers Maroangady, TAN!). 15º19’08”S 50º03’43”E, 430 m, 26.XI.1996, Callman- der, Wohlhauser, Chassot & Laivao 8 (holo-: G Pandanus kimlangii Callm. & Laivao in Bot. Helv. 112: 51. [G00303 540]!; iso-: TAN!), syn. nov. 2002. ϭ Pandanus nosibicus Huynh in Bull. Soc. Neuchâteloise Typus: MADAGASCAR. Prov. Toamasina: Réserve de Sci. Nat. 120: 37. 1997. Typus: MADAGASCAR. Prov. Zahamena, 17°31’28’’S 48°45’29’’E, 1210 m, 24.7.1997, Antsiranana: Lokobe PA, Andranonankomba river, Callmander, Laivao & Wohlhauser 33 (holo-: G [G003035 13°25’13”S 48°19’00”E, 7.V.1996, Laivao 2 (holo-: G 43]!; iso-: MO!, P [P00741927, P00741929]!, TAN!). [G00303539]!; iso-: TAN!) (synonymized by CALL- Pandanus kuepferi Callm., Wohlh. & Laivao in Candollea 58: MANDER, 2001). 66. 2003. Observations. – Pandanus masoalensis Laivao & Callm. Typus: MADAGASCAR. Prov. Antsiranana: Réserve (ϵ Martellidendron masoalensis (Laivao & Callm.) Callm. Intégrale n° IV du Tsaratanana, 14°03’09’’S 48°56’23’’E, & Chassot) was the first species described by MOL and MWC 1750 m, 6.XII.1999, Callmander & Wohlhauser 144 (holo-: in 2000. When we described this species, we followed G [G00303544]!; iso-: P!, TAN!). B.C. Stone’s species concept of Pandanus karaka Martelli Pandanus marojejicus Callm. & Laivao in Candollea 58: 67. (ϵ Martellidendron karaka (Martelli) Callm.) (STONE, 1970), 2003. which included material with grouped infrutescences with up to 13 syncarps enlarged at base (8.5-14 cm in length), with Typus: MADAGASCAR. Prov. Antsiranana: Réserve du pileus holding (2-)3 stigmata at apex. Later, on a visit to Flo- Marojejy, 14°26’51’’S 49°44’34’’E, forêt basse du som- rence Herbarium (FI), MWC was able to examine the type of met de Marojejy, 2100 m, 8.XI.1998, Callmander & Laivao this species and realized that the species had been misunder- 59 (holo-: G [G00303536]!; iso-: MO!, P!, TAN!). stood by STONE (1970: 113, Figs. 4a-e). His interpretation of Pandanus maromokotrensis Callm. & Wohlh. in Adansonia species and the published illustrations correspond to a species 23: 53. 2001. of Pandanus sect. Dauphinensia H. St. John that we described Typus: MADAGASCAR. Prov. Antsiranana: Réserve as P. kimlangii Callm. & Laivao (CALLMANDER & LAIVAO, Intégrale n° IV, massif du Tsaratanana, sur pente dans la 2002) (see CALLMANDER, 2001 for more details). The typical forêt dense, 14°03’09’’S 48°56’23’’E, 1750 m, 29.XI.1999, Martellidendron karaka is characterized by a solitary syncarp Callmander & Wohlhauser 141 (holo-: G [G00303541]!; (c. 20 cm in length) with pileus holding two stigmata forming iso-: P[P00459635]!, TAN!). a cross at apex (Fig. 1c, d). The morphology of the type of M. karaka Martelli is very similar to M. masoalense but initially we chose to maintain the two species based on the con- MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page406

406 – Candollea 66, 2011

D 1 cm

C E 1 m

5 cm

1 cm G H F B

Fig. 1. – Martellidendron karaka (Martelli) Callm. A. Habit; B. Syncarp; C. Lateral view of a drupe; D. Apical view of a drupe; E. Longitudinal section of drupe showing the endocarp; F. Basal part of a leaf; G. Middle part of a leaf; H. Apical part of a leaf. [A, C-H: Callmander & Bemandiny 125, TAN & photo; B: Callmander & al. 7, TAN] [Drawings: R. L. Andriamiarisoa] MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page407

Notes on the flora of Madagascar, 14-21 – 407

Pandanus sambiranensis Martelli in Mém. Inst. Sci. Mada- Madagascar and specially Roger Lala Andriamiarisoa for gascar, Sér. B, Biol. Vég. 3: 54. 1951 (Fig. 2). the fine illustrations. Finally we thank Jean-Michel Gobat Typus: MADAGASCAR. Prov. Antsiranana: massif du (University of Neuchâtel), Laurent Gautier and Pierre-André Manongarivo, bords de Ramena et toute la région du Sam- Loizeau (CJBG) for their assistance with the relocation of birano, 300-1000 m, [13°52’S 48°51’E], XI.1909, Perrier the specimens from NEU to G. Financial support was pro- de la Bâthie 10934 (lecto-: FI [FI001083: 1 sheet exl. st. vided by grants from the U.S. National Science Foundation material]!; isolecto-: P[P00219106]!) (lectotypified by (0743355) and the Andrew W. Mellon Foundation. CALLMANDER & al., 2011). ϭ Pandanus manongarivensis Huynh in Candollea 54: References 153. 1999. Typus: MADAGASCAR. Prov. Antsiranana: CALLMANDER, M. W. (2001). Pandanus subg. Martellidendron Manongarivo, Besinkara, 14º04’S 48º17’E, 300-350 m, (Pandanaceae) part II: revision of sect. Martellidendron Pic. Serm. 13.V.1995, Gautier & Chatelain 2648 (holo-: G [G0030 in Madagascar. Bot. J. Linn. Soc. 137: 353-374. 3535]!; iso-: P!, TAN!), syn. nov. CALLMANDER, M. W., P. CHASSOT, P. KÜPFER & P. P. L OWRY II Observations. – One of us (MWC) was able to study the (2003b). Recognition of Martellidendron, a new genus of Pan- holotype of P. sambiranensis on a visit to FI. This has danaceae, and its biogeographic implications. Taxon 52: 747-762. enabled us to re-interpret the original description of the CALLMANDER, M. W. & M. O. LAIVAO (2002). Révision de Pandanus species and to better understand certain morphological char- sect. Dauphinensia H. St. John (Pandanaceae) à Madagascar. Bot. acters that were not evident in the rather poor type material Helv. 112: 47-67. that we had examined at the Paris Herbarium (P). The CALLMANDER, M. W., M. O. LAIVAO & S. WOHLHAUSER (2003a). Les species is characterized by a narrow drupe (10-15 mm in Pandanus sect. Acanthostyla Martelli (Pandanaceae) d’altitude du Nord de Madagascar, avec description de deux nouvelles width) with a dome-like apex (Fig. 2), and sub-coriaceous espèces. Candollea 58: 63-74. leaves with relatively long marginal prickles (2-3 mm). This morphology is identical to that of the type material of CALLMANDER, M. W., P. B. P HILLIPSON, C. NEPI & M. O. LAIVAO (2011). Lectotypification of the species of Pandanus (Pandana- P. manongarivensis and the latter is therefore considered as ceae) from Madagascar described by U. Martelli and R. Pichi- a synonym. Pandanus sambiranensis is known only from Sermolli. Webbia 64: 3-19. below 800 m in the humid north-western region (the Sam- CALLMANDER, M. W., S. WOHLHAUSER & M. O. LAIVAO (2001). Une birano Domain sensu HUMBERT, 1951, 1955). nouvelle section du genre Pandanus (Pandanaceae) à Madagas- car: Pandanus sect. Tridentistigma. Adansonia 23: 49-57. Pandanus tazoanii Callm. & Wohlh. in Adansonia 23: 50. HUMBERT, H. (1951). Les territoires phytogéographiques du nord de 2001. Madagascar. Compt. Rend. Sommaire Séances Soc. Biogéogr. 246: 176-184. Typus: MADAGASCAR. Prov. Antsiranana: Réserve Inté- grale n° IV, massif du Tsaratanana, sur une crête orientée HUMBERT, H. (1955). Les territoires phytogéographiques de Mada- au sud-ouest, 14°05’48’’S 48°62’00’’E, 1800 m, 6.XII. gascar. Leur cartographie. Année Biol. ser. 3, 31: 439-448. 1999, Callmander & Wohlhauser 145 (holo-: G [G00303 HUYNH, K. L. (1997). The genus Pandanus (Pandanaceae) in Mada- 541]!; iso-: MO!, P [P00741934, P00741939]!, TAN!). gascar part I. Bull. Soc. Neuchâteloise Sci. Nat. 120: 35-44. HUYNH, K. L. (1999). The genus Pandanus (Pandanaceae) in Mada- gascar part III. Candollea 54: 145-170. Acknowledgements LAIVAO, M. O., M. W. C ALLMANDER & S. BUERKI (2006). Sur les The authors would like to thank Prof. Philippe Küpfer Pandanus (Pandanaceae) à stigmates saillants de la Côte Est de and Ernest Fortis (both formerly at the University of Neuchâ- Madagascar. Adansonia 28: 267-285. tel), Chiara Nepi (FI) and Nicolas Fumeaux (CJB) for their LAIVAO , M. O., M. W. CALLMANDER & S. BUERKI (2007). Révision support and help in our research. We are also grateful to de Pandanus sect. Foullioya à Madagascar. Adansonia 29: 39-57. our two colleagues that contributed to the description of LAIVAO, M. O., M. W. CALLMANDER & S. WOHLHAUSER (2000). Une some of the species discussed above: Sven Buerki (Royal espèce nouvelle de Pandanus sect. Martellidendron (Pandanaceae) Kew Gardens, London) and Sébastien Wohlhauser (NGO de la péninsule de Masoala, Madagascar. Bot. Helv. 110: 41-49. Fanamby, Antananarivo) and the local office of the Missouri STONE, B. C. (1970). Observations of the genus Pandanus in Mada- Botanical Garden (MBG) in Antananarivo for assistance in gascar. Bot. J. Linn. Soc. 63: 97-131. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page408

408 – Candollea 66, 2011

2 cm E 1 cm

1 cm

F G D

Fig. 2. – Pandanus sambiranensis Martelli. A. Apical view of a drupe; B. Longitudinal section of drupe showing the endocarp; C. Lateral view of a drupe; D. Syncarp; E. Middle part of a leaf; F. Apical part of a leaf; G. Basal part of a leaf. [A-D: Gautier & Chatelain 2648, TEF; E-H: Perrier de la Bâthie 10934, P] [Drawings: R. L. Andriamiarisoa] MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page409

17. CALLMANDER, Martin W. & Peter B. PHILLIPSON: The genus Vernoniopsis Humbert (Asteraceae) in Madagascar

Introduction genus, Vernoniopsis Humbert, to accommodate them. The genus is distinguished by its characteristic style morphology The family Asteraceae is the largest flowering plant fam- and remarkable Vernonia-like habit and inflorescence. He ily with more than 1,600 genera and 23,000 species currently recognized two subspecies within Vernoniopsis caudata recognized (ANDERBERG & al., 2007). In Madagascar, the fam- (Drake) Humbert: the typical subspecies (which included the ily was studied by the renowned Professor Henri Humbert for material described as Vernonia sanctae-mariae) and V. cau- his PhD. dissertation (HUMBERT, 1923), and this work even- data subsp. lokohensis Humbert (HUMBERT, 1955a). Later, he tually culminated in the publication of the three volumes treat- further divided the typical subspecies into two subvarieties: ment in the Flore de Madagascar et des Comores series (HUM- the typical subvariety and subvar. microcephala Humbert BERT, 1960-1963). At the time of the flora, 521 native species were known from Madagascar (435 [83%] endemic to the (HUMBERT, 1960-1963). He did not recognise varieties. island) placed in 72 native genera (13 [18%] endemic). Rela- An examination of all the available material of Verno niopsis tively little taxonomic research has been undertaken subse- for the Catalogue of the Vascular Plants of Madagascar (MADA- quently on the Malagasy Asteraceae, and the number of cur- GASCAR CATALOGUE, 2011) has enabled us to review taxon delim- rently accepted species has not changed greatly. However itations within the genus. We came to the conclusion that recently, in some problematic groups and particularly those two distinct species should be recognised: V. caudata, which containing large broadly-circumscribed genera, narrower is restricted to the eastern coastal zone from Antalaha to Tao- generic concepts have been adopted. As a result, a number of lagnaro, and the second species, based on V. caudata subsp. new genera occurring in Madagascar have been described lokohense, which is endemic to the Marojejy area, for which recently, and some older genera have been resurrected from we provide a new combination. We designate a lectotype for synonymy. Figures derived from the Catalogue of Vascular V. caudata subsp. lokohensis, and reduce V. caudata subvar. Plants of Madagascar (MADAGASCAR CATALOGUE, 2011), that microcephala to synonymy under V. caudata. We provide obser- we have presented elsewhere (CALLMANDER & al., 2011), vations and a brief discussion of the two species and assessment take into account these taxonomic changes. These figures give of their conservation status using the current IUCN Red List a total of 540 native species, of which 441 (82%) are endemic Categories and Criteria (IUCN, 2001). to Madagascar, divided among 80 native genera, of which 18 (20%) are endemic. The genus Vernonia Schreb. (Tribe Vernonieae Cass.) is Taxonomy and nomenclature characterized by a continuous stigmatic surface over the inner Vernoniopsis Humbert in Mém. Inst. Sci. Madagascar, Sér. B, surface of the style branches. HUMBERT (1955a) discovered Biol. Vég. 6: 154. 1955. that two Malagasy species originally described in Vernonia: (V. caudata Drake and V. sanctae-mariae Drake) possess stig- Type species: V. caudata (Drake) Humbert (ϵ Vernonia matic surfaces that form paired lines (2-banded) on the inner caudata Drake). surface of style branches. He concluded that these species, Vernoniopsis caudata (Drake) Humbert in Mém. Inst. Sci. which he considered to be synonyms, should be transferred to Madagascar, Sér. B, Biol. Vég. 6: 154. 1955. ϵ Vernonia the Tribe Astereae Cass., and he described a new monotypic caudata Drake in Bull. Mus. Hist. Nat. (Paris) 5: 103. 1899.

Addresses of the authors: MWC: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. and Conservatoire et Jardin botaniques de la Ville de Genève, ch. de l’Impératrice 1, CP 60, 1292 Chambésy, Switzerland. E-mail: [email protected] PBP: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. and Muséum national d’Histoire naturelle, Département Systématique et Evolution, UMR 7205 OSEB, case postale 39, rue Cuvier 57, 75231 Paris, cedex 05, France. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page410

410 – Candollea 66, 2011

Typus: MADAGASCAR. Prov. Fianarantsoa: Farafangana, editorial error, and we agree. Certainly the Lantz specimen was 14.IX.1882, fl., Lantz s.n. (holo-: P [P00568726]!; iso-: K seen by Drake and it bears a close resemblance to the specimen [K000273991]!, P [P00568725]!). illustrated by Grandidier, although it is not identical. We there- ϭ Vernonia sanctae-mariae Drake in Bull. Soc. Bot. fore consider the Lantz collection as the type for V. caudata, France 46: 241. 1900. Typus: MADAGASCAR. Prov. following Humbert. Toamasina: Saint-Marie, forêt de Sangous, IX.1909, Vernoniopsis caudata occurs in littoral forest on sand Boivin s.n. (holo-: P [P00435127]!) (synonymized by along the East Coast of Madagascar from Antalaha in the HUMBERT, 1955a). north to Taolognaro in the south (Fig. 2). It is rather vari- able: collections from Antalaha to Tampolo, and on Ile ϭ Vernoniopsis caudata subvar. microcephala Humbert, Sainte-Marie tend to have distinctly longer peduncles than Fl. Madagascar Comores 189: 208. 1960. Typus: those from Farafangana (the type locality) to Taolagnaro MADAGASCAR. Prov. Fianarantsoa: Mananajary, (Fort-Dauphin), but they have similar capitula. On the other zone côtière, [21°13’30’’S 48°21’00’’E], III-IV.1909, hand, collections from near Mananjary have long peduncles fl., Geay 7861 (holo-: P [P00627564]!; iso-: K [K000 and tend to have rather smaller capitula. In the past these 273991]!), syn. nov. variants have been accorded formal taxonomic status, respec- Observations. – Vernoniopsis caudata (ϵ Vernonia caudata) tively as Vernonia sanctae-mariae Drake by DRAKE DEL was illustrated by GRANDIDIER (1897: tab. 476) and then formally CASTILLO (1899b) and Vernoniopsis caudata subvar. micro- described by DRAKE DEL CASTILLO (1899a: 103). The holotype cephala by Humbert. Despite the fact that there are clear designated was “Humblot s.n.”. However, specimens of this morphological trends correlated with geographical distribu- species collected by Humblot have not been found in the Paris tion, we do not believe that our current level of knowledge herbarium. In a type folder, a note written on a copy of Drake permits us to recognize sufficiently clear-cut entities. How- del Castillo’s original published description in Humbert’s ever, the variability requires further study in the field, and it handwriting (Fig. 1) explains that the citation of Humblot as is possible that eventually meaningful infraspecific taxa may the collector was certainly an error and that the accompanying be able to be defined within Vernoniopsis caudata. A com- Lantz collection at P has the words: “Vernonia caudata” in plete listing of specimens examined can be obtained from Drake’s handwriting. Clearly Humbert took this to be simply an the MADAGASCAR CATALOGUE (2011).

Fig. 1. – Reproduction of the P library copy of Bull. Soc. Bot. France (1899b: 242) with Drake del Castillo’s description of Vernonia caudata Drake in his treatment of Malagasy Vernonia, with Humbert’s handwriting annotation (see text). MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page411

Notes on the flora of Madagascar, 14-21 – 411

Lectotypus (designated here): MADAGASCAR. Prov. Antsiranana: Environ d’Andapa, bassin de la Lokoho (NE), [14°39’S 49°38’E], 400-600 m, 25.XI.1948-11.XII. 1948, fl., Humbert & Capuron 21943 (P [P00435119]!; isolecto-: P [P00435120, P00435121]!). Observations. – Vernoniopsis lokohense was originally recognized as a subspecies of V. caudata by HUMBERT (1955a, 1960-1963). This shrubby tree, endemic to the Marorejy area, differs from V. caudata by its much smaller leaves c. 1 ϫ 3-6 cm (vs. 2-3 ϫ 10-12 cm in V. caudata), and its smaller fruits, the achene c. 3 mm long and the pap- pus c. 4 mm long (about twice as long in V. caudata). Ver- noniopsis lokohense is only known from the Lokoho basin in the Marojejy area on bare rocks or in ericoid thicket, and has mostly been collected at elevations over 1300 m (Fig. 2), whereas V. caudata is known from littoral forests on sand from South of Antalaha to Tolagnaro, mostly at under 50 m. This taxon was described on the basis of four collections: Humbert & Capuron 21943, Humbert 23252, Humbert 23556 and Humbert & Cours 23716. We designate one of the three P sheets of Humbert & Capuron 21943 as the lectotype because it is the most representative. Conservation status. – With an EOO of 458 km2, an AOO of 18 km2 and 3 subpopulations, one encompassed in a protected area (Marojejy), V. lokohensis is assigned a preliminary status of Endangered (EN B1ab[i, iii], B2ab[i, iii]) following the IUCN Red List Categories and Criteria (calculation following CALLMANDER & al., 2007).

Acknowledgements Fig. 2. – Distribution of Vernoniopsis Humbert in Madagascar mapped on We thank John Pruski (Missouri Botanical Garden, Saint- phytogeographical map of HUMBERT (1955b): Vernoniopsis caudata (Drake) Humbert (circles) and V. lokohensis (Humbert) Callm. & Phillipson (squares). Louis) for fruitful discussions about Asteraceae taxonomy. Financial support was provided by the U.S. National Science Foundation (0743355) and the Andrew W. Mellon Foundation.

References

ANDERBERG, A. A., B. G. BALDWIN, R. G. BAYER, J. BREITWIESER, Conservation status. – With an EOO of 34,922 km2, an AOO C. JEFFREY, M. O. DILLON, P. ELDENAS, V. FUNK, N. GARCIA- 2 of 78 km and 14 subpopulations, two encompassed in the pro- JACAS, D. J. N. HIND, P. O. KARIS, H. W. LACK, G. NESOM, B. tected area network (Masoala National Park, Manombo Special NORDENSTAM, C. OBERPRIELER, J. L. PANERO, C. PUTTOCK, H. Reserve), V. caudata is assigned a preliminary status of Least ROBINSON, T. F. S TUESSY, A. SUSANNA, E. URTUBEY, R. VOGT, J. Concern (LC) following the IUCN Red List Categories and WARD & L. E. WATSON (2007). Compositae. In: KADEREIT J.W. Criteria (calculation following CALLMANDER & al., 2007). & C. JEFFREY (ed.), The families and genera of vascular plants, 8: 61–588. Vernoniopsis lokohensis (Humbert) Callm. & Phillipson, stat. CALLMANDER, M. W., G. E. SCHATZ, P. P. LOWRY II, M. O. LAIVAO, nov. J. RAHARIMAMPIONONA, S. ANDRIAMBOLOLONERA, T. RAMINOSOA & T. CONSIGLIO (2007). Application of IUCN Red List criteria ϵ Vernoniopsis caudata subsp. lokohensis Humbert in and assessment of Priority Areas for Plant Conservation in Mada- Mém. Inst. Sci. Madagascar, Sér. B, Biol. Vég. 6: 155. gascar: rare and threatened Pandanaceae indicate new sites in 1955. need of protection. Oryx 42: 168-176. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page412

412 – Candollea 66, 2011

DRAKE DEL CASTILLO, E. (1899a). Sur deux genres de Madagascar, de la famille des Composées: Cullumiopsis n. gen. et Centauropsis Bojer. Bull. Mus. Hist. Nat. (Paris) 2: 100-104. DRAKE DEL CASTILLO, E. (1899b). Les Vernonia de Madagascar. Bull. Soc. Bot. France 46: 225-244. GRANDIDIER, A. (1897). Histoire physique, naturelle et politique de Madagascar 6. Paris, Hachette. HUMBERT, H. (1923). Les Composées de Madagascar. Thèse de Doctorat, Université de Caen. HUMBERT, H. (1955a). Une merveille de la nature à Madagascar. Première exploration botanique du massif du Marojejy et ses satel- lites. Mém. Inst. Sci. Madagascar, Sér. B, Biol. Vég. 6: 1-211. HUMBERT, H. (1955b). Les territoires phytogéographiques de Mada- gascar. Leur cartographie. Année Biol. ser. 3, 31: 439-448. HUMBERT, H. (1960-1963). Composées. In: HUMBERT, H. (ed.), Fl. Madagascar Comores 189. Muséum national d’Histoire naturelle, Paris. IUCN (2001). IUCN Red List Categories and Criteria: Version 3.1. IUCN Species Survival Commission, IUCN, Gland, Switzerland; Cambridge, United-Kingdom. MADAGASCAR CATALOGUE (2011). Catalogue of the Vascular Plants of Madagascar [http://www.efloras.org/madagascar]. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page413

18. CRIBB, Phillip, Laurent GAUTIER, Sonia TRIGUI & Louis NUSBAUMER: Two new records of Orchidaceae from Madagascar, with an updated list of species shared with continental Africa

Introduction Distribution and ecology. – South Africa (N KwaZulu- Natal) and western Madagascar. In coastal forest on sand in Didymoplexis verrucosa J. Stewart & Hennessy and Het- South Africa and in forest on tsingy (eroded limestone) in aeria heterosepala (Rchb. f) Summerh., two orchid species Madagascar; 0-100 m. hitherto only known from Africa, are newly reported from Observations. – Jean-Philippe Castillon recently sent a pho- Madagascar. Didymoplexis verrucosa was discovered by Jean- tograph to Kew of a distinctive white holomycotrophic orchid, Philippe Castillon near Bekopaka in the Tsingy de Bemaraha from Bekopaka in the Tsingy de Bemaraha in western Mada- in western Madagascar. It is the second species of the palaeo - gascar (Fig. 1), in which the flower and lip morphology was tropical genus Didymoplexis Griff. to be recorded from clearly visible. It proved to be Didymoplexis verrucosa, an Madagascar. Hetaeria heterosepala was collected during an orchid originally described by STEWART & HENNESSY (1980) inventory of the upper part of Montagne d’Ambre in northern from a plant collected in KwaZulu-Natal in South Africa from Madagascar (TRIGUI, 2010). It represents the first record for coastal forest on sand. Unfortunately, no material of it was pre- Madagascar of this paleotropical genus of ca. 30 species. served as the photographer was on a holiday trip. Didymoplexis verrucosa J. Stewart & Hennessy in Amer. Orchid Soc. Bull. 49: 841. 1980 (Fig. 1). Hetaeria heterosepala (Rchb. f.) Summerh. in Bull. Misc. Typus: SOUTH AFRICA: KwaZulu-Natal Province, Mtun- Inform. Kew 5: 207. 1934. zini Distr.: Farm “Twinstreams”, 15.VIII.1978, Garland ϭ Cheirostylis heterosepala Rchb. f., Otia Bot. s.n. (holo-: NU [NU0015603-0]; iso-: E, K [000242 212], Hamburg: 110. 1881. ϭ Zeuxine heterosepala (Rchb. PRE [PRE0633590-0], S [G-G-6945]). f.) Geerinck in Bull. Jard. Bot. Natl. Belg. 50: 120. Herb, holomycotrophic, glabrous, leafless. Tuber elongate, 1980. moniliform, branching, up to 90 mm long, 20 mm in diam. Typus: CAMEROON: Cameroon Mountains, 3000 f. [915 m], Scape erect, 10-15 cm tall, bearing two sheathing cataphylls XI.1862, Mann 2130 p. p. (holo-: K [K000106606]; iso-: W). at the base, 8- to 12-flowered; bracts spreading, ovate to trian- Herb, perennial up to 20 cm tall. Rhizome fleshy, creeping, gular-ovate, 1-2 mm long. Flowers non-resupinate, white, 10- 3-4 mm in diam., rooting at the nodes. Leaves ovate, acute, 13 mm across; pedicel and ovary 10-16 mm long; sepals and 2.2-4.5 ϫ 1-1.8 cm; petiole and sheath 8-12 mm long. Inflo- petals connate into a short tube at the base; dorsal sepal oblong, rescence densely racemose, 2-6 cm long, up to 20-flowered; obtuse or rounded at the tip, 6-8 3-4 mm, connate for 1-2 mm ϫ bracts lanceolate, acuminate, 5-7 mm long, almost as long as with the petals; lateral sepal similar but connate for 1 mm or the ovary. Flowers usually non-resupinate, with green sepals less to the dorsal sepal; petals oblong, obtuse or rounded, 5-6 ϫ and white petals and lip; dorsal sepal elliptic, obtuse, 4- 5 ϫ 2-4 mm; lip erect-spreading, free, shortly clawed, transversely 1.7-2.5 mm, adnate to the petals to form a hood over the col- triangular, 3-lobed at apex, 5-6 ϫ 8-9 mm; callus of two lines umn; lateral sepals obliquely ovate, similar to dorsal sepal; of small warts from base to tip, with a fleshy yellow ligule on petals obliquely oblong-elliptic, 4-4.5 ϫ 1.5 mm; lip bipartite, the claw; column straight, 4 mm long. Capsule narrowly ovoid- 3.5-5 ϫ 3-3.5 mm; basal part saccate, bearing 2 recurved, elliptic, 20 mm long, 10 mm in diam., pedicel extending erect hooked calli at the base within; the apical lobe transversely to 100-200 mm as the capsule develops. oblong, emarginate, apiculate; column 2.5 mm long, with two linear, acute to obtuse, porrect arms at apex.

Addresses of the authors: PC: The Herbarium, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AB, UK. E-mail: [email protected] LG, ST & LN: Conservatoire et Jardin botaniques de la Ville de Genève and Laboratoire de Systématique Végétale et Biodiversité de l’Université de Genève, ch. de l’Impératrice 1, CP 60, 1292 Chambésy, Switzerland. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page414

414 – Candollea 66, 2011

Fig. 1. – Didymoplexis verrucosa J. Stewart & Hennessy. First observation in Madagascar. [Photo: J.-P. Castillon]

Distribution and ecology. – West, Central and East Tropi- by non-resupinate flowers and a column that lacks the long cal Africa and Northern Madagascar. This species occurs on apical arms. These genera are also each represented by a sin- forest floor in deep shade in dense wet forest; around 990 m gle species in Madagascar. in Madagascar. An annotated list of orchid species shared between mainland Specimens examined. – MADAGASCAR. Prov. Antsiranana, DIANA Africa and Madagascar is provided (Table 1). With the two new Region: Montagne d’Ambre, versants est, alt. 990 m, 10.V.2008, Trigui, records, the total number of shared species amounts to 40, rep- Razanajatovo & Ramandimbimanana 390 (G, TEF, K, P, MO, WAG). resenting about 4.6% of the Malagasy orchid flora (862 species Observations. – A terrestrial orchid (Trigui & al. 390) col- according to CALLMANDER & al., 2011). Looking at their distri- lected on Montagne d’Ambre in northern Madagascar proved bution, 95% of them occur in East Africa; 50% occur in south- to be a new record of Hetaeria heterosepala, previously known ern Africa and 45% are distributed as far as West Africa. Some only from Tropical Africa where it ranges from Liberia to species also occur in the Comoros, the Mascarenes or Cameroon and across the Democratic Republic of the Congo the Seychelles. The two new Malagasy species records pre- to North-east Tanzania. It could only be confused with species sented in this note illustrate very different cases: Didymoplexis of Cheirostylis or Zeuxine, both of which are also characterised verrucosa has a very restricted range in hot dry parts of South MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page415

Notes on the flora of Madagascar, 14-21 – 415

Africa and western Madagascar, where it grows on quite differ- References ent substrates, whereas Hetaeria heterosepala occurs in moist forest throughout a large part of West, Central and East Africa, CALLMANDER, M. W., P. B. P HILLIPSON, G. E. SCHATZ, S. ANDRI- AMBOLOLONERA, M. RABARIMANARIVO, N. RAKOTONIRINA, and a single known locality in northern Madagascar. J. RAHARIMAMPIONONA, C. CHATELAIN, L. GAUTIER & P. P. LOWRY II (2011). The endemic and non-endemic vascular flora of Mada- gascar updated. Pl. Ecol. Evol. 144: 121-125. Acknowledgements CRIBB, P. J. & J. H ERMANS (2009). Field guide to the Orchids of Madagascar. Royal Botanic Gardens, Kew. Financial support for fieldwork allowing collection of HERMANS, J., C. HERMANS, D. DU PUY, P. J. CRIBB & J. BOSSER (2007). Hetaeria heterosepala was provided by grants from Sud Orchids of Madagascar. Royal Botanic Gardens, Kew. Experts Plantes (SEP 348) and Marc Birkigt Fund from the NUSBAUMER, L., P. J. CRIBB & L. GAUTIER (2011). Nervilia gassneri Geneva Academic Society. This fieldwork was conducted Börge Pett. from Africa is conspecific with the Malagasy N. under collaboration between the Département d’Ecologie végé- lilacea Jum. & H. Perrier. Candollea 66: 133-145. tale of the Université d’Antananarivo (DBEV) and the Con- SIEDER, A. & P. STÜTZ. (2011). Finding Disa in Madagascar. Orchid servatoire et Jardin botaniques de la Ville de Genève (CJBG). Review 119: 20-25. We thank Prof. Charlotte Rajeriarison, Dr. Edmond Roger, STEWART, J. & HENNESSY, E. F. (1980). Orchids of Africa: Didymo- Dr. Patrick Ranirison, Mialy Harindra Razanajatovo and Solo- plexis verrucosa – a new saprophytic Orchid from South Africa. tiana Deraharilanto Raman dimbimanana from the DBEV and Amer. Orchid Soc. Bull. 49: 841-847. Madagascar National Parks (MNP). Many thanks to Nicolas TRIGUI, S. M. (2010). Etude floristique et biogéographique des alti- Fumeaux at CJBG for sharing his large bibliographic knowl- tudes supérieures de la Montagne d’Ambre (Nord de Madagascar). edge. We would like to compliment J.-P. Castillon, Reunion Travail de Maîtrise universitaire (Master) en Biologie. Université Island University, for having found Didymoplexis verrucosa de Genève / Laboratoire de Systématique végétale et biodiversité for the first time in Madagascar and to thank him for allowing et Conservatoire et Jardin botaniques de la Ville de Genève. us to reproduce his photograph. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page416

416 – Candollea 66, 2011

Table 1. – Orchid species distributed in both mainland Africa and Madagascar (based on HERMANS & al., 2007; CRIBB & HERMANS, 2009; and further contributions); WA, CA, EA and SA: West, Central, East and southern Africa, respectively.

Species Comments 1. Acampe pachyglossa Rchb. f. EA, SA, also Comoros and Seychelles 2. Angraecopsis parviflora (Thouars) Schltr. CA, EA, also Mascarenes 3. Angraecum eburneum Bory Subsp. eburneum also Comoros, Mascarenes and Seychelles; subsp. superbum (Thouars) H. Perrier also Comoros and Seychelles; var. giryamae (Rendle) Senghas & P. J. Cribb in EA 4. Brachycorythis pleistophylla Rchb. f. WA, CA, EA 5. Brownleea parviflora Lindl. WA, CA, EA 6. Bulbophyllum longiflorum Thouars EA, also Comoros, Seychelles, Mascarenes, Tropical Asia and the SW Pacific islands 7. Calanthe sylvatica (Thouars) Lindl. EA, SA, also Comoros, Seychelles and Mascarenes 8. Cheirostylis nuda (Thouars) Ormerod EA, SA, also Comoros and Mascarenes 9. Corymborkis corymbis Thouars EA, SA, also Mascarenes 10. Didymoplexis verrucosa J. Stewart & Hennessy SA 11. Disa caffra Bolus EA, SA 12. Disa brevicornis (Lindl.) Bolus SA (cf. SIEDER & STÜTZ, 2001) 13. Disperis anthoceros Rchb. f. var. humbertii (H. Perrier) la Croix endemic to Madagascar; var. anthoceros WA, CA, EA (absent from Madagascar) 14. Eulophia clitellifera (Rchb. f.) Bolus WA, CA, EA, SA 15. Eulophia cucullata (Sw.) Steud. WA, CA, EA, SA 16. Eulophia hians var. nutans (Sond.) S. Thomas EA, SA 17. Eulophia livingstoniana (Rchb. f.) Summerh. EA, SA, also Comoros 18. Habenaria cirrhata (Lindl.) Rchb. f. WA, CA, EA, also Comoros 19. Habenaria clareae Hermans EA 20. Hetaeria heterosepala (Rchb. f.) Summerh. WA, CA, EA 21. Liparis caespitosa (Lam.) Lindl. EA, also Comoros, Mascarenes, Tropical Asia and SW Pacific islands 22. Microcoelia aphylla (Thouars) Summerh. EA, also Comoros and Mascarenes 23. Microcoelia exilis Lindl. WA, CA, EA, SA 24. Microcoelia physophora (Rchb. f.) Summerh. EA 25. Nervilia bicarinata (Blume) Schltr. WA, CA, EA, SA, also Comoros, Mascarenes and Tropical Asia 26. Nervilia kotschyi (Rchb. f.) Schltr. Var. kotschyi in WA, CA, EA, also Comoros; var. purpurata (Rchb. f. & Sond.) Börge Pett. in EA, SA 2 7. Nervilia lilacea Jum. & H. Perrier EA, SA; see NUSBAUMER & al. (2011) 28. Nervilia petraea (Sw.) Summerh. WA, CA, EA 29. Nervilia renschiana (Rchb. f.) Schltr. EA, SA 30. Nervilia simplex (Thouars) Schltr. WA, CA, EA, SA, also Tropical Asia to Australia 31. Oberonia disticha (Lam.) Schltr. EA, SA, also Comoros and Mascarenes 32. Oeceoclades decaryana (H. Perrier) Garay & P. Taylor EA, SA, possibly Australia 33. Oeceoclades lonchophylla (Rchb. f.) Garay & P. Taylor EA, also Comoros 34. Oeceoclades maculata (Lindl.) Lindl. WA, CA, EA, also Mascarenes and Tropical America 35. Oeceoclades pulchra (Thouars) P. J. Cribb & M. A. Clem. WA, CA, EA, also Mascarenes, and across to SW Pacific. (cited as Eulophia pulchra in HERMANS & al., 2007) 36. Platycoryne pervillei Rchb. f. EA 3 7. Polystachya concreta (Jacq.) Garay & H. R. Sweet WA, CA, EA, also Comoros, Mascarenes and Tropical Asia and America 38. Polystachya cultriformis (Thouars) Spreng. WA, CA, EA, SA, also Comoros, Mascarenes and Seychelles 39. Polystachya fusiformis (Thouars) Lindl. WA, CA, EA, SA, also Mascarenes and Seychelles (unconfirmed) 40. Satyrium trinerve Lindl. WA, CA, EA, SA MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page417

19. ROGERS, Zachary S., Patrick W. SWEENEY & Laurent GAUTIER: Malagasy Garcinia (Clusiaceae) revisited: new combinations and lectotypifications for Ochrocarpos goudotianus and O. evonymoides and the lectotypification of Garcinia comorensis

Introduction The name Ochrocarpos evonymoides (PLANCHON & TRI- ANA, 1860) was not discussed in the Garcinia synopsis pub- New information bearing on the nomenclatural status of lished in SWEENEY & ROGERS (2008). Three syntypes (Boivin several names belonging to Malagasy species of Clusiaceae 2614bis, Richard 53 [2 sheets], 661), misfiled in the P herbar- has come to light since the publication of the taxonomic syn- ium under the unpublished name “Mammea evonymoides opsis of Malagasy and Comorian Garcinia L. by SWEENEY & (Planch. & Triana) Kosterm.”, have recently been found. An ROGERS (2008). Two new combinations are provided in examination of the original material clearly shows that Ochro- Garcinia for Ochrocarpos evonymoides Planch. & Triana and carpos evonymoides is synonymous with the more recently O. goudotianus Planch. & Triana. Two new synonyms are published Garcinia disepala (VESQUE, 1889), which was the established for Garcinia evonymoides (Planch. & Triana) name used for the species in SWEENEY & ROGERS (2008). A P. Sweeney & Z. S. Rogers, whereas two new synonyms are new combination in Garcinia is provided here for Planchon provided for G. goudotiana (Planch. & Triana) P. Sweeney & and Triana’s earlier name. Z. S. Rogers. Lectotypes are designated for three of the names With regard to the typification of Ochrocarpos evony- treated here (G. comorensis Drake, Ochrocarpos evonymoides, moides, at least two unnumbered collections were referenced O. goudotianus). in the protologue via the citation, “Vohémar (Richard, Boivin in herb. Mus.)”. In the Clusiaceae treatment published for the Garcinia evonymoides (Planch. & Triana) P. Sweeney & Z. S. Flore de Madagascar et des Comores, PERRIER DE LA BÂTHIE Rogers, comb. nova (1951) accepted O. evonymoides as a distinct species based on ϵ Ochrocarpos evonymoides Planch. & Triana in Ann. three collections, which were cited as “Richard 121, 161 [pre- Sci. Nat., Bot. ser. 4, 14: 366. 1860. sumably a typographical error for 661], Boivin (Richard comm.) 2614bis).” The sheet of Boivin 2614bis (P00852334) Lectotypus (designated here): MADAGASCAR: Voyage de M. Boivin, 1847-1852, imm. fr., Boivin 2614bis is the most complete of the examined syntypes and is desig- (P [P00852334]!). nated here as the lectotype of O. evonymoides. DRAKE DEL CASTILLO (1896) published a plate engraved as ϭ Ochrocarpos multiflorus O. Hoffm., Sert. Pl. Madagasc.: “Ochrocarpus evonymoides” and “Madagascar” without any 7. 1881. ϵ Garcinia disepala Vesque, Epha r mosis: tab. explanatory text linking the plate to the homonym published 159. 1889. [nom. illeg.]. Typus: MADAGASCAR. Prov. Antsiranana: Nosy Komba, II.1880, fl., Hildebrandt by Planchon and Triana in 1860. Thus, the two homonyms can- 3337 (lecto-: G [G00090050, stamp 724413]!; isolecto-: not be considered homotypic, but the plant illustrated in the G [G00090049, 2 sheets, G00090050, stamp 724412]!, plate belongs taxonomically to Garcinia evonymoides as first GOET [GOET003975]!, JE [JE00002367, JE00002368]!, indicated by PERRIER DE LA BÂTHIE (1951). K [K000380200]!, M [M0109452]!, P [P030820, P030821, P00592816]!, WU [WU032899]!) (lectotypified by SWEENEY & ROGERS, 2008: 532), syn. nov.

Addresses of the authors: ZR: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. Email: [email protected] PWS: Yale University Herbarium, Peabody Museum of Natural History, 170 Whitney Avenue, New Haven, Connecticut, 06511, U.S.A. LG: Conservatoire et Jardin botaniques de la Ville de Genève and Laboratoire de Systématique Végétale et Biodiversité de l’Université de Genève, ch. de l’Impératrice 1, CP 60, 1292 Chambésy, Switzerland. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page418

418 – Candollea 66, 2011

Garcinia goudotiana (Planch. & Triana) P. Sweeney & part of the same collection and is thus regarded as an isolecto- Z. S. Rogers, comb. nova type. P00568801, a scantly labeled sheet composed of five ϵ Ochrocarpos goudotianus Planch. & Triana in Ann. detached leaves, was cited as the “holotype” of O. goudotianus Sci. Nat., Bot. ser. 4, 14: 365. 1860. in SWEENEY & ROGERS (2008). Although the label may bear the annotation of the scientific name in Triana’s handwriting and the Lectotypus (designated here): MADAGASCAR. Prov. Toa- leaves on the sheet correspond taxonomically to Garcinia goudo- masina: Tamatave [Toamasina], VII.1830, imm. fr., tiana, it is very unlikely that P00568801 represents type mate- Goudot s.n. (P [P00592815]!; isolecto-: G [G00090054]!; rial of Ochrocarpos goudotianus as it was part of the herbarium MPU [MPU014394] image seen). of Petit-Thouars and bears none of the provenance information = Garcinia melleri Baker in J. Linn. Soc., Bot. 20: 92. mentioned in the protologue, nor any mention of a collector. 1883. Typus: MADAGASCAR. Prov. Toamasina: Between SWEENEY & ROGERS (2008) cited a P sheet of Humblot 144 Tamatave [Toamasina] and Antananarivo, VII-VIII.1862, as the “holotype” of Garcinia comorensis Drake. At the time imm. fr., Meller s.n. (lecto-: K [K000240242]!; isolecto-: of the publication a barcode number was not available to cite, P [P030819]!) (lectotypified by SWEENEY & ROGERS, but the sheet has since been barcoded as P00592817. Two 2008: 533), syn. nov. branches are affixed to the sheet (one with flower buds and the = Garcinia comorensis Drake in Bull. Mens. Soc. Linn. other is sterile). In 2009 we located a second sheet of Humblot Paris: 1221. 1896. ϵ Ochrocarpos comorensis (Drake) 144 at P (P00701479) that has three branches affixed to the R. Vig. & Humbert in Rev. Gén. Bot. 25bis: 634. 1914. sheet (one branch with several open staminate flowers and Lectotypus (designated here): MADAGASCAR. Prov. buds, one branch only with flower buds, and one completely Toamasina: Lac de Nossi-Vé, 11.IV.1882, fl., sterile branch). Here we designate the lectotype of G. comoren- Humblot 144 (P [P00701479, branch labeled A1]!; sis as the largest and most complete branch (labeled “A1”) on isolecto-: K [K000240247]!, P [P00592817, P00701 P00701479. The rest of material associated with P00701479 479, the remaining specimens]!), syn. nov. and both specimens affixed to P00592817 are regarded as This species was mistakenly referred to as Garcinia mel- isolectotypes. leri Baker (BAKER, 1883) in SWEENEY & ROGERS (2008) because of the overlooked priority of Ochrocarpos goudo- Acknowledgements tianus (PLANCHON & TRIANA, 1860). The new combination G. goudotiana (Planch. & Triana) P. Sweeney & Z. S. Rogers The authors thank Gérard Aymonin (P) for locating the is proposed here. misfiled syntypes of Ochrocarpos evonymoides, Cécile Aupic (P) and Peter Schäfer (MPU) for their curatorial assistance, In the protologue of Ochrocarpos goudotianus (PLANCHON and Danièle Domeyne (MPU) and Gregory Wahlert (MO) & TRIANA, 1860: 365), three elements provide information for providing digital images of a few type specimens. Peter regarding the provenance of the original material: the manuscript Stevens (MO) kindly reviewed the manuscript. name “Tovomita Goudotiana, Cambess. mss. in herb.” cited in synonymy, the locality of the collection “Madagascar (Goudot in herb. Mus. Par. et Cambess., ann. 1830)”, and the following References sentence in the diagnosis “Arbor excelsa, Hippocastani facie BAKER, J. G. (1883). Contributions to the Flora of Madagascar. Part (Goudot).” These statements indicate that at least two specimens I. Polypetalae. J. Linn. Soc., Bot. 20: 87-158. were used for the description, namely Goudot material deposited DRAKE DEL CASTILLO, E. (1896). Ochrocarpos evonymoides. In: at P and MPU (Herb. Cambessèdes). Two sheets (MPU014394, GRANDIDIER, A., Hist. Phys. Madagascar: tab. 353. Imprimerie P00592815), labeled with information mentioned explicitly in nationale, Paris. the protologue, are clearly identifiable as syntypes (e.g., MPU PERRIER DE LA BÂTHIE, H. (1951). Guttifères. In: HUMBERT, H. (ed.), sheet annotated as “Tovomita Goudotiana Camb. ined.”, P sheet Fl. Madagascar Comores 136. Muséum national d’Histoire annotated as “Arbor excelsa Hippocastani facie”). naturelle, Paris. Comparing the two sheets, P00592815 has the more com- PLANCHON, J. E. & J. T RIANA (1860). Mémoire sur la famille des plete specimen and includes the most detailed locality and date Guttifères. Ann. Sci. Nat., Bot. ser. 4, 14: 226-367. (viz Tamatave, juillet 1830). The entire sheet is designated as the SWEENEY, P. W. & Z. S. R OGERS (2008). Nomenclatural notes on lectotype of O. goudotianus. A further specimen collected by Garcinia (Clusiaceae) from Madagascar and the Comoros. Novon Goudot in 1830 is deposited in G (G00090054). It bears the anno- 18: 524-537. tation: “Grand arbre, ayant le port d’un maronnier et à peu près VESQUE, J. (1889). Epharmosis, sive, materiae ad instruendam anato- son volume moyen. Tamatave. Juillet.” and can be considered miam systematis naturalis, part 2. Delapierre, Vincennes. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page419

20. CALLMANDER, Martin W., Bruno VILA, John DRANSFIELD & Henk BEENTJE: The legacy of Henri Jumelle in Marseille: an overlooked collection of Palms from Madagascar

Introduction The main themes of Jumelle’s research on the Malagasy Flora were on certain families, notably Apocynaceae, Arecaceae The Colonial Museum of Marseille was founded in 1893 and Rubiaceae. Among these, Arecaceae was certainly the by Edouard Heckel to illustrate to the general public the pros- most important numerically with the publication of 90 taxa perity of the French colonies (JUMELLE, 1916). In the same (5 genera, 77 species and 8 varieties) and 11 combinations. year, Heckel initiated the Annales de l’Institut botanico- The first work that came out on the Malagasy palms (JUMELLE géologique colonial de Marseille where researchers were able & PERRIER DE LA BÂTHIE, 1912) was followed by 23 other arti- to publish the results of numerous studies linked with scien- cles between 1915 and 1933 (see CHOUX, 1936 for a complete tific material brought back from the colonies. A year later Henri list), culminating in the palm volume for the Flore de Mada- Jumelle arrived in Marseille as an Associate Professor at the gascar et des Comores, which was published after Jumelle’s HOUX Science Faculty of the University (C , 1936) (Fig. 1). death (JUMELLE & PERRIER DE LA BÂTHIE, 1945). This long He immediately followed Heckel’s areas of interest and pub- series of scientific publications was based mainly on the 247 lished the results of his research on aspects of applied botany, palm collections made by Perrier de la Bâthie. notably on potential industrial and food sources such as oil-pro- ducing plants. He became Professor of General Botany when Henri Jumelle’s palm collection in Marseille Heckel retired in 1913 and Director of the Museum Colonial de Marseille after Heckel’s death in 1916 (CHEVALIER, 1936). Initially situated at the Boulevard des Dames in Marseille, the Colonial Museum was transferred after World War I to the In 1901, Jumelle initiated a correspondence with the French 1st floor of the Natural Science building at the Science Faculty explorer and botanist Henri Perrier de la Bâthie, and a fruitful (AILLAUD, 2002, 2006). After World War II, Pierre Choux collaboration started between the two scientists that introduced incorporated the Jumelle herbarium with the herbarium main- Jumelle to the Malagasy Flora. Perrier de la Bâthie made a tained by the city of Marseille (MARS) at the University of total of more than 20,000 plant collections in Madagascar Provence, where it is still housed today (http://www.univ- between 1898 and 1928 (HUMBERT, 1958) accompanied by provence.fr/). In the 1990s, HB and JD tried in vain to obtain precise field notes and wrote several volumes of the Flore de information on the fate of the palm collections at the Colonial Madagascar et des Comores, while Jumelle studied the mate- Museum of Marseille when they were working towards their rial in his Laboratory in Marseille (AILLAUD, 2002). Jumelle monograph of the palms of Madagascar (DRANSFIELD & BEEN- described 328 new taxa (12 genera, 293 species, TJE, 1995). The presence in the Paris Herbarium of many 19 varieties and 4 forma) and 31 new name combinations in of the collections cited in protologues seemed to add evidence 23 different families before his death in 1935, 230 (c. 65%) of to what they had been told – that the Colonial Museum palm which were published jointly with Perrier de la Bâthie (MADA- collection had been incorporated into the Paris Herbarium. GASCAR CATALOGUE, 2011). Among the genera published Yet, some types (e.g., that of Ravenea latisecta Jum.) were by Jumelle, six are still accepted. These include: Arophyton missing. Other French herbaria were suggested, and HB and Jum. and Carlephyton Jum. (Araceae); Ischnolepis Jum. & JD attempted to track down the missing specimens by corre- H. Perrier and Secamonopsis Jum. (Apocynaceae); Beccario- sponding with herbarium curators, but to no avail. The only phoenix Jum. & H. Perrier and Masoala Jum. (Arecaceae). specimens studied by Jumelle & Perrier de la Bâthie that were

Addresses of the authors: MWC: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. & Conservatoire et Jardin botaniques de la Ville de Genève, ch. de l’Impératrice 1, CP 60, 1292 Chambésy, Switzerland. E-mail: [email protected] BV: Université de Provence, IMEP UMR CNRS 6116, Centre Saint-Charles, case 4, 3, place Victor Hugo, 13331 Marseille Cedex 03, France. JD, HB: The Herbarium, Library, Art & Archives Royal Botanic Gardens Kew, RichmondSurrey TW9 3AB, England, United Kingdom. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page420

420 – Candollea 66, 2011

Fig. 1. – Personal of the Colonial Institute and Colonial Museum of Marseille in 1900. Henri Jumelle (the second from the left) is sitting next to Edouard Heckel (on his left) (repro- duced from HECKEL & al., 1900).

available to them when they published the complete revision – Perrier de la Bâthie 11735 (Syntype [ST] of Neo dypsis of the family for Madagascar were the specimens deposited heteromorphus Jum. ϵ Dypsis heteromorpha (Jum.) in P and duplicates subsequently distributed to TAN, and there- Beentje & J. Dransf.). fore several important specimens could not be examined. – Perrier de la Bâthie 11982 (ST of Chrysalidocarpus Aware of the important historical significance of the oleraceus Jum. & Perrier ϭ Dypsis madagascariensis Jumelle legacy, in 2009 BV managed to alert the botanical (Becc.) Beentje & J. Dransf.). community via tela-botanica, a French botanical community – Perrier de la Bâthie 11986, 11987 (ST of Neophloga network (http://www.tela-botanica.org), to the existence of the lanceolata Jum. ϵ Dypsis jumelleana Beentje & important botanical collection at the herbarium of the Univer- J. Dransf.). sity of Provence. Two years later, MWC made a visit to MARS – Perrier de la Bâthie 12008 (Holotype of Ravenea lati- and was able to confirm that the collections studied by Jumelle secta Jum.). are still deposited at MARS, including a nearly complete set of palms collected by Perrier de la Bâthie. The MARS collec- – Decary 10295, 10297, Perrier de la Bâthie 18654 (ST tion of Malagasy palms includes 208 collections of Perrier de of Neodypsis decaryi Jum. ϵ Dypsis decaryi (Jum.) la Bâthie, six by Raymond Decary, and one each of Charles Beentje & J. Dransf.). I. Forsyth-Major, Henri Humbert and George F. Scott-Elliot, The re-discovery of the Malagasy palm collection housed making a total of 217 collections. Among the MARS collec- at the University of Provence will provide a new stimulus to tion, 76 (35 %) are nomenclatural types and 69 represent gath- research on Malagasy Palms, and will eventually allow us to erings that are not present at the P herbarium. The main inter- better understand several species that were poorly known due est remains in the types not deposited at P and therefore not to unavailability of the relevant collections. Several nomen- seen by the specialists: clatural and taxonomic works will certainly result from this discovery. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page421

Notes on the flora of Madagascar, 14-21 – 421

Acknowledgements Our research was supported by grants from the U.S. National Science Foundation (0743355) and the Andrew W. Mellon Foundation.

References

AILLAUD, G. (2002). La botanique, de la science aimable à la botanique appliquée soutien du Marseille colonial. In: Marseille, 2600 ans de découvertes scientifiques. III. Découvreurs et découvertes: 241-290. Publications de l’Université de Provence. AILLAUD, G. (2006). Edouard Heckel et l’institut colonial de Mar- seille. In: Désirs d’ailleurs. Les expositions coloniales de Mar- seille 1906 et 1922: 45-53. Editons Alors Hors du Temps. CHEVALIER, A. (1936). Professeur Henri Jumelle (1866-1935). Rev. Bot. Appl. Agric. Colon. 16: 165-172. CHOUX, P. (1936). Henri Jumelle (1866-1935). Ann. Fac. Sci. Mar- seille 9: 1-158. DRANSFIELD, J. & H. BEENTJE (1995). The Palms of Madagascar. Royal Botanic Gardens, Kew & International Palm Society, Kew. HECKEL, E., H. JUMELLE, E. JACOB DE CORDEMOY, L. LAURENT & M. EBERLIN (1900). L’Institut et le Musée Colonial de Marseille. Notice publiée à l’occasion de l’Exposition Universelle de 1900. H. Roberge, Paris. HUMBERT, H. (1958). Henri Perrier de la Bâthie (1873-1958). J. Agric. Trop. Bot. Appl. 5: 863-867. JUMELLE, H. (1916). Le Dr. Heckel. Ann. Mus. Colon. Marseille ser. 3, 3: 5-13. JUMELLE, H. & H. PERRIER DE LA BATHIE. (1912). Un nouveau genre de palmiers de Madagascar. Compt. Rend. Hebd. Séances Acad. Sci. 155: 410-411. JUMELLE, H. & H. PERRIER DE LA BATHIE (1945). Palmiers. In: HUM- BERT, H. (ed.), Fl. Madagascar Comores 30. Muséum national d’Histoire naturelle, Paris. MADAGASCAR CATALOGUE (2011). Catalogue of the vascular plants of Madagascar [http://www.efloras.org/madagascar]. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page422

21. SCHATZ, George E. & Porter P. LOWRY II: A new name for the endemic species of Diospyros L. (Ebenaceae) in the Seychelles

Introduction Baill.). It seems likely that as Baillon gathered together material for his paper on Olax, a single duplicate of Pervillé 36 In the course of evaluating Ebenaceae for the Catalogue became separated from the others, and was mistakenly given of the Vascular Plants of Madagascar project (MADAGASCAR the printed label with the footer “Madagascar Voyage de CATALOGUE, 2011), we discovered a folder in the Paris herbar- M. Boivin 1847-1852”, which the holotype of Olax pervil- ium containing the specimen that served as the type of Olax leana (Pervillé 513) also bears, as well as the handwritten boiviniana Baill., a collection that is referred to in the proto- labels indicating the provenance as Ile Sainte Marie. It would logue only as «In S. Maria madagasc. a Bernier olim lecta et appear that many of the collections of Pervillé from the cum Boivin, ann. 1846, communicatum» (BAILLON, 1862). The Seychelles and Madagascar were sent by Bernier to Boivin, specimen bears three original labels along with a handwritten and that some of them bear a second number associated with annotation label by René Capuron dated 17 July 1960. The either Bernier or Boivin (SCHATZ & LOWRY, in press). three original labels all indicate the provenance of the collec- As Capuron correctly discerned, the type specimen of Olax tion as Ile Sainte Marie (Madagascar), as follows: 1) a hand- boiviniana clearly represents part of the same gathering as the written label with “Olax, Ste Marie B”; 2) another handwrit- four Paris duplicates of Pervillé 36, which are isotypes of Maba ten label with “Ste Marie de Madagascar, Bernier comm. seychellarum, one of which is labeled “Ile Mahé 13 fév. 1840 1846”; and 3) a printed label with the heading “Herb. Mus. Arbrisseau de 10 à 12 pieds de hauteur Montagnes près de Paris” and the footer “Madagascar Voyage de M Boivin 1847- Cascade”. Insofar as the Kew holotype of Maba seychellarum 1852”, and in Baillon’s handwriting “Olax? Boiviniana H. Bn. (determined as such because of a handwritten annotation (Comm. à Boivin par Bernier) Ste Marie.” Capuron’s annota- “Maba seychellarum, sp. nov.” followed by Hiern’s initials tion indicates that the specimen is not an Olax, but rather rep- “W.P.H.”) has only a Paris printed label with the heading resents Maba seychellarum Hiern (1873), and further suggests “Herb. Mus. Paris” and the footer “Voyage de M. Pervillé. Iles that it is a mislabeled duplicate of Pervillé 36, the type of Seychelles. 1841.”, Hiern clearly must have examined some M. seychellarum. Capuron concluded that a new combination of Paris duplicates of Pervillé 36 because he included the infor- “Maba boiviniana” was thus necessary, but never formally mation on the specific island and locality, date, and habit in published the name. the protologue, but evidently did not see the holotype of Olax In their revision of Olax for Madagascar and the Comoro boiviniana Baill. with its labels incorrectly indicating Ile Ste Islands, ROGERS & al. (2006) cited Olax boiviniana as an Marie. Furthermore, as Capuron also correctly determined, excluded name, stating that they were unable to find the orig- since Baillon’s epithet “boiviniana” is the oldest available inal material, the reason for which has now become evident. for the Seychelles taxon, a new combination is necessary. Mod- In addition to Olax boiviniana, BAILLON (1862) described three ern generic circumscription within Ebenaceae now recognizes other species of Olax from Madagascar that have also proved Maba as a synonym of Diospyros (DUNGJAI & al., 2006; to be species of Diospyros, two of which are also based upon SCHATZ & LOWRY, in press), and we therefore propose the Pervillé collections (Olax pervilleana Baill.; Olax quercina following new combination.

Addresses of the authors: GES: Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, 63166-0299, U.S.A. E-mail: [email protected] PPL: Missouri Botanical Garden, P.O. Box 299, St. Louis, MO, 63166-0299, U.S.A. and Département Systématique et Evolution (UMR 7205), Muséum National d’Histoire Naturelle, case postale 39, 57 rue Cuvier 57, 75231 Paris CEDEX 05, France. MEP Candollea 66-2 COMPLET_. 03.01.12 08:50 Page423

Notes on the flora of Madagascar, 14-21 – 423

Diospyros boiviniana (Baill.) G. E. Schatz & Lowry, comb. nova ϵ Olax boiviniana Baill. in Adansonia 3: 121. 1862. Typus: SEYCHELLES: Pervillé 36 (holo-: P! [P00722532] [mislabeled as coming from Ile Sainte Marie, Madagas- car]; iso-: K [000350813] image seen, L [0049302] image seen, P [P00722526, P00722528, P00722530]!). ϭ Maba seychellarum Hiern in Trans. Cambridge Philos. Soc. 12: 130. 1873. ϵ Diospyros seychellarum (Hiern) Kosterm. in Blumea 23: 469. 1977. Typus: SEYCHELLES: Pervillé 36 (holo-: K [000350813] image seen; printed “Herb. Mus. Paris” label with “Voyage de M. Pervillé. Iles Seychelles. 1841.”]; iso-: L [0049302] image seen, P [P00722530: “Ile Mahé 13 fév 1840 Arbrisseau de 10 a 12 pieds de hauteur Montagnes près de Cascade”, P00722526, P00722528]!).

References

BAILLON, H. (1862). Deuxième mémoire sur les Loranthacées. Adan- sonia 3: 50-128. DUANGJAI, S., B. WALLNÖFER, R. SAMUEL, F. MUNZINGER & M. W. CHASE (2006). Generic delimitation and relationships in Ebenaceae sensu lato: evidence from six plastid DNA regions. Amer. J. Bot. 93: 1808-1827. HIERN, W. P. (1873). A monograph of Ebenaceae. Trans. Cambridge Philos. Soc. 12: 27-300. MADAGASCAR CATALOGUE (2011). Catalogue of the vascular plants of Madagascar [http://www.efloras.org/madagascar]. ROGERS, Z. S., V. MALÉCOT & K. G. SIKES (2006). A synoptic revision of Olax L. (Olacaceae) in Madagascar and the Comoro Islands. Adansonia 28: 71-100. SCHATZ, G. E. & P. P. LOWRY II (in press). Nomenclatural notes on Malagasy Diospyros L. (Ebenaceae). Adansonia 33.