TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Phylogeny and systematics of African Melastomateae
(Melastomataceae)
Marie Claire Veranso-Libalah,1,2,3 Robert Douglas Stone,4 Augustina G.N. Fongod,1,5
Thomas L.P. Couvreur5,6 & Gudrun Kadereit2,3
1
2
3
4
5
Department of Botany and Plant Physiology, University of Buea, P.O. Box 63, Buea, Cameroon
Institut für Molekulare Physiologie, Johannes Gutenberg-Universität Mainz, 55099 Mainz, Germany
Institut für Organismische und Molekulare Evolutionsbiologie, Johannes Gutenberg-Universität Mainz, 55099 Mainz, Germany
School of Life Sciences, University of KwaZulu-Natal, Private Bag X01, Pietermaritzburg 3209, South Africa
Université de Yaoundé I, Ecole Normale Supérieure, Département des Sciences Biologiques, Laboratoire de Botanique
Systématique et d’Ecologie, B.P. 047, Yaoundé, Cameroun
6 Institut de Recherche pour le Développement, UMR-DIADE, BP 64501, 34394 Montpellier cedex 5, France
Author for correspondence: Marie Claire Veranso-Libalah, mario_clario@yahoo.ca
ORCID MCVL, http://orcid.org/0000-0001-7847-1740; RDS, http://orcid.org/0000-0003-2752-7788
DOI https://doi.org/10.12705/663.5
Abstract Melastomateae are the most species-rich (185 spp./13 gen.) and morphologically diverse tribe within the African
Melastomataceae. In this study we present the first in-depth phylogenetic analysis of African Melastomateae using sequences
generated from one nuclear (nrITS) and two plastid spacers (accD-psaI, psbK-psbL) sampling 183 accessions representing 75
African, 10 Madagascan and 7 Asian species and a broad outgroup sampling. We infer phylogenetic relationships using maximum
likelihood, maximum parsimony and Bayesian approaches and propose a revised generic classification of African Melastomateae
based on our molecular trees in combination with a careful reassessment of morphological variation. Our phylogenetic analyses
support the monophyly of Old World Melastomateae except Dinophora which clearly falls outside Melastomateae. Asian and
Madagascan endemic genera of Melastomateae are nested among the African genera. Of the African genera only two monophyletic groups, Dichaetanthera and Tristemma, are recovered in addition to the monospecific Pseudosbeckia lineage, while
Antherotoma, Dissotis and Heterotis as currently circumscribed are polyphyletic. However, the subgenera and sections of these
genera are mostly recovered as morphologically distinct and molecularly well-supported clades. Isomorphism versus dimorphism
of the stamens is found to be an unreliable character when delimiting genera in Melastomateae. We propose the reinstatement
of Argyrella and Dupineta, a new status at generic rank for Dissotis subg. Dissotidendron, a broadened circumscription of
Guyonia (to include the species previously placed in Heterotis sect. Cyclostemma), and description of a new, monospecific genus
Anaheterotis for the Heterotis pobeguinii lineage. Heterotis is maintained at generic rank but with a narrower circumscription
in comparison to earlier treatments. We accommodate the rare Osbeckia porteresii lineage from Mt. Nimba (Guinea) within
an expanded Melastomastrum. To complete our generic realignment, a total of 42 new combinations are effected at the species
level and below, and a key is provided for the 12 genera presently recognised in African Melastomateae. Denser taxonomic
sampling and sequencing of more loci will be needed to resolve relationships within the large clade of Dissotis s.str. and allies
(including Antherotoma as well as three West African species erroneously placed in the otherwise Asian Osbeckia).
Keywords Africa; Dissotis; Heterotis; Melastomataceae; Melastomateae; molecular phylogenetics; systematics; taxonomy
Supplementary Material Electronic Supplement (Figs. S1 & S2) and DNA sequence alignment files are available in the
Supplementary Data section of the online version of this article at http://ingentaconnect.com/content/iapt/tax
INTRODUCTION
Melastomataceae s.l. (incl. Memecylaceae, cf. APG, 2009)
constitute a characteristic element of numerous tropical ecosystems and range among the largest tropical plant families worldwide. According to Renner & al. (2007–), Melastomataceae
comprise ca. 170 genera and 5000 species with greatest diversity in the Neotropics (ca. 3000 spp.), followed by Southeast
Asia (ca. 1000), with moderate diversities in Madagascar (ca.
330) and Africa (ca. 320). They exhibit a wide variety of life
forms from trees, shrubs, woody climbers to herbs and grow
in a wide range of habitats, from lowland to montane tropical
forests, in savannas as well as in disturbed secondary vegetation (Renner, 1993; Clausing & Renner, 2001). However, their
systematics and phylogenetic relationships, particularly in the
Paleotropics, remain poorly understood.
The tribe Melastomeae was first proposed by Bartling
(1830) to include species having poricidal anthers. This tribe
Received: 24 Sep 2016 | returned for (first) revision: 6 Dec 2016 | (last) revision received: 26 Feb 2017 | accepted: 27 Feb 2016 || publication date(s):
online fast track, n/a; in print and online issues, 23 Jun 2017 || © International Association for Plant Taxonomy (IAPT) 2017
584
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
was later split based on biogeography, anther structure and
nature of pubescence (Triana, 1865, 1872; Cogniaux, 1891).
Cogniaux (1891) placed the African and Asian genera in
Osbeckieae and the American genera in Tibouchineae. This
classification scheme was widely accepted and used (Van
Tieghem, 1891; Krasser, 1893; Jacques-Félix, 1981, 1995; Van
Vliet & al., 1981) until Renner (1993) merged these two segregate tribes resurrecting Melastomeae [from here on referred
to as “Melastomateae”. The generic name Melastoma L. is
of Greek origin and is a neuter noun. The plural of stoma is
stomata, and “Melastomataceae” is thus the correct family
name corresponding to the generic name Melastoma, not
“Melastomaceae” as used by Cogniaux (1981) and other earlier authors. As such, “Melastomateae” is the correct tribal
name corresponding to the generic name Melastoma, not
“Melastomeae” as used by Renner (1993) and other later authors. See the online Indices Nominum Supragenericorum
Plantarum Vascularium by Reveal (2011–)].
This pantropical tribe now comprises about 870 species in
47 genera (Michelangeli & al., 2013). In Africa, Melastomateae
are the most diverse melastome tribe with about 185 species in
13 genera (Wickens, 1975; Renner, 1993) although only ten genera were recognised by Jacques-Félix (1995). While currently
ca. 66% of the African species belong to the diverse genus
Dissotis Benth., seven genera comprise only 2–15 species and
three are monospecific (Jacques-Félix, 1995) (Table 1).
Morphologically, African Melastomateae are diverse, offering a multitude of characters for generic and species delimitation (Figs. 1, 2). Growth forms range from herbs to trees,
plants are glabrous to densely pubescent, and the number of
longitudinal leaf nerves varies between 3 and 11. Floral morphology has been particularly important in delimiting taxa,
e.g., floral merosity (4- or 5- or rarely 6-merous); persistent
or caducous calyx-lobes; presence or absence of intersepalar
appendages; type and nature of calyx-tube pubescence; stamens isomorphic or dimorphic; pedoconnective and staminal
appendages present or absent; arcuate or linear pedoconnectives; and lobed, bilobed or clavate to 3-cuspidate staminal
appendages (Figs. 1, 2). Capsule and seed morphology have
also offered diagnostic characters for species delimitation. On
the basis of this morphological diversity, several generic and
infrageneric classification schemes have been proposed (Table
1) for African Melastomateae (Bentham, 1849; Naudin, 1850,
1853; Triana, 1872; Cogniaux, 1891; Fernandes & Fernandes,
1969; Jacques-Félix, 1981, 1995; Van Vliet & al., 1981).
Bentham (1849) recognised only four genera: Heterotis
Benth., Dissotis, Tristemma Juss. and Osbeckia L., and he proposed the first infrageneric classification of Heterotis with
four sections, though only two of them (sect. Heterotis, sect.
Cyclostemma Benth.) were later recognised by Jacques-Félix
(1981, 1995). Triana (1872) recognised five additional genera (Dichaetanthera Endl., Antherotoma Hook.f., Guyonia
Naudin, Dinophora Benth., Nerophila Naudin), and he also
transferred all species treated in Heterotis sect. Leiocalyx
Planch. ex Benth., H. sect. Wedeliopsis Planch. ex Benth. and
Melastomastrum Naudin to Tristemma. Fernandes & Fernandes
(1969) recognised four additional genera (Pseudosbeckia
A.Fern. & R.Fern., Cailliella Jacq.-Fél., Dionychastrum
A.Fern. & R.Fern., Derosiphia Raf.) and transferred all species
of Heterotis to Dissotis. They also proposed a new infrageneric
classification of Dissotis which consisted of six subgenera, i.e.,
subg. Dissotis, subg. Dupineta (Raf.) A.Fern. & R.Fern., subg.
Dissotidendron A.Fern. & R.Fern., subg. Osbeckiella A.Fern.
& R.Fern., subg. Heterotis (Benth.) A.Fern. & R.Fern. and
sect. Argyrella (Naudin) A.Fern. & R.Fern., and within subg.
Dissotis they recognised four sections, i.e., sect. Dissotis, sect.
Macrocarpae A.Fern. & R.Fern., sect. Squamulosae A.Fern. &
R.Fern. and sect. Sessilifoliae A.Fern. & R.Fern. This classification scheme by Fernandes & Fernandes (1969) was further
modified by Jacques-Félix (1981, 1995) who proposed an additional subgenus Paleodissotis Jacq.-Fél. within Dissotis, and
transferred the species of subg. Osbeckiella to Antherotoma
and reinstated Heterotis at generic rank. Within Heterotis he
recognised three sections, i.e., sect. Heterotis, sect. Argyrella
(Naudin) Jacq.-Fél. and sect. Cyclostemma. He also transferred
Nerophila to Chaetolepis (DC.) Miq., an otherwise New World
genus of Melastomateae (Table 1; see also Grimm & Almeda,
2013).
While most genera of African Osbeckieae (Melastomateae)
have always been placed in this tribe, the positions of Dinophora
and Guyonia remain controversial. Dinophora has been treated
in Rhexieae (Bentham, 1849), Bertolonieae (Van Tieghem,
1891), Melastomateae (Triana, 1872; Baillon, 1877; Cogniaux,
1891; Gilg, 1898; Engler, 1921; Renner, 1993; Stevens, 2001–)
and Dissochaeteae (Jacques-Félix, 1983, 1995; Leistner, 2005).
Several authors have treated Guyonia within the Old World
Melastomateae (Triana, 1872; Van Tieghem, 1891; Krasser,
1893; Engler, 1921; Renner, 1993). However, Jacques-Félix
(1995) treated it in Tibouchineae (New World Melastomateae)
because of its tibouchinoid seeds which are elongate-cochleate,
somewhat smooth or asperulate having very shallowly raised
surface cells (see description of Chaeotolepis seeds; Whiffin
& Tomb, 1972) and also because of its chromosome number
n = 9 (Favarger, 1962).
Molecular phylogenetic analyses have been widely used
in resolving phylogenetic relationships at tribal (Meyer, 2001;
Fritsch & al., 2004; Michelangeli & al., 2004, 2013; Penneys &
Judd, 2013) and generic levels in Melastomataceae (Ionta & al.,
2007; Reginato & al., 2010; Stone & Andreasen, 2010; Stone,
2014; Kriebel, 2016; Reginato & Michelangeli, 2016). The study
by Clausing & Renner (2001) suggested that Melastomateae
are monophyletic, but Michelangeli & al. (2013), studying the
phylogenetic relationships of New World Melastomateae based
on a wide sampling, showed that Melastomateae sensu Renner
(1993) are diphyletic. Two non-sister clades, the “Marcetia
and allies” and the “core Melastomateae” were recovered
and supported by molecular and morphological characters
(Michelangeli & al., 2013). Also, the Old World Melastomateae
formed a monophyletic lineage, but the Madagascan endemic
and Asian Melastomateae genera were found to be nested
among the African Melastomateae (Clausing & Renner, 2001;
Michelangeli & al., 2013). Nevertheless, the African members
of Melastomateae are notoriously understudied and only barely
represented with only 7 (< 4%) of the ca. 185 species so far
Version of Record
585
Fernandes & Fernandes (1954a, 1969)
Jacques-Félix (1981, 1995)
This study
Dissotis Benth.
Dissotis Benth.
sect. Eudissotis Triana
sect. Heterotis Benth.
sect. Argyrella Naudin
Dissotis Benth.
subg. Dissotis Benth.
sect. Dissotis Benth.
sect. Macrocarpae A.Fern. & R.Fern.
sect. Squamulosae A.Fern. & R.Fern.
sect. Sessilifoliae A.Fern. & R.Fern.
subg. Dissotidendron A.Fern. & R.Fern.
subg. Dupineta (Raf.) A.Fern. & R.Fern.
subg. Osbeckiella A.Fern. & R.Fern.
sect. Osbeckiella A.Fern. & R.Fern.
subg. Heterotis Benth.
sect. Heterotis Benth.
sect. Argyrella (Naudin) A.Fern. & R.Fern.
Dissotis Benth. (± 120)
subg. Dissotis Benth.
sect. Dissotis Benth.
sect. Macrocarpae A.Fern. & R.Fern.
sect. Squamulosae A.Fern. & R.Fern.
sect. Sessilifoliae A.Fern. & R.Fern.
subg. Paleodissotis Jacq.-Fél.
subg. Dissotidendron A.Fern. & R.Fern.
subg. Dupineta (Raf.) A.Fern. & R.Fern.
subg. Osbeckiella A.Fern. & R.Fern.
sect. Osbeckiella A.Fern. & R.Fern.
sect. Senegambia Jacq.-Fél.
*Dissotis Benth. (± 50; D. grandiflora Benth.)
subg. Dissotis Benth.
sect. Dissotis Benth.
(D. grandiflora Benth.)
sect. Macrocarpae A.Fern. & R.Fern.
(D. speciosa Taub.)
sect. Squamulosae A.Fern. & R.Fern.
(D. longicaudata Cogn.)
sect. Sessilifoliae A.Fern. & R.Fern.
(D. welwitschii Cogn.)
subg. Paleodissotis Jacq.-Fél.
(D. leonensis Hutch. & Dalziel)
Treated in Dissotis sect.
Argyrella Naudin
Treated in Dissotis sect. Argyrella
(Naudin) A.Fern. & R.Fern.
Treated in Heterotis sect. Argyrella
(Naudin) Jacq.-Fél.
Argyrella Naudin (6; A. canescens (E.Mey.
ex Graham) Harv.)
Treated in Dissotis subg. Dupineta
(Raf.) A.Fern. & R.Fern.
Treated in Dissotis subg. Dupineta (Raf.)
A.Fern. & R.Fern.
Dupineta (Sm.) Raf. (6; D. multiflora (Sm.)
Raf.)
Treated in Dissotis subg. Dissotidendron
A.Fern. & R.Fern.
Treated in Dissotis subg. Dissotidendron
A.Fern. & R.Fern.
Dissotidendron (A.Fern. & R.Fern.) VeransoLibalah & G.Kadereit stat. nov. (11; D. melleri (Hook.f.) Veranso-Libalah & G.Kadereit
Treated in Tristemma Juss. Melastomastrum Naudin
Melastomastrum Naudin (6)
Melastomastrum Naudin (6; including
O. porteresii Jacq.-Fél. (M. capitatum (Vahl)
A.Fern. & R.Fern.)
Tristemma Juss.
Tristemma Juss.
(including Melastomastrum,
Heterotis sect. Leiocalyx
and sect. Wedeliopsis)
Tristemma Juss. (15)
Tristemma Juss. (16; T. mauritianum
J.F.Gmel.)
Dichaetanthera Endl.
Dichaetanthera Endl.
Dichaetanthera Endl. (8)
Dichaetanthera Endl. (8; D. articulata Endl.)
Heterotis Benth.
sect. Heterotis Benth.
sect. Cyclostemma Benth.
sect. Leiocalyx Benth.
sect. Wedeliopsis Benth.
Treated in Dissotis sect.
Heterotis Benth.
Treated in Dissotis subg. Heterotis
(Benth.) A.Fern. & R.Fern.
Dissotis sect. Heterotis (Benth.) A.Fern.
& R.Fern.
Heterotis Benth. (14)
sect. Heterotis (Benth.) Jacq.-Fél.
sect. Argyrella (Naudin) Jacq.-Fél.
sect. Cyclostemma Benth.
Heterotis Benth. (6; H. rotundifolia (Sm.)
Jacq.-Fél.)
Treated in Heterotis sect.
Cyclostemma Benth.
Treated in Dissotis sect.
Heterotis
Treated in Dissotis subg. Heterotis
Treated in Heterotis sect. Cyclostemma
Treated in Guyonia
Pseudosbeckia A.Fern. & R.Fern.
Pseudosbeckia A.Fern. & R.Fern. (1)
Pseudosbeckia A.Fern. & R.Fern. (1; P. swynnertonii (Baker f.) A.Fern. & R.Fern.)
Treated in Dissotis sect. Sessilifoliae
Treated in Heterotis sect. Cyclostemma
Anaheterotis Veranso-Libalah & G.Kadereit
gen. nov.
Tristemma Juss.
TAXON 66 (3) • June 2017: 584–614
Triana, 1872
Version of Record
Bentham, 1849
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
586
Table 1. Summary of the major inter- and infrageneric classifications proposed for the African Melastomateae and the revised classification based on the molecular phylogenetic analyses conducted
in this study.
Dinophora Benth. (1; treated in
Dissochaeteae)
Dinophora Benth. (treated
in Osbeckieae)
Types are only provided for the classification proposed in this study. The number of species per genus as suggested by Jacques-Félix (1995) and in this study are indicated in brackets and bold.
* Indicates all the genera included in the informal Dissotis and allies clade.
Dinophora Benth. (treated in
Osbeckieae)
Dinophora Benth.
(treated in Rhexiaeae)
Unplaced (1; not a Melastomateae)
Derosiphia Raf. (included in Dissotis subg. *Dissotis tubulosa (Sm.) Triana (1)
Dupineta (Raf.) A.Fern. & R.Fern
as Dissotis tubulosa (Sm.) Triana)
Derosiphia Raf.
Treated in Osbeckia sect.
Podocaelia Benth.
Chaetolepis Miq. (1; treated in Dissochaeteae)
Chaetolepis Miq. (1; treated in
Dissochaeteae)
Nerophila Naudin (treated in
Osbeckieae)
Nerophila Naudin (treated
in Osbeckieae)
*African Osbeckia (3)
Osbeckia L. (4)
Cailliella Jacq.-Fél. (1)
Cailliella Jacq.-Fél.
Osbeckia L.
Guyonia Naudin (1; treated in
Tibouchineae)
Guyonia Naudin
Guyonia Naudin
Osbeckia L.
Osbeckia L.
Antherotoma Hook.f. (11;
included subg. Osbeckiella)
Antherotoma Hook.f.
Antherotoma Hook.f.
Dionychastrum A.Fern. & R.Fern. (1)
Dionychastrum A.Fern. & R.Fern.
Cailliella Jacq.-Fél. (1; C. praerupticola
Jacq.-Fél.)
Guyonia Naudin (14; G. tenella Naudin,
including most species treated in H. sect.
Cyclostemma Benth.)
included in any phylogenetic analyses (Clausing & Renner,
2001; Michelangeli & al., 2013).
It is unknown if African Melastomateae genera, subgenera
and sections as currently recognized by Jacques-Félix (1981,
1995) represent natural lineages. Therefore, a well-sampled molecular phylogeny is needed to set the basis for an updated generic and infrageneric classification of African Melastomateae.
The present phylogenetic study is based on sequence variation
in three DNA regions (ITS, accD-psaI, psbK-psbL), and includes representatives of almost all currently recognised genera, subgenera and sections of African Melastomateae The
specific goals of this study are (1) to resolve phylogenetic relationships among genera, subgenera and sections within the
African Melastomateae, and (2) to propose a revised generic
classification.
*Antherotoma Hook.f. (11; A. naudinii
Hook.f.)
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Dionychastrum A.Fern. & R.Fern.
(1; D. schliebenii A.Fern. & R.Fern.)
TAXON 66 (3) • June 2017: 584–614
MATERIALS AND METHODS
Sampling and outgroups. — We included a total of 183
accessions of which 116 represent 75 species and 10 of the
13 traditionally recognised African genera (Fernandes &
Fernandes, 1969; Wickens, 1975; Renner, 1993; Jacques-Félix,
1995). We were unable to obtain samples of three monospecific genera, namely Cailliella, Dionychastrum and Nerophila.
Within Dissotis s.l. (Fernandes & Fernandes, 1969), we sampled 58 species to test if the updated inter- and infrageneric
groups proposed by Jacques-Félix (1995) represent natural lineages (Table 1). Three of the four recognised Dissotis
subgenera (i.e., subg. Dissotis, subg. Dissotidendron, subg.
Dupineta) are represented in this study. We were unable to sample Dissotis leonensis Hutch. & Dalziel, an endemic of Sierra
Leone and Guinea-Conakry and the only species of Dissotis
subg. Paleodissotis. However, we sampled all four sections
(sect. Dissotis, sect. Macrocarpae, sect. Sessilifoliae, sect.
Squamulosae) of D. subg. Dissotis and all three sections (sect.
Agryrella, sect. Cyclostemma, sect. Heterotis) of Heterotis
(Feranandes & Fernandes, 1969; Jacques-Félix, 1981) (see
Table 1). Seventeen Madagascan and Asian Melastomateae
were also sampled as part of the Old World Melastomateae. Ten
of ca. 48 Madagascan species were sampled, with 5 species representing all three Madagsacan endemic genera (Amphorocalyx
Baker, Rousseauxia DC., Dionycha Naudin). Two (Melastoma,
Osbeckia) of the three Asian Melastomateae genera representing 7 of ca. 83 species were also sampled. The Madagascan
endemic Dissotis pauciflora (Baker) Jacq.-Fél. (formerly
Rhodosepala pauciflora Baker) was not sampled. As outgroups, we included 3 species of Miconieae, 3 of Merianieae,
2 of Rhexieae, 1 of Physeterostemon, 3 of Microlicieae, 37
(14/30 genera) of New World Melastomateae and 2 species
of Cambessedesia DC. Voucher information and GenBank
accession numbers are available in Appendix 1.
Morphological data were obtained from the literature
and supplemented by our own observations. Field observations were conducted in Cameroon (July–October 2014), and
study of herbarium material was made through visits to BR
and BRLU, loans from BR, BRLU, C, EA, K, MO, WAG and
Version of Record
587
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
Fig. 1. Flowers of representatives of African Melastomateae with focus on stamen morphology. A, Features of typical Melastomateae stamens
(an, anther; fi, filament; p, pedoconnective; vap, ventral appendage; arrow indicates position of dorsal appendages if present); B, Dissotis grandiflora (subg. Dissotis sect. Dissotis); C & D, Dissotis longisetosa (subg. Dissotis sect. Dissotis); E, Dissotis welwitschii (subg. Dissotis sect.
Sessilifoliae); F, Dissotis perkinsiae (subg. Dissotis sect. Macrocarpae); G, Dissotis princeps (subg. Dissotis sect. Squamulosae); H, Dissotis
thollonii var. elliotii (subg. Dissotis sect. Squamulosae); I, Dissotidendron melleri; J, Dissotidendron johnstonianum; K, Dupineta brazzae;
L, Dichaetanthera corymbosa; M, Dupineta multiflora. ― Photographs by M.C. Veranso-Libalah (A, C, D, F–H, L), P.B., Phillipson (B), D.
Goyder (E, K), R. v. Blittersdorff (I, J), E. Bidault (M).
588
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Fig. 2. Representatives of African Melastomateae. A, Argyrella canescens; B, Heterotis prostrata; C, Antherotoma irvingiana; D, Osbeckia
decandra; E, Antherotoma naudinii; F, Dissotis tubulosa; G & H, Melastomastrum porteresii; I, Melastomastrum segregatum; J, Melastomastrum
theifolium; K, Tristemma mauritianum; L, Dinophora spenneroides; M, Pseudosbeckia swynertonii. ― Photographs by B. Wursten (A, E, I, M),
V. Droissart (B), M.C. Veranso-Libalah (C, F, J–L), E. Bidault (D, G, H).
Version of Record
589
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
UPS, and through online repositories from BM (http://data.
nhm.ac.uk/), LISC (http://actd.iict.pt/), P (https://science.mnhn.
fr/all/search), WAG (http://bioportal.naturalis.nl/), Z (http://
www.herbarien.uzh.ch/en/Datenbanken.html) and COI (raw
catalogue of specimens was provided but not yet publicly available online). Most of the type specimens and protologues of
African Melastomateae were seen except for those types that
were kept in Berlin and destroyed during World War II. A list
of the specimens examined for morphological characters is
available from the first author on request.
DNA extraction, amplification and sequencing. — Total
genomic DNAs were extracted from silica-gel dried or herbarium material using the DNeasy Plant Mini Kit (Qiagen,
Hilden, Germany), with modification to the manufacturer’s
protocol following Goldenberg & al. (2008). During the lysis
stage, 30 µl Proteinase-K was added to all samples which were
incubated at 42°C for 20–22 hours.
In order to resolve the phylogenetic relationships amongst
the African Melastomateae, we sequenced one nuclear locus
(the nuclear ribosomal internal transcribed spacer [nrITS] region) and two plastid spacers, accD-psaI and psbK-psbL. The
ITS region has been used widely in resolving phylogenetic relationships within several tribes and genera in Melastomataceae
(Michelangeli & al., 2004, 2008, 2013; Ionta & al., 2007;
Goldenberg & al., 2008; Stone & Andreasen, 2010; Penneys
& Judd, 2013; Stone, 2014; Majure & al., 2015). We designed
some new specific ITS primers for amplification and sequencing (Table 2). The accD-psaI and psbK-psbL intergenic spacers have proven to be variable at tribal and generic levels in
Melastomataceae. These regions have already been employed
to resolve the phylogeny of the Blakeeae (Penneys & Judd,
2013), New World Melastomateae (Michelangeli & al., 2013)
and several other genera of Melastomataceae (Reginato & al.,
2010; Kriebel & al., 2015; Reginato & Michelangeli, 2016). An
added advantage of using these markers was to obtain numerous previously published (GenBank) sequences of New World
Melastomateae and other tribes for use as outgroups in this
study. The psbK-psbL plastid region was amplified using the
primers published by Reginato & al. (2010), and the accD-psaI
region was amplified using the primers designed by Small &
al. (1998). All primers used in this study are shown in Table 2.
Polymerase chain reactions (PCR) were performed with a
total volume of 25 μl containing the following: 16.0 μl water,
1 μl template DNA, 2.5 μl 10× PCR buffer (Sigma-Aldrich,
Hamburg, Germany), 2.5 μl MgCl2 (25 mM), 1.0 μl each of
10 μM primers, 0.5 μl dNTPs (10 mM), 0.25 μl bovine serum
albumin (10 μg/μl) and 0.25 μl of Taq polymerase. The PCR
protocols for amplifying accD-psaI and psbK-psbL followed
those in Penneys & Judd (2013) with minor modification as
follows: (accD-psaI) 94°C for 3 min, 33 cycles of 94°C for 45 s,
64°C for 45 s and 72°C for 105 s, followed by 72°C for 6 min;
(psbK-psbL) 94°C for 2 min, 40 cycles of 94°C for 30 s, 55°C
for 30 s and 72°C for 30 s, followed by 72°C for 7 min. The
ITS region was amplified in two parts (ITS 1, ITS 2), since
amplifying the whole ITS fragment was problematic. Both ITS
parts were amplified as follows: 94°C for 1 min, 35 cycles of
94°C for 20 s, 63°C for 30 s and 72°C for 1 min, followed by
94°C for 20 s, 55°C for 80 s and 72°C for 8 min.
The amplified products were cleaned using “ExoSAP-IT’’
(Affymetrix, Santa Clara, California, U.S.A.). Cycle sequencing was carried using Big Dye Terminator v.3.1 reagents
(Applied Biosystems, Darmstadt, Germany) with the same
forward and reverse primers as used in the amplification of
accD-psaI, psbK-psbL and ITS 1. However, for ITS2 a different reverse primer was used for sequencing (see Table 2).
Sequencing was done at StarSEQ (Mainz, Germany) or using a
3130xl Genetic Analyzer (Applied Biosystems) at the Institute
of Ecology, Johannes Gutenberg-University, Mainz, Germany.
Phylogenetic inference. — Contigs of forward and reverse sequences were assembled and manually edited using
Sequencher v.4.1.4 (Gene Codes Corp., Ann Arbor, Michigan,
U.S.A.). Sequence alignment was done manually using
Mesquite v.3.04 (Maddison & Maddison, 2015). We conducted
phylogenetic analyses using maximum likelihood (ML), maximum parsimony (MP) and Bayesian inference (BI) methods.
First, we separately analysed each of the three datasets (ITS,
accD-psaI, psbK-psbL). Then we combined the two plastid
datasets (accD-psaI, psbK-psbL) into a single one. Since trees
based on the three single and the combined plastid markers
revealed no well-supported conflict (ML or MP bootstrap values ≥ 70% or BI posterior probability ≥ 0.95), we concatenated
all three datasets for all subsequent analyses. The best-fitting
Table 2. Primers of the three markers used in this study.
590
Marker
Primer name
Primer sequence 5′→3′
Source
ITS
ITS-241r
CAG TGC CTC GTG GTG CGA CA
Michelangeli & al. (2004)
ITS
ITS2-MEL2 R
CCA CCT TCC AGT GAA CTT GG
This study (sequencing primer ITS 2 reverse)
ITS
ITS2-MEL F
CGG CTC TTG CAT CGA TGA AG
This study
ITS
ITS1-MEL F
GGA GAA GTC GTA ACA AGG TTT C
This study
ITS
ITS1-MEL R
CTT GCG TTC AAA GAA TTG ATG G
This study
accD
accD F
GGG ATA TCA TTA TTG CCG AAC C
Small & al. (1998)
psaI
psaI R
AGA AGC CAT TGC AAT TGC CGG AAA
Small & al. (1998)
psbK
psbK F
TTA GCC TTT GTT TGG CAA G
Reginato & al. (2010)
psbL
psbL R
AGA GTT TGA GAG TAA GCA T
Reginato & al. (2010)
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
model for nucleotide substitutions was determined using the
Akaike information criterion (AIC) in jModelTest2 v.2.1.6
(Darriba & al., 2012). The best substitution model for the individual and combined data was GTR + I + G, so the combined
dataset was analysed using this single model.
The ML trees were generated using RAxML v.8.2.8
and inferred under default settings (Stamatakis, 2014). The
MP analyses were carried out in PAUP* v.4.0b10 (Swofford,
2002). A heuristic search was conducted of 1000 replicates
with random sequence addition, TBR branch swapping and
MulTrees on. Swapping was performed on a maximum of 50
trees (nchuck = 50). All characters were unordered and equally
weighted, and gaps were treated as missing data. Bootstrap values (Felsenstein, 1985) were calculated from 10,000 replicates
using a heuristic search using TBR branch swapping and one
random addition-sequence replicate. BI was performed using
BEAST v.1.8.2 (Drummond & Rambaut, 2007). All BI analyses
were conducted under the GTR + I + G substitution model with
gamma distribution in four categories and Birth-Death process
chosen as the tree prior. We ran two independent Markov chain
Monte Carlo (MCMC) runs of 20 million generations each,
sampling every 2000 generations, starting with a random tree.
The individual output log files were examined using Tracer
v.1.5 (Rambaut & Drummond, 2009) to assess convergence.
Both runs had an ESS value greater than 200. The resultant tree
files were combined using LogCombiner v.1.8.0 (Drummond &
Rambaut, 2007). The first 2000 trees (20%) were discarded as
burn-in, and a maximum clade credibility (MCC) tree was constructed using TreeAnnotator v.1.8.0 (Drummond & al., 2012).
Major phylogenetic relationships among clades. — Tree
topologies from the ML, MP and BI analyses were very similar
with a well-resolved backbone (Fig. 3) allowing for comparison between trees. We considered the clades as resolved when
the bootstrap values from the ML (BS) and MP (PBS) analyses were ≥ 70% and the posterior probability (PP) was ≥ 0.95
for the BI analysis. Old World Melastomateae formed a wellsupported, monophyletic group (BS 100/PBS 100/PP 1) in the
ML, MP and BI analyses (Fig. 3), except for the monospecific
genus Dinophora which was recovered in a polytomy including Rhexieae, one New World Melastomateae clade (Aciotis
D.Don and allies), Microlicieae and a large clade containing
both the New World (Tibouchina Aubl. and allies; Pterolepis
Schrad. + Pterogastra Naudin) and Old World Melastomateae
(Figs. 3, 4A).
The Old World Melastomateae are part of a trichotomy
with two New World Melastomateae clades, i.e., Pterolepis
plus Pterogastra on the one hand and Tibouchina and allies on
the other. At their root, the Old World Melastomateae are split
into two primary clades, Pseudoheterotis and Euheterotis (these
clade names are informal and proposed for the first time in this
study, Fig. 4B; Electr. Suppl.: Figs. S1, S2). The Pseudoheterotis
clade is composed of Guyonia s.l. (including Heterotis sect.
Cyclostemma), Argyrella Naudin (Heterotis sect. Argyrella),
Anaheterotis Veranso-Libalah & G.Kadereit (Heterotis pobeguinii (Hutch. & Dalziel) Jacq.-Fél.; see Taxonomic Treatment),
Melastomastrum s.l. (including Osbeckia porteresii Jacq.Fél.) and Tristemma s.l. (including M. cornifolium (Benth.)
Jacq.-Fél.). The Euheterotis clade comprises Dichaetanthera,
Dissotidendron Veranso-Libalah & G.Kadereit (Dissotis
subg. Dissotidendron; see Taxonomic Treatment), Heterotis
(H. sect. Heterotis), Asian Melastomateae (Osbeckia s.str. and
Melastoma), Madagascan endemic genera (Amphorocalyx,
Rousseauxia, Dionycha), Dupineta Raf. (Dissotis subg.
Dupineta; see Taxonomic Treatment), Pseudosbeckia, Dissotis
subg. Dissotis (all four sections), Antherotoma, and the remaining African Osbeckia species recognised by Jacques-Félix
(1995). According to the most recent generic delimitation
within the African Melastomateae (Jacques-Félix, 1981, 1995),
in addition to the monospecific Pseudosbeckia lineage, only
two well-supported monophyletic groups are recovered in our
analyses, i.e., Dichaetanthera and Tristemma. The remaining
genera, i.e., Dissotis, Heterotis, Antherotoma, Osbeckia and
Melastomastrum are all polyphyletic (Fig. 4B; Electr. Suppl.:
Figs. S1, S2).
RESULTS
Molecular variation. — We generated a total of 286 new
DNA sequences for African Melastomateae, i.e., 103 sequences for ITS, 115 for psbK-psbL and 78 for accD-psaI (see
Appendix 1). The combined dataset contained 3061 characters
from all three markers (ITS, accD-psaI, psbK-psbL). Of the
total of 183 accessions, 138 were represented by sequence data
from all three loci and 45 accessions were represented by two
loci. The plastid and nuclear markers are relatively variable
(Table 3) and therefore most of the major clades were recognised in the plastid and nuclear gene trees, respectively, even
though the backbone was generally not well supported (Electr.
Suppl.: Figs. S1, S2).
Table 3. Summary statistics of genomic regions sequenced for phylogenetic analyses of African Melastomateae.
% Missing
data
Aligned
base pairs
Marker
Accessions
sampled
nrITS
173
4
1030
Variable
sites
482
% Variable
sites
47
Conserved
sites
509
Parsimonyinformative
sites (PIS)
% PIS
Model
399
39
GTR + I + G
accD-psaI
151
17
1458
448
31
810
288
20
GTR + I + G
psbK-psbL
180
2
573
237
41
265
180
31
GTR + I + G
Combined
183
8
3061
1167
38
1584
879
29
GTR + I + G
Best-fit substitution model estimated under the AIC criterion using jModelTest.
Version of Record
591
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
Within the Pseudoheterotis clade, two of the three Heterotis
sections are recovered as monophyletic (Fig. 4B), i.e., H. sect.
Argyrella and H. sect. Cyclostemma (now included in Guyonia).
In turn, these two sections plus H. pobeguinii form a well-supported monophyletic group (BS 100/PBS 98/PP 1) that is sister
to a well-supported clade (BS 100/PBS 100/PP 1) containing
Melastomastrum s.l. and Tristemma s.l. (Fig. 4B; Electr. Suppl.:
Figs. S1, S2).
Within the Euheterotis clade, there is successive branching
of clades with Dichaetanthera branching off first, followed
by the well-supported monophyletic Dissotidendron (BS 100/
PBS 100/PP 1). Next is Heterotis sect. Heterotis (Heterotis),
a strongly supported clade (BS 100/PBS 100/PP 1), followed
by the Asian Melastomateae genera and another clade consisting of the Madagascan endemic genera (Fig. 3, 4B; Electr.
Suppl.: Figs. S1, S2). Next to the Madagascan endemic genera
is Dupineta, a well resolved clade (BS 100/PBS 100/PP 1),
although its sister-group relationship to Pseudosbeckia plus
the Dissotis and allies clade is not strongly supported (BS
89/PBS 77/PP 0.56). Next to Dupineta is the monospecific
Pseudosbeckia lineage. Lastly, we recovered a large clade
of Dissotis and allies with partially supported internal subclades, composed of Dissotis s.str. (excluding subg. Dupineta
and Dissotidendron), Antherotoma, Dissotis tubulosa (Sm.)
Triana and African Osbeckia (BS 94/PBS 91/PP 0.95; Figs.
3, 4B; Electr. Suppl.: Figs. S1, S2). Within Dissotis and allies,
D. subg. Dissotis and Antherotoma are polyphyletic (Fig. 4;
Electr. Suppl.: Figs. S1, S2).
Placement of Dinophora. — Dinophora was first placed in
Rhexieae, and Spennera Mart. ex DC. (now Aciotis) was suggested as its closest relative (Bentham, 1849). It was later treated
in Bertolonieae because its seed anatomy and geographic distribution are similar to those of African Ochthocharis G.Don
(Van Tieghem, 1891). Dinophora has also been treated in
Osbeckieae (Triana, 1872; Baillon, 1877; Cogniaux, 1891;
Gilg, 1898; Engler, 1921; Renner, 1993; Stevens, 2001–) although its fruits are fleshy compared to the rest of the Old
World Melastomateae (except Tristemma and Melastoma which
have berry-like fruits). Lastly, because of its baccate fruit and
chromosome number n = 12 (Favarger, 1962) it has been treated
in Dissochaeteae (Jacques-Félix, 1983, 1995; Leistner, 2005).
In our molecular phylogenetic analyses, Dinophora is part
of a polytomy with Rhexieae, one New World Melastomateae
clade (Aciotis and allies), Microlicieae and the large clade consisting of the further two New World Melastomateae clades
plus the Old World clade (Figs. 3, 4A; Electr. Suppl.: Figs. S1,
S2). Our analyses therefore suggest that Dinophora does not
belong in the “core Melastomateae” sensu Michelangeli & al.
(2013). In subsequent analyses it will be necessary to increase
the sampling of Dissochaeteae before any decision concerning
the tribal placement of Dinophora can be made.
Revised generic classification of African Melastomateae.
— Our molecular phylogenetic study provides the first in-depth
sampling of African Melastomateae including most genera
(10/13) and approximately 42% of the species. With this dense
sampling, we succeeded to recover as monophyletic several
genera, subgenera and sections that have been suggested in
earlier systematic treatments based on morphology (Bentham,
1849; Fernandes & Fernandes, 1969; Jacques-Félix, 1981,
1995) (compare Table 1; Figs. 1–4; Electr. Suppl.: Figs. S1,
S2). To accommodate phylogenetically isolated members of
previously recognised genera and to avoid polyphyletic taxa,
we propose several taxonomic adjustments and new combinations in African Melastomateae (see Taxonomic Treatment).
All taxonomic groups proposed here ensure that all taxa represent monophyletic lineages (Fig. 4; Electr. Suppl.: Figs. S1,
S2; Table 1).
In previous taxonomic treatments, staminal morphology
has been very important for the delimitation of genera in Old
World Melastomateae. For instance, isomorphic stamens were
used to separate Osbeckia from Dissotis (Bentham & Hooker,
1867; Fernandes & Fernandes, 1954a). Isomorphic stamens
have the same colour and are equal to subequal in size while
dimorphic stamens differ in colour and are markedly unequal
in size (Figs. 1, 2). Mapping of stamen-type (isomorphic versus
dimorphic) on the species included in our phylogeny shows
that this character is very unstable and changes frequently in
many subclades of Melastomateae respectively (Fig. 4B). For
DISCUSSION
Overview of Melastomateae. — The traditional circumscription of the Old World Melastomateae (Osbeckieae) as first
suggested by Cogniaux (1891) is recovered and strongly supported as monophyletic in our ML, MP and BI analyses (BS
100/PBS 100/PP 1). The Old World Melastomateae are in a
trichotomy with two New World Melastomateae clades (a small
Pterolepis and Pterogastra clade and a large Tibouchina and
allies clade; Figs. 3, 4; Electr. Suppl.: Figs. S1, S2) and these
results are similar to those earlier published by Michelangeli
& al. (2013). The core Melastomateae sensu Michelangeli & al.
(2013) are characterised by cochleate seeds with tuberculate
testa (Whiffin & Tomb, 1972), stamens usually with a pedoconnective with one or more ventral appendages and the ovary
with a crown of persistent bristles (absent in Guyonia ciliata
Hook.f.). The ovary in New World Melastomateae is free (including Guyonia) whereas the Old World Melastomateae have
ovaries more or less fused to the hypanthium (Jacques-Félix,
1995).
Fig. 3. Major clades recovered in the phylogenetic analyses of African Melastomateae based on combined nuclear (nrITS) and plastid (accDpsaI, psbK-psbL) data. Tree topology resulting from the Bayesian inference analysis (left) with posterior probabilities above branches. Tree
resulting from the maximum likelihood (ML) analysis (right) with ML bootstrap support values above branches and bootstrap support values
resulting from the parsimony analysis below branches. In the OW Melastomateae clade, the major clades recognised in this study are given in
bold and old taxonomic names in brackets.
592
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
1
Merianieae
1
Miconieae
1
1
Cambessdesia (Unplaced)
Rhexieae (Rhexia)
Dinophora
1
1
1
1
NW Melastomateae
(Aciotis and allies)
98
100
99
100
100
100
100
100
100/100
100
100
100/100
100
100
Microlicieae
0.99
NW Melastomateae
(Tibouchina and allies)
0.98
1
1
1
1
1
NW Melastomateae
(Pterolepis+Pterogastra)
Guyonia (H. sect. Cyclostemma+Guyonia)
Anaheterotis (H. pobeguinii)
Argyrella (H. sect. Argyrella)
73
100
100 100
100
99/93
99
96
94
71
1
Pseudoheterotis
100
100
1
1
Euheterotis
OW Melastomateae
Melastomastrum
(+O. porteresii)
1
0.99
1
1
1
1
Tristemma
(+M. cornifolium)
100
100
94
86
1
0.95
100
100
100
100
Pseudoheterotis
100
96
100
100
Euheterotis
Dichaetanthera
100
100
1
Dissotidendron
(D. subg. Dissotidendron)
100
100
1
Heterotis
(H. sect. Heterotis)
100
100
97
86
1
Asian Melastomateae
92
90
100
100
1
Madagascan endemic
Melastomateae
99
97
1
Dupineta
(D. subg. Dupineta)
100
100
OW Melastomateae
85
62
95
94
98
98
Pseudosbeckia
89
77
1
94
85
0.95
Dissotis and allies
(Dissotis+Antherotoma+
African Osbeckia)
Version of Record
94
91
593
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
example, within Heterotis and Antherotoma, H. buettneriana
(Cogn. ex Büttner) Jacq.-Fél. and A. naudinii Hook.f. are the
only species with isomorphic stamens. In contrast to some
previous taxonomic treatments we consider isomorphism versus dimorphism of stamens as an unreliable character when
delimiting genera in Melastomateae.
Generic limits within the Pseudoheterotis clade. — The
Pseudoheterotis clade comprises the following subclades:
Guyonia, Argyrella, Melastomastrum, Tristemma and the
monospecific Anaheterotis lineage. This clade contains all the
species earlier treated in Heterotis sect. Cyclostemma, sect.
Leiocalyx, and sect. Wedeliopsis by Bentham (1849) and in
H. sect. Argyrella and sect. Cyclostemma by Jacques-Félix
(1981, 1995) (Table 1).
Guyonia. — Heterotis sect. Cyclostemma s.str. (JacquesFélix, 1981, 1995) comprises ca. 12 species in three morphologically and ecologically distinct groups. The first group consists
of six species: H. arenaria Jacq.-Fél., H. cinerascens (Hutch.)
Jacq.-Fél., Dissotis glandulosa A.Fern. & R.Fern., H. obamae
Lejoly & Lisowski, H. rupicola (Gilg ex Engl.) Jacq.-Fél., and
H. seretii (De Wild.) Jacq.-Fél. This group is characterised by
subsessile leaves with flattened petioles, no intersepalar appendages and is mainly found in montane grassland habitats.
The second group consists of five species including H. antennina (Sm.) Benth., H. entii (J.B.Hall) Jacq.-Fél., Dissotis humilis A.Chev. & Jacq.-Fél., H. pygmaea (A.Chev. & Jacq.-Fél.)
Jacq.-Fél., and H. sylvestris (Jacq.-Fél.) Jacq.-Fél., which are
mainly found in wet habitats. They are also characterized by
adventitious roots, intersepalar appendages, and distinctly petiolate leaves. The third group is monospecific and consists of
H. jacquesii (A.Chev.) Aké Assi (≡ Dissotis jacquesii A.Chev.),
which is morphologically different from all other species of
H. sect. Cyclostemma. Initially H. jacquesii had been included
in Dissotis, but already Feissly (1964) suggested to remove it
92
100/1
95/0.99
99
100/1
100/1
100
90/1
100
100/1
100
*/*
100/1
100
92
100/1
100
100
100/1
100
99
100/1
100/1
100
100/1
96/1
100
100/1
100
100/0.99
81
92/1
100
100
100/1
100/*
91
100
100/1
100
*
100/1
100
83/1
100/1
100
100/1
100
73
100/1
73/0.96
100
86/*
100/1
100
100/1
73
93/1
79
92/1
100
87/0.95
85
100
100/1
76/0.97
100
*
100/1
100
73/0.98
100/1
Meriania longifolia
Graffenrieda moritziana
Merianieae
Graffenrieda latifolia
Physeterostemon thomasii
Miconia dodecandra
Miconieae
Leandra kleinii
Miconia tomentosa
Cambessedesia hilariana Cambessdesia (Unpl.)
Cambessedesia espora
Dinophora spenneroides 21 Dinophora (Unplaced)
Rhexia aristosa
Rhexieae
Rhexia virginica
Trembleya parviflora
Lavoisiera mucorifera
Microlicieae
Lavoisiera imbricata
Aciotis paludosa
Aciotis rubricaulis
Aciotis indecora
Aciotis and allies
Acisanthera quadrata
Ernestia glandulosa
Comolia microphylla
Acisanthera hedyotoidea
Tibouchina gracilis
Tibouchina confertiflora
Tibouchina citrina
Brachyotum incrassatum
Brachyotum microdon
Chaetolepis microphylla
Chaetolepis cufodontisii
Monochaetum discolor
Monochaetum meridense
Monochaetum vulcanicum
Heterocentron elegans
Tibouchina
Pilocosta nubicola
Pilocosta nana
and allies
Tibouchina pulchra
Tibouchina dubia
Svitramia pulchra
Tibouchina cristata
Tibouchina candolleana
Tibouchina radula
Tibouchina clavata
Tibouchina cryptadena
Tibouchina arborea
Microlepis oleifolia
Tibouchina granulosa
Tibouchina arenaria
Tibouchina cardinalis
Pterogastra divaricata
Pterogastra minor
Pterolepis+
Pterolepis glomerata
Pterogastra
Pterolepis repanda
to Fig. 4B (ingroup)
Fig. 4A & B. Maximum likelihood (ML) phylogenetic cladogram of African Melastomateae with representatives of Asian and Madagascan
Melastomateae, and outgroups (Rhexieae, Microlicieae, Miconieae, Merianieae, Cambessedesia, NW Melastomateae) derived from the combined data matrix (nrITS, accD-psaI, psbK-psbL). Values above branches refer to bootstrap values resulting from the parsimony analysis (only
values ≥ 70), values below branches refer to bootstrap values resulting from the ML analysis (only values ≥ 70) and posterior probabilities resulting from Bayesian inference (only values ≥ 0.95). An asterisk (*) denotes a branch supported by either bootstrap or posterior probability only.
594
Version of Record
NW Melastomateae
98
100
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
89
96/1
98
*/0.98
100
82/1
100/1
100
100/1
*
*/0.98
93
99/1
to Fig. 4A (outgroup)
100/1
100
96
70/*
99/1
100
100
100/*
100/1
71
100
94/0.95
*
100/1
100
*/0.96
100
100
100/1
100/1
73/*
98
96
75/1
100/1
*
71/*
100
*
*/0.97
100/1
96
D
OW Melastomateae
*
99/*
99/1
99
81
100/1
100
89/*
100/1
*
100
87/*
91/1
100
62
100
*
81/*
100/1
85/0.99
100/1
100
81
91/*
100/1
100
86
100/1
97/1
100
98/1
86
72/1
100
100
91
100/1
93/1
100/1
100
*
98/*
100/1
94
98
95/1
96/1
*
99/*
*
*/1
*
*/1
99
98
100
100/1
98/1
100/1
94
*
*/1
75
85/1
*
*/0.98
89
98/1
93
94/1
91/1
77
89/*
73
100
85
81/*
100/1
94/1
*
*/0.97
91
94/0.95
92
Dissotis and allies
97
91/1
95/1
99
100
100/1
95/*
75
= Dimorphic stamens (markedly
unequal in both colour and size)
94/1
100
100/1
*
*/0.98
100
= Isomorphic stamens (equal or
subequal in size and same colour)
100/1
*
*/1
100
90
82/1
100/1
80
*/1
98
97/1
*
79/0.99
84
88/1
83
100/1
85
94/1
100
98
97/1
100/1
100
98/1
100
100/1
Version of Record
75
94/1
99
95/1
Heterotis antennina 148
Heterotis seretii 48
Heterotis arenaria 46
Heterotis jacquesii 144
Guyonia ciliata
Guyonia ciliata 131
Heterotis pobeguinii 132
Heterotis angolensis 39
Heterotis angolensis 8
Heterotis canescens 36
Heterotis amplexicaulis 54
Argyrella sp. 94
Heterotis canescens 1
Melastomastrum theifolium 10
Melastomastrum theifolium 83
Melastomastrum theifloium I95
Osbeckia porteresii 17
Osbeckia porteresii 129
Melastomastrum afzelii 155
Melastomastrum capitatum 162
Melastomastrum segregatum 135
Melastomastrum cornifolium 166
Melastomastrum cornifolium 152
Tristemma albiflora 158
Tristemma involucrata 154
Tristemma mauritianum
Tristemma hirtum
Tristemma demeusei 159
Tristemma akeasii 147
Tristemma coronatum
Tristemma littorale
Tristemma oreophilum 153
Tristemma vestitum 157
Tristemma demeusei 146
Dichaetanthera oblongifolia
Dichaetanthera arborea 176
Dichaetanthera arborea 172
Dichaetanthera articulata 173
Dichaetanthera erici rosenii 156
Dichaetanthera africana 170
Dichaetanthera africana
Dichaetanthera strigosa 174
Dissotis polyantha 89
Dissotis melleri 106
Dissotis cordata 163
Dissotis cordata 81
Dissotis caloneura 93
Dissotis caloneura 87
Dissotis caloneura 86
Heterotis rotundifolia
Heterotis congniauxiana 164
Heterotis buetteriana 121
Heterotis decumbens
Heterotis decumbens 37
Heterotis rotundifolia 34
Heterotis rotundifolia 35
Heterotis rotundifolia 5
Osbeckia stellata
Osbeckia nepalensis
Osbeckia australiana
Melastoma denticulatum
Melastoma candidum
Melastoma sanguineum
Melastoma malabathricum
Amphorocalyx multiflorus
Amphorocalyx rupestris
Rousseauxia andringitrensis
Dionycha bojerii
Rousseauxia minimifolia
Dissotis multiflora 30
Dissotis multiflora 108
Dissotis multiflora 32
Dissotis multiflora
Dissotis hensii 57
Dissotis hensii 62
Dissotis brazzae 19
Dissotis brazzae 29
Pseudosbeckia swynertonii 4
Dissotis thollonii 24
Dissotis princeps 20
Dissotis princeps 42
Dissotis romiana 45
Dissotis trothae 90
Dissotis trothae 117
Dissotis sizenandii 137
Dissotis trothae 84
Dissotis trothae 118
Dissotis pulchra 2
Dissotis thollonii 26
Dissotis thollonii 12
Dissotis thollonii var. thollonii 44
Dissotis crenulata 165
Dissotis princeps 6
Dissotis princeps 61
Dissotis trothae 102
Dissotis denticulata 122
Dissotis falcipila 127
Dissotis princeps 27
Antherotoma gracilis 43
Antherotoma angustifolia 140
Antherotoma naudinii 3
Antherotoma debilis 185
Antherotoma naudinii
Antherotoma phaeotricha 120
Antherotoma phaeotricha 119
Antherotoma naudinii 38
Antherotoma debilis 59
Dissotis longisetosa 23
Dissotis longisetosa 25
Dissotis grandifolia 105
Dissotis longisetosa 11
Antherotoma irvingiana 19
Antherotoma irvingiana 33
Dissotis gilgiana 52
Dissotis gilgiana 161
Dissotis welwitschii 184
Dissotis cryptantha 98
Dissotis chevalieri 28
Dissotis perkinsiae 160
Dissotis perkinsiae 7
Antherotoma densiflora 143
Dissotis tubulosa 31
Dissotis tubulosa 22
Osbeckia togoensis 130
Osbeckia decandra 125
Osbeckia decandra 124
Antherotoma senegambiensis 113
Antherotoma senegambiensis 60
Dissotis congolensis 58
Dissotis congolensis 41
Dissotis congolensis 18
Guyonia
(Heterotis sect.
Cyclostemma)
Anaheterotis (Heterotis pobeguinii)
Argyrella (Heterotis
sect. Argyrella)
Melastomastrum
Pseudoheterotis Clade
100
86
Tristemma
Dichaetanthera
Dissotidendron
(Dissotis subg.
Dissotidendron)
Heterotis
(Heterotis sect.
Heterotis)
Asian
Melastomateae
Madagascan
endemic
Euheterotis Clade
86
94/1
Dupineta
(Dissotis subg.
Dupineta)
Pseudosbeckia
Dissotis sect.
Squamulosae
Antherotoma
Dissotis sect.
Dissotis
Antherotoma
Dissotis sect.
Sessilifoliae
Dissotis sect.
Macrocarpae
Antherotoma
Dissotis tubulosa
African Osbeckia
(Unplaced)
Antherotoma
Dissotis sect.
Sessilifoliae
595
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
from Dissotis because it lacks intersepalar appendages and
the whole plant is entirely covered by glandular hairs. It has a
chromosome number of n = 16 (Favarger, 1952, 1962).
Our molecular analyses included 4 of these 12 species
(H. antennina, H. arenaria, H. jacquesii, H. seretii) representing all three groups. These species formed a monophyletic clade together with Guyonia (Figs. 3, 4B; Electr. Suppl.:
Figs. S1, S2), a genus with two species, i.e., the widespread
Guinean-Congolian G. ciliata and the poorly known type
G. tenella Naudin from Guinea. Guyonia has been suggested
to be closely related to the neotropical Aciotis (Jacques-Félix,
1995). However, according to our results Guyonia clearly belongs to Old World Melastomateae and is sister to H. jacquesii
(Fig. 4B; Electr. Suppl.: Figs. S1, S2; Table 1).
The phylogenetic position of H. sect. Cyclostemma plus
Guyonia is well apart from Heterotis sect. Heterotis which
clearly indicates that it should be recognised at the generic level.
As defined here, an expanded Guyonia comprises 14 species,
including the 12 previously treated in H. sect. Cyclostemma.
Morphologically, Guyonia s.l. are decumbent, rarely erect
herbs, geophytes (sometimes forming tuber-like underground
structures during unfavourable conditions) with one to few
flowers and persistent calyx-lobes.
Heterotis sect. Argyrella. — The genus Argyrella was
first described by Naudin (1850) and later treated as Dissotis
sect. Argyrella (Naudin) Triana (Triana, 1872). Fernandes &
Fernandes (1969, 1970) transferred D. canescens (E.Mey. ex
Graham) Hook.f. and another species, D. angolensis Cogn.
to Dissotis subg. Argyrella. These two species together with
four other Dissotis species previously treated in the “séries
des Dissotis canescentsˮ (Jacques-Félix, 1953) were then
transferred to Heterotis sect. Argyrella (Naudin) Jacq.-Fél.
(Jacques-Félix, 1981, 1995). These species are found mainly
in moist habitats such as marshes, swamps, wet depressions
and mountain valleys.
In our phylogenetic analyses, we sampled three of the
six species of this section. These three representative species
clearly formed a monophyletic group (BS 99/PBS 93/PP 1)
which is sister to Heterotis pobeguinii (Figs. 3, 4B; Electr.
Suppl.: Figs. S1, S2). Argyrella species are also morphologicaly
well defined and are characterised by being erect herbs with
an indumentum of stellate and/or glandular hairs, paniculate
inflorescences and tri-cuspidate anther appendages (Fig. 2).
According to Jacques-Félix (1995), H. sect. Argyrella should
be upgraded to generic rank if more substantial taxonomic
evidence is found. Based on their distinct morphology and the
molecular results presented here, we strongly support JacquesFélix’s (1995) proposition and here raise H. sect. Argyrella to
generic rank (Table 1; see Taxonomic Treatment).
Heterotis pobeguinii. — This species, endemic to Guinea
and Sierra Leone, is a glabrous herb, morphologically unique
and easy to recognise by its prominently ciliate leaf margins
and long, lax, terminal panicles. It was initially described
as Dissotis pobeguinii Hutch. & Dalziel, with D. cinerascens Hutch. suggested as its closest relative (Hutchinson &
Dalziel, 1928) although they are morphologically very distinct. Dissotis pobeguinii was later treated in Dissotis sect.
Sessilifoliae together with D. buraeavii (Cogn.) A.Fern. &
R.Fern and D. welwitschii Cogn. mainly because of being glabrous (Fernandes & Fernandes, 1954a). Most recently, it was
treated in H. sect. Cyclostemma by Jacques-Félix (1981, 1995).
We included D. welwitschii, type of D. sect. Sessilifoliae in
our analyses and found it nested within the large Dissotis and
allies clade. Also, Heterotis pobeguinii is strongly supported
as sister to Argyrella although they are morphologically very
different (Fig. 4B; Electr. Suppl.: Figs. S1, S2). Given its distinct
morphology together with our molecular results, we neither
support the placement of H. pobeguinii in D. sect. Sessilifoliae
nor in H. sect. Cyclostemma as suggested earlier. We instead
propose that H. pobeguinii be recognized as the separate,
monospecific genus Anaheterotis (see Taxonomic Treatment).
Melastomastrum. — Melastomastrum was first described
by Naudin (1850) and consisted of one species, M. erectum
(Guill. & Perr.) Naudin. This species had been earlier described as Tristemma erectum Guill. & Perr. (Guillemin &
al., 1833) and later treated as Dissotis erecta (Guill. & Perr.)
Dandy (Andrews, 1950). This species was then reinstated to
Melastomastrum (Fernandes & Fernandes, 1954b, 1956a;
Jacques-Félix, 1975) (compare Table 1) and is now treated
as a synonym of M. capitatum (Vahl) A.Fern. & R.Fern.
Melastomastrum sensu Jacques-Félix (1975) comprises six
species with eight varieties in three morphologically distinct
groups (Fig. 2). The first group consists of M. capitatum,
M. afzelii (Hook.f.) A.Fern. & R.Fern., M. segregatum (Benth.)
A.Fern. & R.Fern., and M. autranianum (Cogn.) A.Fern. &
R.Fern. (the latter not sampled in our study), which are mainly
herbs with large 5-merous flowers, distinctly dimorphic stamens, pubescent pedicels and calyx-tube, attenuate to rostrate
anthers, linear anther appendages and a valvate capsular fruit.
The second group is monospecific and consists of M. theifolium with its two varieties, i.e., var. theifolium and var. controversum (A.Chev. & Jacq.-Fél.) Jacq.-Fél. This species is a
hemicryptophyte with slender woody stems, oblong sepals,
completely glabrous calyx-tube, clavate anther appendages and
a dorsal spur on the internal anthers. The third group is also
monospecific and consists of M. cornifolium which is similar
to the species in the first group except for its poorly developed
ring of bristles about mid-way on its calyx-tube, similar to the
well-developed complete or incomplete ring of bristles found
on the calyx-tube of Tristemma species.
Our molecular analyses included five of the six species representing all three morphological groups. Groups one and two
formed a monophyletic clade together with Osbeckia porteresii
(BS 94/PBS 71/PP 0.95; see Fig. 4B; Electr. Suppl.: Figs. S1,
S2) while M. cornifolium was resolved as sister to Tristemma
(BS 100/PBS 96/PP 1).
Osbeckia porteresii is a small shrub endemic to Mt.
Nimba in Guinea. It has a glandular indumentum, persistent
calyx-lobes and bracts (leafy), and isomorphic stamens with
attenuate anthers lacking both a pedoconnective and an appendage. The original description of O. porteresii (JacquesFélix, 1939) was based on two collections, i.e., Portères 3175
and Jacques-Félix 1933. The latter collector was not able to observe flowers at the time of collection, but based on the former
596
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
collection, the flowers were described as yellow (Jacques-Félix,
1939). Observations of mounted herbarium specimens show
that the flowers appear yellowish, a colour not found in African
Melastomateae species except in Nerophila (now treated in the
otherwise neotropical genus Chaetolepis; see Jacques-Félix,
1995; Grimm & Almeda, 2013). Recent collections in MO
(Bidault 200 & 667, Lerat 36 & 11) and online photographs
(http://www.tropicos.org/ImageSearch.aspx) show that the
flowers of O. porteresii are pink and not yellow as stated in
the protologue (M.C. Veranso-Libalah, pers. obs.; Fig. 2). More
surprisingly, no previous studies of this species have mentioned
the presence of large leafy persistent bracts which enclose
the flowers (a character it shares with Melastomastrum and
Tristemma). It greatly differs from the other African Osbeckia
species, i.e., O. decandra (Sm.) DC., O. praviantha Jacq.-Fél.
and O. togoensis Leuenb. in being a small shrub with glandular hairs and persistent calyx-lobes and bracts while the other
species are herbs with caducous calyx-lobes and bracts. Our
molecular results clearly suggest O. porteresii should not be
placed within Osbeckia, but instead it is nested within the
Melastomastrum clade (Fig. 4B; Electr. Suppl.: Figs. S1, S2).
The phylogenetic position of O. porteresii was quite astonishing at first sight, but based on several synapomorphies
shared with Melastomastrum, we suggest broadening the
circumscription of Melastomastrum to include O. porteresii.
Melastomastrum as defined here will include five species (excluding M. cornifolium) and O. porteresii.
Tristemma. — This genus consists of 15 species and 7
varieties (Jacques-Félix, 1976, 1986, 1995), mainly distributed in Africa (from the Sudanian to Zambezian regions, but
prominently present in East and West Africa especially in the
Guinean-Congolian region) with only T. mauritianum J.F.Gmel.
also known from Madagascar and the Mascarene Islands.
Morphologically, Tristemma is close to Melastomastrum, and
all species within these two genera were treated in Tristemma
by Triana (1872). They share calyx-tubes without intersepalar
appendages, persistent calyx-lobes, and two large persistent
bracts concealing the flower. Based on staminal characters and
the nature of the calyx-tube pubescence, Jacques-Félix (1975)
separated these two genera again. Traditionally, Tristemma has
been characterised by brachymorphic flowers with isomorphic
stamens, calyx-tubes with one or more complete to incomplete
rings of bristles and baccate fruits bursting irregularly (Fig. 2).
Our molecular analyses included 10 of the 15 species of
Tristemma which formed a well-supported monophyletic group
(BS 100/PBS 96/PP 1) that also includes M. cornifolium (Figs.
3, 4B; Electr. Suppl.: Figs. S1, S2). The latter species is endemic
to humid areas and marshes in West Africa (Guinean region)
and was originally described as Heterotis cornifolia Benth.
(Bentham, 1849), later treated as Dissotis cornifolia (Benth.)
Hook.f. (Hooker, 1871; Keay, 1954), as Tristemma cornifolium
(Benth.) Triana (Triana, 1872) and lastly as M. cornifolium
(Jacques-Félix, 1975). Morphologically, M. cornifolium is very
similar to the other Melastomastrum species, except that it
has a poorly developed ring of bristles about mid-way on its
calyx-tube, very similar to that seen in Tristemma species. This
ring of bristles in M. cornifolium was described by Favarger
(1962) as an incomplete form of the bristles found in Tristemma,
while Feissly (1964) described them as a rudimentary crown
of bristles similar to those in Tristemma. These bristles also
have glandular hairs (Feissly, 1964), which are absent in the
rest of the Tristemma and Melastomastrum species, except in
Osbeckia porteresii which our phylogenetic results have shown
to be a member of Melastomastrum.
In view of our molecular results, we feel it is necessary to
broaden the circumscription of Tristemma to include M. cornifolium in order to avoid the need to create several new combinations and monospecific genera. Tristemma as defined here
will now comprise 16 species with 7 varieties.
Generic limits within the Euheterotis clade. — The
Euheterotis clade is very large, with several morphologically
well-defined and phylogenetically well-supported subclades
(except for the Dissotis and allies clade). Within Euheterotis
the following successively branching clades are recognised:
Dichaetanthera, Dissotis subg. Dissotidendron, Heterotis s.str.,
Asian Melastomateae, Madagascan endemic Melastomateae,
Dupineta, Pseudosbeckia and the large clade of Dissotis and
allies. We will discuss in detail these individual clades in the
order of their successive branching below. However, due to the
poor internal resolution of clades within Dissotis and allies, we
will only briefly discuss Dissotis, Antherotoma, and African
Osbeckia.
Dichaetanthera. — Dichaetanthera has ca. 30 species with
about 8 in Africa and the rest in Madagascar (Jacques-Félix,
1983; Renner, 1993). With only six of these species sampled
in this study, Dichaetanthera is recovered as a strongly supported monophyletic clade (BS 100/PBS 100/PP 1; Figs. 3,
4B; Electr. Suppl.: Figs. S1, S2). Dichaetanthera species are
mainly deciduous trees or large shrubs with leathery leaves,
4-merous flowers with very short persistent calyx-lobes, and
seeds cuneate to subcochleate with a lateral hilum and hyaline
papillae on the back (Wickens, 1975; Fernandes & Fernandes,
1978; Jacques-Félix, 1983) (Fig. 1). Within Dichaetanthera,
four sections have been proposed: sect. Pseudodionycha Cogn.,
sect. Hysteranthia Cogn., sect. Rutenbergianae H.Perrier and
sect. Dichaetanthera (Cogniaux, 1891; Perrier de la Bâthie,
1932; 1951). These sections are yet to be tested in a phylogenetic context.
Dissotis subg. Dissotidendron. — This subgenus was
originally thought to be endemic to East Africa (Fernandes &
Fernandes, 1969; Wickens, 1975; Jacques-Félix, 1981, 1995).
Morphologically, ten species were placed in it, characterised as
trees or shrubs with rugose leaf surfaces and 5-merous flowers
borne in panicles with the calyx-tube campanulate and calyxlobes persistent (Fernandes & Fernandes, 1969) (Fig. 1). Our
phylogenetic analyses included three of the ten species and
strongly support the monophyly of D. subg. Dissotidendron
(BS 100/PBS 100/PP 1). Dissotis cordata Gilg, a species endemic to the Katanga region of the Democratic Republic of
the Congo also falls in this group (Figs. 3, 4B; Electr. Suppl.:
Figs. S1, S2; Table 1).
Together with its unique morphology, our molecular
results clearly suggest that D. subg. Dissotidendron can no
longer be treated as a subgenus within Dissotis. We therefore
Version of Record
597
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
propose to elevate D. subg. Dissotidendron to generic rank (see
Taxonomic Treatment). Dissotidendron as defined in this study
will now include 11 species, 10 previously treated in D. subg.
Dissotidendron and with D. cordatum as a new addition.
Heterotis sect. Heterotis. — Heterotis was first described
by Bentham (1849) comprising four sections: sect. Heterotis,
sect. Cyclostemma, sect. Leiocalyx and sect. Wedeliopsis. All
species treated in the latter two sections were later included
in Melastomastrum (Naudin, 1850) while the species in sect.
Heterotis and sect. Cyclostemma were later treated in Dissotis
sect. Heterotis (Benth.) Hook.f. (Hooker, 1871). More recently,
Fernandes & Fernandes (1969) treated all species originally
placed in H. sect. Heterotis and sect. Cyclostemma in D. subg.
Heterotis and D. subg. Dissotis, respectively. After studying
staminal morphology, Jacques-Félix (1981) reinstated the species in D. subg. Heterotis as H. sect. Heterotis. This section currently includes seven species characterised by being decumbent
herbs rooting at nodes, calyx-tubes with stellate emergences
(except in H. decumbens (P.Beauv.) Jacq.-Fél. which has simple
hairs), intersepalar appendages with stellate hairs on the apex
and calyx-lobes persistent (Fig. 2). The group is widely distributed in Africa, from the Guinean-Congolian region through
the Sudanian to the Zambezian region. Dissotis fruticosa
(Brenan) Brenan & Keay, H. buettneriana, and Dissotis cogniauxiana A.Fern. & R.Fern. are restricted to the Congolian region whereas H. prostrata (Thonn.) Benth., H. decumbens and
H. rotundifolia (Sm.) Jacq.-Fél. are widely distributed in Africa.
Heterotis prostrata (often wrongly identified as H. rotundifolia)
is also quite weedy and has been documented as invasive in
parts of Asia, North America, Oceania, Central America and
the Caribbean (CABI, 2014; Judd & al., 2011).
Our molecular analyses included four of the seven species
of H. sect. Heterotis, and for these a well-supported clade is recovered (BS 100/PBS 100/PP 1; Figs. 3, 4B; Electr. Suppl.: Figs.
S1, S2). Our molecular results clearly exclude H. sect. Argyrella
and sect. Cyclostemma from Heterotis sensu Jacques-Félix
(1981) (see Table 1). We here suggest maintaining Heterotis
at generic rank as originally proposed by Bentham (1849), but
with a considerably narrower circumscription in comparison
to the more recent treatments of this genus by Jacques-Félix
(1981, 1995) (see Table 1 and Taxonomic Treatment).
Asian Melastomateae. — Melastoma (> 50 spp.) is the
most diverse genus of Asian Melastomateae, followed by
Osbeckia with ca. 30 species and Otanthera Blume with 3
species (Renner, 1993; Meyer, 2001). Our molecular analyses
included only seven of these Asian species (four of Melastoma
and three of Osbeckia), which are recovered in a well-supported
(BS 93/PBS 91/PP 1) monophyletic group (Figs. 3, 4B; Electr.
Suppl.: Figs. S1, S2). This clade is found nested within African
Melastomateae. Renner & Meyer (2001) suggested that the
Asian Melastoma and Osbeckia were likely derived from
African or Madagacan ancestors. However, little is currently
known about the phylogenetic relationships of these Asian
species.
Madagascan endemic Melastomateae. — Madagascar
currently has about 48 Melastomateae species in eight genera (Perrier de la Bâthie, 1951; Renner, 1993, Jacques-Félix,
1995). These are: Dichaetanthera (ca. 23 spp.), Rousseauxia
(ca. 13 spp.), Amphorocalyx (5 spp.), Dionycha (3 spp.),
Antherotoma (1 sp.), Melastoma (1 introduced sp.), Rhodosepala
(now Dissotis; 1 sp.) and Tristemma (1 sp.). Our analyses included six Dichaetanthera species (two from Madagascar and
four from Africa), and these together formed a monophyletic
group as already discussed above. We also sampled T. mauritianum and Antherotoma naudinii which are also distributed in
Africa, and these species were recovered in the Tristemma and
Dissotis and allies clades, respectively. Also, five species representing all three Madagascan endemic genera (Amphorocalyx,
Rousseauxia, Dionycha) were sampled, and these species
formed a monophyletic group (BS 96/PBS 98/PP 1; Figs. 3, 4B;
Electr. Suppl.: Figs. S1, S2). Nevertheless, we recommend more
in-depth studies of this Madagascan Melastomateae lineage, to
understand better their morphology, phylogenetic relationships
and biogeography.
Dissotis subg. Dupineta. — The genus Dupineta was first
described as monospecific with only Dupineta multiflora
(Sm.) Raf. (Rafinesque, 1838). It was later treated with an expanded circumscription as a subgenus of Dissotis (Fernandes
& Fernandes, 1969; Jacques-Félix, 1981, 1995). Dissotis subg.
Dupineta consists of six species widely distributed in Africa
(Jacques-Félix, 1981, 1995). This subgenus is morphologically
characterised by 5-merous flowers in panicles of cymes, intersepalar appendages absent, hypanthial emergences caducous, stamens dimorphic, appendages bifid, and calyx-tube
accrescent in fruit with distinct longitudinal ribs (Fig. 1). The
West African D. tubulosa was included in this subgenus by
Jacques-Félix (1995) because of its accrescent calyx-tube.
However, it is very different from the rest of the species by
being an annual herb with stamens isomorphic, hypanthial
emergences persistent and fruits lacking longitudinal ribs.
Our molecular analyses included four of the six species
treated in this subgenus by Jacques-Félix (1995), namely:
D. brazzae Cogn., D. multiflora (Sm.) Triana, D. hensii Cogn.
and D. tubulosa. Our phylogenetic results strongly support
(BS 100/PBS 100/PP 1) the monophyly of D. subg. Dupineta
s.str. (excluding D. tubulosa) although its relationship to
Pseudosbeckia and the Madagascan endemic clade is not well
resolved (see Figs. 3, 4B; Electr. Suppl.: Figs. S1, S2). Based
on the unique morphology of species in D. subg. Dupineta
together with molecular evidence, we strongly support the exclusion of D. tubulosa from D. subg. Dupineta and suggest
reinstating D. subg. Dupineta to generic rank (see Rafinesque,
1838) (Table 1 and Taxonomic Treatment). Dissotis tubulosa
is found nested within the large clade of Dissotis and allies in
our molecular results.
Pseudosbeckia. — Pseudosbeckia swynnertonii (Baker f.)
A.Fern. & R.Fern. is a monospecific lineage only known
from the Chimanimani Mountains of Zimbabwe and adjacent Mozambique. It was originally described as Osbeckia
swynnertonii Baker f. because of its isomorphic stamens and
5-merous flowers (Rendle & al., 1911). Fernandes & Fernandes
(1954a, c) found that this species has a distinct but short pedoconnective with two ventral appendages when compared
to Osbeckia species, and they placed it first in Dissotis as
598
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
D. swynnertonii (Baker f.) A.Fern. & R.Fern. before transferring it to Pseudosbeckia (Fernandes & Fernandes, 1956a, b).
It is a shrub with 5-merous flowers in cymes of 1–3 flowers,
sepals tardily caducous, and anthers erostrate and opening by
a terminal introrse pore (Fig. 2). In our molecular analyses this
monospecific genus is recovered as sister to the large, internally
unresolved clade consisting of Dissotis s.str. and allies (Figs.
3, 4B; Electr. Suppl.: Figs. S1, S2).
Dissotis and allies. — Our molecular phylogenetic results
recovered a large clade of Dissotis and allies with partial
internal resolution (BS 94/PBS 91/PP 0.95; see Fig. 3). This
large clade contains all species of D. subg. Dissotis and its four
sections (sect. Dissotis, sect. Macrocarpae, sect. Sessilifoliae,
sect. Squamulosae) as well as Antherotoma, D. tubulosa (previously placed in D. subg. Dupineta; Fernandes & Fernandes,
1969; Jacques-Félix, 1995) and two African Osbeckia species
(O. decandra, O. togoensis; see. Fig. 4B; Electr. Suppl.: Figs. S1,
S2). Three Dissotis sections (sect. Dissotis, sect. Macrocarpae,
sect. Squamulosae) each formed well-supported clades, as did
the species of Antherotoma treated in D. sect. Osbeckiella sensu
Jacques-Félix (1981). Nevertheless, the poor resolution of this
part of the phylogenetic tree does not allow us to propose a new
generic classification for this very large and morphologically
diverse clade. To resolve phylogenetic relationships between all
genera, sections and species, it will be necessary to sequence
additional loci and also to sample more extensively the species
belonging to this large clade. For the time being, we suggest
maintaining Dissotis s.str. (including subg. Dissotis and subg.
Paleodissotis) and Antherotoma. However, one firm conclusion
is that the genus Osbeckia s.str. (Asian Osbeckia) does not
occur in Africa.
Morphologically, this is a diverse clade of shrubs, subshrubs, perennial or annual herbs with paniculate, cymose,
capitulate inflorecences or solitary flowers, which are 4- or
5-(6-)merous, calyx-tubes glabrous or pubescent, calyx-lobes
persistent, tardily or early caducous, stamens isomorphic or
dimorphic, pedoconnective and appendages present or absent.
Dissotis. — Our phylogenetic analyses showed that
Dissotis s.l. (sensu Fernandes & Fernandes, 1969; Wickens,
1975; Jacques-Félix, 1981, 1995) is polyphyletic with two of
its subgenera (subg. Dissotidendron, subg. Dupineta) distantly
related to Dissotis s.str. as defined by its type D. grandiflora
Benth. We sampled ten species of D. sect. Squamulosae and
two species of D. sect. Dissotis (D. grandiflora, D. longisetosa
Gilg & Ledermann ex Engl.). These two sections were each
recovered as well-supported clades (BS 100/PBS 100/PP 1;
see Fig. 4B; Electr. Suppl.: Figs. S1, S2). We also sampled four
species previously treated in Dissotis sect. Macrocarpae, and
this group is recovered as monophyletic in our analyses (BS 82/
PBS 90/PP 1; see Fig. 4B; Electr. Suppl.: Figs. S1, S2). Our molecular analyses included three species of D. sect. Sessilifoliae,
i.e., D. congolensis (Cogn. ex Büttner) Jacq.-Fél., D. gilgiana
Hutch. & Dalziel and D. welwitschii (the type, Fig. 1). Dissotis
congolensis is part of a clade with Antherotoma senegambiensis
(Guill. & Perr.) Jacq.-Fél., O. togoensis and O. decandra, while
D. gilgiana is recovered as sister to D. welwitschii and D. sect.
Macrocarpae (Fig. 4B; Electr. Suppl.: Figs. S1, S2).
According to Wickens (1975), the four sections (sect.
Dissotis, sect. Squamulosae, sect. Macrocarpae, sect.
Sessilifoliae) of D. subg. Dissotis as proposed by Fernandes &
Fernandes (1969) were ill-defined, and he did not use them in
his treatment for the Flora of tropical East Africa. Although
we are able to recognise all these sections in our molecular
and morphological analyses, we do not provide any taxonomic
suggestions for these clades because of the only partially resolved backbone and internal branches within the large clade
of Dissotis s.str. and allies.
Antherotoma. — Antherotoma s.l. (sensu Jacques-Félix
(1995) comprises 11 species and is distributed in continental
Africa, apart from the type A. naudinii Hook.f. which is also
found in the Comoro Islands and Madagascar. Our molecular
analyses included five and three species of D. sect. Osbeckiella
and sect. Senegambia, respectively. We recovered a well-supported clade (BS 91/PBS 92/PP 1) containing all species earlier
treated in D. sect. Osbeckiella (Jacques-Félix, 1981). On the
other hand, the three species sampled of D. sect. Senegambia,
i.e., A. senegambiensis, A. irvingiana (Hook.f.) Jacq.-Fél. and
A. densiflora (Gilg) Jacq.-Fél., do not form a monophyletic
clade but fall in different positions in the tree (Fig. 4B; Electr.
Suppl.: Figs. S1, S2). Antherotoma senegambiensis, for example, groups with Dissotis congolensis which is also supported
by the morphological similarity between these two species.
However, from our current molecular sampling, we cannot
definitely conclude whether Antherotoma should retain all
species formerly treated in D. sect. Senegambia and D. sect.
Osbeckiella as done by Jacques-Félix (1995).
African Osbeckia. — In studying staminal morphology of
African Melastomateae, Jacques-Félix (1981) confirmed the
occurrence of four Osbeckia species (O. decandra, O. porteresii, O. praviantha, O. togoensis) in Africa. He further suggested that Antherotoma is sister to O. decandra which has
characteristic osbeckioid stamens, but he could not suggest any
relatives for O. porteresii or O. togoensis. The last species is a
small shrub with a dense indumentum of white hairs, stamens
subequal but not dimorphic and endemic to Togo. Osbeckia
praviantha on the other hand is morphologically very similar
to O. decandra: these are annual herbs having isomorphic stamens with neither a pedoconnective nor an appendage and are
endemic to West Africa (Fig. 2). As discussed earlier, our molecular results place O. porteresii in the Melastomastrum clade
and further support Osbeckia s.str. as a strictly SE Asian genus with no known representatives in Africa. Nonetheless, the
African species previously placed in Osbeckia are morphologically very different from Dissotis and Antherotoma and would
require a new generic or subgeneric classification after a more
comprehensive phylogenetic and morphological study. Two
poorly collected Osbeckias (O. albiflora Cogn. ex De Wild. &
Th.Dur., O. pusilla De Wild.) from the Democratic Republic
of the Congo are morphologically very close to D. congolensis
and are likely synonyms of that species. Also O. calotricha
Gilg which is currently known from a single type specimen
(P00412554!) and O. drepanosepala Gilg which has no type
but a well described species in Gilg (1898), are all probably
synonyms of Antherotoma senegambiensis.
Version of Record
599
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Dissotis tubulosa. — This species, distributed from
Guinea and Sierra Leone to Cameroon, is morphologically
unique but taxonomically problematic. It was first described
as Osbeckia tubulosa Sm. (Smith, 1813), and then transferred
to Dissotis by Triana (1872). The species has also been placed
in two monospecific genera, i.e., Derosiphia Raf. (Rafinesque,
1838) and Podocaelia A.Fern. & R.Fern. (Fernandes &
Fernandes, 1962). Dissotis tubulosa was again transferred to
Dissotis subg. Dupineta (Jacques-Félix, 1981, 1983) although
morphologically different from the rest of the species in this
group. It is an annual herb, calyx-tube with persistent stellate emergences, stamens isomorphic and fruits accrescent
but lacking longitudinal ribs (Fig. 2). From our phylogenetic
results, this species cannot be included in D. subg. Dupineta.
Nevertheless, whether to treat D. tubulosa as a monospecific
genus as suggested by previous authors cannot be decided in
this study.
Placement of unsampled genera. — Our current molecular sampling of African Melastomateae lacks two monospecific genera; Cailliella and Dionychastrum. However, based
on morphology we suggest that Cailliella may be a close relative of Melastomastrum. They share persistent calyx-lobes
and 5-merous flowers subtended by an involucre of bracts.
Cailliella is different from the other African Melastomateae
species by being a shrub, glabrescent with age, young branches
yellow with dense strigose hairs, nodes swollen, flowers subtended by 2–3 pairs of leaves, pedoconnective absent, connectives with sharp dorsal spur and two lateral linear lobes.
Morphologically, Dionychastrum is very similar to Dionycha, a Madagascan endemic genus. They are both deciduous shrubs with cymose inflorescence, isomorphic stamens,
calyx-tube and capsule not contracted above. Dionychastrum
differs from Dionycha by the following characters: 5-merous
flowers, calyx-lobes persistent, pedoconnectives curved,
anthers with one anterior pore, capsule apex glabrous and
exserted.
Also Chaetolepis, the only NW genus which is represented
in Africa by one species, C. gentanoides was unsampled. Based
on its morphology, this species would probably be placed within
the Dissotis and allies clade and not in NW Melastomateae.
Nevertheless, we treat this species in Chaetolepis until it is
phylogenetically sampled.
Recommendations. — Although several representatives
of African Melastomateae are included in our phylogenetic
analyses, a denser taxonomic sampling and further morphological studies of Guyonia (species formerly treated in H. sect.
Cyclostemma), Melastomastrum, Dichaetanthera (all four
sections) and Dissotis s.str. and allies (Antherotoma, Dissotis
and African Osbeckia) will be necessary to understand better their ecology, systematics and morphological divergence.
Also the monospecific genera Cailliella, Dionychastrum,
Nerophila (now Chaetolepis) and Rhodosepala (now Dissotis)
need to be sampled. Lastly, the Madagascan and Asian
Melastomateae genera are poorly represented. Only 7 of ca.
48 Madagascan Melastomateae species and two of the three
Asian Melastomateae genera (ca. 7/83 species) are included in
our phylogenetic analyses.
600
TAXONOMIC TREATMENT
Based on the phylogenetic and morphological results presented in this paper, we here propose the following changes/
revisions to the classification of African Melastomateae: (1)
the genera Argyrella Naudin and Dupineta Raf. are reinstated;
(2) a new status at generic rank is proposed for Dissotis subg.
Dissotidendron A.Fern. & R.Fern.; (3) the new monospecific
genus Anaheterotis Veranso-Libalah & G.Kadereit is described
for Heterotis pobeguinii (Hutch. & Dalziel) Jacq.-Fél.; and (4)
a broader circumscription of Guyonia is provided to include
all species formerly treated in H. sect. Cyclostemma Benth.
Jacques-Félix (1981, 1995) had previously reinstated Heterotis
Benth., and our results support the continued recognition of
this genus, but with a narrower circumscription that excludes
H. sect. Cyclostemma and sect. Argyrella (Naudin) Jacq.-Fél.
A total of 42 new combinations are effected at the species level
and below, some of which represent taxa that were not included
in our phylogenetic analyses. The new combinations for these
unsampled taxa are thus based on comparative morphology
and not on molecular results. The positions of the specific nucleotide substitutions used for molecular diagnosis are based
on the sequence alignments we have provided in the electronic
supplementary material.
We also provide a key to the 12 African Melastomateae
genera recognised in this study (Table 1). In addition, we treat
Dissotis Benth. and allies in one informal clade. This clade consists of Dissotis s.str. (including D. subg. Paleodissotis Jacq.Fél. and the four sections of D. subg. Dissotis) together with
Antherotoma Naudin (including Dissotis subg. Osbeckiella
A.Fern. & R.Fern. and D. sect. Senegambia Jacq.-Fél.),
Dissotis tubulosa (Sm.) Triana (previously treated in D. subg.
Dupineta (Raf.) A.Fern. & R.Fern.), and three African species
of Osbeckia L. (i.e., O. decandra (Sm.) DC., O. praviantha
Jacq.-Fél., O. togoensis Leuenb.). In a follow-up study, we will
aim to resolve further the phylogenetic relationships and classification within the Dissotis and allies clade.
Taxonomic note. — The genus Kadalia Raf. was published with two types, Osbeckia antennina Sm. and Osbeckia
rotundifolia Sm. (Rafinesque, 1838). In our phylogenetic
analyses both types were recovered in two different genera,
O. antennina Sm. in Guyonia Naudin and O. rotundifolia Sm.
in Heterotis Benth. Since the generic name Kadalia is older
and has priority over both Heterotis and Guyonia, it will be
required that either Kadalia antennina (Sm.) Raf. or Kadalia
rotundifolia (Sm.) Raf. be chosen as the type of Kadalia if
this name were to be used. Stone & Veranso-Libalah (2017),
discuss the nomenclatural issues concerning the name Kadalia
in detail and suggested it should be outrightly rejected, since
it has not been used and wrongly typified. Also, rejecting
Kadalia will maintain nomenclatural stability and avoid the
creation of several new combinations (i.e., six in the case
of Heterotis and fifteen in Guyonia). As a result, we do not
use the name Kadalia but instead treat Guyonia (including
G. antennina (Naudin) Veranso-Libalah & R.D.Stone) and
Heterotis (including H. rotundifolia (Sm.) Jacq.-Fél.) as two
segregate genera.
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Key to genera
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
10
10
11
11
Creeping plants rooting at the nodes ...................... 2
Erect plants not rooting at the nodes ...................... 3
Calyx-tube with stalked stellate emergences (except H. decumbens); intersepalar appendages acute to oblong, the
apex with stellate hairs; seeds cochleate, visibly arillate
........................................................ Heterotis
Calyx-tube glabrescent, glandular or with simple setose
emergences; intersepalar appendages subulate or dentritic,
the apex without stellate hairs; seeds elongate-cochleate
with sparse tubercules ............................. Guyonia
Calyx-lobes persistent ....................................... 4
Calyx-lobes caducous ...................................... 11
Calyx-tube with 1 or more rings of bristles; fruit a berry
or indehiscent capsule .......................... Tristemma
Calyx-tube glabrous or with bristles but these not in rings;
fruits capsular, dehiscent .................................... 5
Deciduous shrubs or trees ................................... 6
Perennial herbs or subshrubs ............................... 9
Shrubs glabrescent with age; leaf surface with dense strigose hairs; flowers solitary with an involucre of 2–3 pairs
of leaves ............................................. Cailliella
Shrubs or trees not glabrescent with age; leaf surface with
setae having bulbous bases, becoming rugose or coarse;
flowers in heads, cymes or panicles and lacking an involucre ............................................................ 7
Leaves 1.5–2.5 × 1–1.5 cm; flowers 5-merous, borne in
simple cymes of 1–3 flowers; intersepalar appendages
present; stamens isomorphic ............ Dionychastrum
Leaves 6.5–12 × 2.5–7 cm; flowers 4- or 5-merous, borne
in panicles; intersepalar appendages absent; stamens dimorphic ........................................................ 8
Flowers 4-merous; calyx-lobes short or absent, exposing the
corolla in bud; pedoconnectives with linear appendages;
seeds with hyaline papillae on the back .. Dichaetanthera
Flowers 5-merous; calyx-lobes large, concealing the corolla in bud; pedoconnectives with lobed, broad and obtuse
appendages; seeds lacking hyaline papillae on the back
................................................. Dissotidendron
Flowers in heads or solitary surrounded by large persistent
bracts ....................................... Melastomastrum
Flowers in racemes or panicles surrounded by small or
large leafy caducous bracts ............................... 10
Plant glabrous; leaves and petioles reddish to purplish,
margins densely serrate ending in very prominent ciliate
hairs; each flower surrounded by two large leafy bracts,
calyx-tube glabrous ........................... Anaheterotis
Plant shortly stellate-pubescent; leaves and petioles appearing dusty or tomentose, margins entire to serrulate;
each flower subtended by a pair of small bracts; calyx-tube
with only stellate or intermixed with simple, often glandtipped hairs ......................................... Argyrella
Flowers 5-merous; calyx-tube accrescent in fruit, developing a long neck with longitudinal ribs ...... Dupineta
Flowers 4- or 5-merous; calyx-tube not accrescent in fruit
and lacking a long neck with longitudinal ribs ........ 12
12 Leaves distinctly bicoloured, dark green above and
yellowish-green beneath; flowers 5-merous; anthers
erostrate, opening by an extrorse pore; seeds cochleate, covered with rather elongate tubercules arranged in parallel lines
................................................. Pseudosbeckia
12 Leaves mostly monocoloured, if bicoloured then not as
above; flowers 4- or 5-merous; anthers rostrate or oblong,
opening by an introrse pore; seeds cochleate but tubercules
not in parallel lines . .................... Dissotis and allies
Chromosome numbers. — Chromosome numbers of many
African Melastomateae species remain unknown, even though
it is very important in the delineation of certain genera. Based
on the studies of Favarger (1952, 1962) and Farron & Favarger
(1952), the following African Melastomateae genera can be
identified with their respective chromosome numbers in brackets; Heterotis (n = 15), Dupineta (n = 10), Melastomastrum
(n = 17), Tristemma (n = 17), Argyrella (n = 17) and Guyonia
(n = 9, 12 [only G. jacquesii]). So far, all the species in the
Dissotis allies clade whose chromosome numbers are known,
have n = 10. It will be necessary to study the chromosome
number of more African Melastomateae species especially
in the Dissotis and allies clade, since this together with other
taxonomic evidence, may help resolve the relationships in this
unresolved clade.
1. Anaheterotis Veranso-Libalah & G.Kadereit, gen. nov.
– Type: Anaheterotis pobeguinii (Hutch. & Dalziel)
Veranso-Libalah & G.Kadereit (≡ Dissotis pobeguinii
Hutch. & Dalziel).
Morphological diagnosis. – Anaheterotis differs from
Argyrella, its closest relative, by the following combination of
characters: erect, entirely glabrous herb except for its prominently ciliate leaf margins, leaf blades with 5 conspicous adaxial longitudinal nerves, few-flowered paniculate inflorescence,
and flowers enclosed by two leafy bracts. Argyrella is an erect
herb entirely covered with tomentose stellate and/or glandular
hairs and leaf margins more-or-less entire.
Molecular diagnosis. – Anaheterotis also differs from the
other African Melastomateae genera by one nrITS nucleotide
substitution: (indel)-to-G at position 913; and one accD-psaI
nucleotide substitution: C-to-A at position 1070 (summarised
in Table 4).
Anaheterotis pobeguinii (Hutch. & Dalziel) Veranso-Libalah
& G.Kadereit, comb. nov. ≡ Dissotis pobeguinii Hutch.
& Dalziel, Fl. W. Trop. Afr. 1: 211. 1927 ≡ Heterotis
pobeguinii (Hutch. & Dalziel) Jacq.-Fél. in Adansonia,
sér. 2, 20: 419. 1981 – Lectotype (designated here):
Guinea, Douné (Baffing), H. Pobéguin 1695 (K barcode
K000313170!; isolectotypes: P barcodes P00412581! &
P00412582!).
Note. – In the protologue of Dissotis pobeguinii,
Hutchinson & Dalziel (1927: 211) cited Pobéguin 1695 as the
type specimen. We located three specimens, two housed in P
and one in K, all part of the same field collection as the type
in the original description but none of them was designated as
Version of Record
601
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
Table 4. Diagnostic nucleotide substitutions in plastid (accD-psaI, psbK-psbL) and nuclear ribosomal (ITS) markers among African Melastomateae.
Position in
alignment
Specific
substitutions
Comment
Anaheterotis
913
G
– (rest of the African Melastomateae genera have an indel)
1070
A
C (rest of the African Melastomateae genera)
293–294
AA
AC (Dichaetanthera Endl., Dissotidendron (A.Fern. & R.Fern.) Veranso-Libalah & G.Kadereit,
Heterotis Benth.)
AT (Anaheterotis Veranso-Libalah & G.Kadereit, Guyonia Naudin, Melastomastrum Naudin,
Tristemma Juss.)
GC (Dissotis and allies, Dupineta)
621–623
TGG
CCG (Dichaetanthera, Dissotis Benth. and allies, Dupineta (Sm.) Raf.)
TCG (Dissotis tubulosa (Sm.) Triana)
GCG (Pseudosbeckia)
CGG (Anaheterotis, Guyonia, Heterotis, Melastomastrum, Tristemma)
GCA (Dissotidendron)
716–720
AGC GA
TGT GT (Anaheterotis, Guyonia)
TGC GC (Dichaetanthera, Dissotidendron, Dissotis and allies, Dupineta, Heterotis)
TGG AC (Dissotis tubulosa)
775–777
GAA
CAT (Dissotis and allies, Dupineta)
GAG (Anaheterotis, Dichaetanthera, Guyonia, Heterotis, Melastomastrum, Tristemma)
TAG (Dissotidendron)
293–294
CT
TC (Dupineta)
CC (rest of the African Melastomateae genera)
ITS
accD-psaI
Argyrella
ITS
psbK-psbL
accD-psaI
108
G
T (rest of the African genera included in this study)
503
A
G (rest of the African genera included in this study)
518
A
G (rest of the African genera included in this study)
1296
G
T (rest of the African genera included in this study)
CCGTCC
Duplication of CCGTCC from position 258–263
Dissotidendron
ITS
264–269
–––––––
Deletion of GAGGCCC from position 271–277
A
C (rest of the African genera)
G (Argyrella)
623–624
GA
GG (Anaheterotis, Argyrella, Dichaetanthera, Heterotis, Melastomastrum, Pseudosbeckia, Tristemma)
CG (Dissotis and allies, Dupineta, Guyonia)
635–636
TG
CA (Anaheterotis, Argyrella, Guyonia, Pseudosbeckia)
TA (Dichaetanthera, Dissotis and allies, Dupineta, Heterotis, Melastomastrum, Tristemma)
GA (Dissotis tubulosa)
861
C
A (Dissotis sect. Dissotis)
T (rest of the African Melastomateae genera)
psbK-psbL
93
A
T (Dupineta, Heterotis)
G (Dichaetanthera, Guyonia)
– (rest have an indel [deletion])
accD-psaI
334
G
T (rest of the African Melastomateae genera)
– (Anaheterotis, Argyrella, and some Dissotis sect. Squamulosae)
427
A
C (rest of the African Melastomateae genera)
– (some Dissotis sect. Squamulosae)
325–326
AC
GT (A. irvingiania, Dissotis sect. Macrocarpae)
703–704
TT
– – (rest of the African Melastomateae genera have indels, deletion)
705–706
TG
CA (Dissotis tubulosa)
CG (rest of the African Melastomateae genera)
271–277
606
Dupineta
ITS
AT (rest of the African Melastomateae genera)
602
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Table 4. Continued.
Position in
alignment
241
psbK-psbL
Specific
substitutions
Comment
T
G (rest of the African Melastomateae genera)
259
A
T (rest of the African Melastomateae genera)
292–293
T
C (rest of the African Melastomateae genera)
G
T (rest of the African Melastomateae genera)
TT
CG (Guyonia)
AT (Melastomastrum theifolium)
CT (rest of the African Melastomateae genera)
128
A
C (rest of the African Melastomateae genera)
435
C
G (rest of the African Melastomateae genera)
– (some Dissotis sect. Squamulosae)
450
G
T (rest of the African Melastomateae genera)
GT
AT (rest of the African Melastomateae genera)
AG (some Melastomastrum)
T
G (rest of the African Melastomateae genera)
342
446–447
accD-psaI
984–985
Guyonia
ITS
127
711
T
C (rest of the African Melastomateae genera)
713
T
C (rest of the African Melastomateae genera)
A
G (rest of the African Melastomateae genera)
739
––––––––
accD-psaI 1032–1039
CATTATTT (rest of the African Melastomateae genera)
––
CA (rest of the African Melastomateae genera)
T
C (Dichaetanthera, Dissotis and allies, Guyonia)
A (Anaheterotis, Argyrella, Dissotidendron, Dupineta, Pseudosbeckia, Melastomastrum, Tristemma)
203–205
CAA
Duplication of CAA from the position 200–202
235–236
TT
TG (Anaheterotis, Argyrella, Dichaetanthera, Dissotidendron, Guyonia, Melastomastrum, Tristemma)
CG (Dissotis and allies, Dupineta)
1045–1046
Heterotis
ITS
64
psbK-psbL
accD-psaI
253
T
A (rest of the African Melastomateae genera)
323
C
A (rest of the African Melastomateae genera)
1154
G
T (rest of the African Melastomateae genera)
Melastomastrum
psbK-psbL
163
T
C (rest of the African Melastomateae genera)
261
T
C (rest of the African Melastomateae genera)
CA
CC (Tristemma)
TG (rest of the African Melastomateae genera)
C
– (rest of the African Melastomateae genera have an indel)
accD-psaI 1181–1182
Tristemma
ITS
psbK-psbL
accD-psaI
71
C
T (rest of the African Melastomateae genera)
67–68
––
AT (rest of the African Melastomateae genera)
78–79
CG
CC (Anaheterotis, Argyrella, Dichaetanthera, Guyonia, Melastomastrum,)
GC (Dissotidendron, Dissotis and allies, Heterotis, Pseudosbeckia)
47
130
C
T (rest of the African Melastomateae genera)
86
A
T (rest of the African Melastomateae genera)
T
C (rest of the African Melastomateae genera)
CC
CA (Melastomastrum)
TG (rest of the African Melastomateae genera)
1045
1174–1175
Melastomateae genera: Anaheterotis, Argyrella, Dichaetanthera, Dissotidendron, Dissotis and allies, Dupineta, Heterotis, Guyonia, Melastomastrum, Pseudosbeckia and Tristemma.
Positions are based on the sequence alignments provided in the electronic supplementary material.
Version of Record
603
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
the holotype. We selected K000313170 as the lectotype because
it has Pobéguin’s original label with his collection number, the
specimen probably studied by Hutchinson and Dalziel, is in
agreement with the original description, is complete and well
preserved.
1981 – Holotype: Angola, Huíla, Lubango, Ferrão da
Sola, Welwitsch 914 (G-DC barcode G00319500!; isotypes:
BM barcode BM000902303!, C barcode C10000969!, K
barcode K000313397!, LD barcode 1246147!, LISU barcode LISU209421!, M barcode M-0105939!, P barcode
P00412567!, PRE barcode PRE0341342-0!).
Note. – One collection of Heterotis angolensis from Poli,
north Cameroon (Raynal 13096, P barcode P05222487!) was
cited in the Flore du Cameroun (Jacques-Félix, 1983). However,
we identify this collection as Argyrella canescens (E.Mey. ex
Graham) Harv.
2. Argyrella Naudin in Ann. Sci. Nat., Bot., sér. 3, 13: 300.
1850 ≡ Dissotis sect. Argyrella (Naudin) Triana in Trans.
Linn. Soc. London 28: 58. 1872 (“1871”) ≡ Dissotis subg.
Argyrella (Naudin) A.Fern. & R.Fern. in Bol. Soc. Brot.,
sér. 2, 43: 289. 1969 ≡ Heterotis sect. Argyrella (Naudin)
Jacq.-Fél. in Adansonia, sér. 2, 20: 419. 1981 – Type (designated by Fernandes & Fernandes in Bol. Soc. Brot., sér.
2, 43: 289. 1969): A. incana (E.Mey. ex Walpers) Naudin
(= A. canescens (Graham) Harv.).
Morphological diagnosis. – Argyrella differs from Anaheterotis, its closest African relative, by the following combination of characters: erect herbs with soft tomentose stellate and/
or glandular hairs on the entire plant, inflorescence paniculate,
intersepalar appendages reduced or absent and large stamens
with tricuspidate appendages while smaller stamens have
bilobed appendages. Anaheterotis is an erect glabrous herb
except for its prominently ciliate leaf margins.
Molecular diagnosis. – Argyrella also differs from the
other African Melastomateae genera at the following nrITS nucleotide positions: AC/AT/GC-to-AA substitution at positions
293–294, CCG/TCG/GCG/CGG/GCA-to-TGG substitution
at positions 621–623, TGTGT/TGCCC/TGGAC-to-AGCGA
substitution at positions 716–720 and CAT/GAG/TAG-to-GAA
substitutions at positions 775–777; two psbK-psbL nucleotide
substitutions, TC/CC-to-CT at positions 293–294; and four
accD-psaI nucleotide substitutions, T-to-G at positions 108
and 1296, and G-to-A at positions 503 and 518 and T-to-G at
position 1296 (summarised in Table 4).
Argyrella amplexicaulis (Jacq.-Fél.) Veranso-Libalah &
G.Kadereit, comb. nov. ≡ Dissotis gilgiana Hutch.
& Dalziel, Fl. W. Trop. Afr. 1: 212. 1927, nom. illeg. ≡
Dissotis amplexicaulis Jacq.-Fél. in Bull. Mus. Natl. Hist.
Nat., sér. 2, 8: 108. 1936 ≡ Heterotis amplexicaulis (Jacq.Fél.) Aké Assi in Bull. Mus. Natl. His. Nat., sér. 4, 9:
459. 1987 – Lectotype (designated here): Mali, Sicoro,
Chevalier 228 (K barcode K000313145!; isolectotype: P
barcode P00412566!).
Note. – In the protologue of Dissotis gilgiana, Hutchinson
& Dalziel (1927: 212) cited Chevalier 228 as the type specimen.
We located two specimens housed in P and K, both part of the
same field collection. However, neither of them was designated
as the holotype. Since Hutchinson and Dalziel mainly worked
in K, they likely studied the specimen housed in K. As such,
we selected K000313145 as the lectotype and P00412566 as
the isolectotype.
Argyrella angolensis (Cogn.) Veranso-Libalah & G.Kadereit,
comb. nov. ≡ Dissotis angolensis Cogn. in Candolle &
Candolle, Monogr. Phan. 7: 371. 1891 ≡ Heterotis angolensis (Cogn.) Jacq.-Fél. in Adansonia, sér. 2, 20: 420.
604
Argyrella bambutorum (Gilg & Ledermann ex Engl.)
Veranso-Libalah & G.Kadereit, comb. nov. ≡ Dissotis
bambutorum Gilg & Ledermann ex Engl., Veg. Erde
9(3,2): 750. 1921 ≡ Heterotis angolensis var. bambutorum
(Gilg & Ledermann ex Engl.) Jacq.-Fél. in Adansonia,
sér. 2, 20: 420. 1981 – Neotype (designated by JacquesFélix in Satabié & Leroy, Fl. Cameroun 24: 47. 1983):
Cameroon, Bambuto Mts., Jacques-Félix 2662 (P barcode
P05222490!).
Note. – Dissotis bambutorum (≡ Argyrella bambutorum in
the present treatment) was treated by Jacques-Félix (1983) in
Flore du Cameroun as an isomorphic variety of the otherwise
dimorphic Heterotis angolensis (≡ Argyrella angolensis in the
present treatment). However, the stamens of A. bambutorum
are isomorphic.
Argyrella linearis (Jacq.-Fél.) Veranso-Libalah & G.Kadereit,
comb. nov. ≡ Dissotis linearis Jacq.-Fél. in Bull. Inst. Franç.
Afrique Noire 15: 980. 1953 – Lectotype (designated
here): Guinea, Dinguiraye, Jacques-Félix 1467 (P barcode P00412580!; isolectotype: K barcode K000313147!).
Note. – In the protologue of Dissotis linearis, JacquesFélix (1953: 80) cited Jacques-Félix 1467 as the type specimen.
We located two specimens housed in P and K, all part of the
same field collection as the type. However, neither of them
was designated as the holotype. We selected P00412580 as the
lectotype because it has Jacques-Félix’s original label with
his collection number, the specimen is in agreement with the
original description, is likely the specimen studied by JacquesFélix, is complete and well preserved.
Argyrella sessilis (Hutch. ex Brenan & Keay) Veranso-Libalah
& G.Kadereit, comb. nov. ≡ Dissotis sessilis Hutch. ex
Brenan & Keay in Kew Bull. 7: 545. 1953 (“1952”) –
Holotype: Sierra Leone, Bitumane Peak, Glanville 386
(K barcode K000313144!).
Other included species. – Argyrella canescens (E.Mey.
ex Graham) Harv.
3. Dissotidendron (A.Fern. & R.Fern.) Veranso-Libalah &
G.Kadereit, stat. nov. ≡ Dissotis subg. Dissotidendron
A.Fern. & R.Fern. in Bol. Soc. Brot., sér. 2, 43: 289. 1969
– Type: Dissotidendron melleri (Hook.f.) Veranso-Libalah
& G.Kadereit.
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Morphological diagnosis. – Dissotidendron can be distinguished from Dichaetanthera Endl. by the following combination of characters: trees or shrubs, mostly deciduous; leaf
surfaces rugose; flowers 5-merous; calyx-lobes persistent with
no intersepalar appendages. Dichaetanthera are trees or shrubs
with 4-merous flowers.
Molecular diagnosis. – Dissotidendron also differs from
the other African Melastomateae genera in the following nrITS
nucleotide positions: a tandem duplication of CCGTCC (264–
269) from positions 258–263, deletion of GAGGCCC at positions 271–277, C/G-to-A substitution at position 606, GG/CGto-GA substitution at positions 623–624, CA/TA/GA-to-TG at
positions 635–636, A/T-to-C at position 861; one psbK-psbL
nucleotide subtitution, T/G/indel-to-A at position 93; and two
accD-psaI nucleotide substitutions: T/indel-to-G at position
334, C/indel-to-A at position 427 (summarised in Table 4).
Wild. in Ann. Mus. Congo Belge, Bot., sér. 4, 2(1): 116. 1913
≡ Dissotis caloneura Gilg ex Engl., Veg. Erde 9(3,2): 749.
1921, isonym – Neotype (designated here): Tanzania, T4,
Kigoma District, Mt. Livandabe (Lubalisi), Bidgood & al.
4178 (P barcode P05281276!; isoneotypes: C!, K).
= Dissotis venulosa Hutch., Botanist S. Afr.: 512. 1946, in
adnot. – Type: Zambia, Mbala District, Lake Chila,
Hutchinson & Gillett 3887 (K barcode K000313393!).
Note. – In the protologue of Dissotis caloneura, De
Wildeman (1913: 116) cited a specimen collected from Mt.
Minta, M’Lolose by Kässner with collection number 2960 (13
May 1908) and housed in Berlin as the type. Unfortunately,
this specimen was probably destroyed during World War II
since there was no trace of it in B but we were able to locate
three specimens from the same field collection housed in BM
(BM000902299!), P (P05223126!) and K!. The BM specimen is
a mounted sheet of D. trothae Gilg and a single flower fragment
of D. caloneura in an envelope while the other specimens in P
and K are only mounted sheets of D. trothae.
Also, in another protologue of Dissotis caloneura, Engler
(1921: 749) cited a specimen collected from Katanga, Mt.
Nunta by Kässner with collection number 2960 (1908 expedition). We were unable to locate this specimen which was
also likely stored in B and destroyed during the World War
II. However, we think this was the same specimen cited by
De Wildeman of which only a flower fragment is available
and housed in BM. As a result, we selected P05281276 as the
neotype.
Dissotidendron apricum (Gilg ex Engl.) Veranso-Libalah &
G.Kadereit, comb. nov. ≡ Dissotis aprica Gilg ex Engl.,
Veg. Erde 9(3,2): 749. 1921 – Lectotype (designated here):
Tanzania, Lindi District, Rondo (Muera) Plateau, Busse
2577 (EA barcode EA000003766!).
Note. – In the protologue of Dissotis aprica, Engler (1921:
749) cited a specimen collected from Tanzania, Lindi District,
Rondo (Muera) Plateau by Busse 2577 and housed in B as the
type. Unfortunately, this specimen was likely destroyed during
World War II since there was no trace of it in B. We located and
selected EA000003766 as the lectotype because it is probably
part of the same field collection and in good condition.
Dissotidendron arborescens (A.Fern. & R.Fern.) VeransoLibalah & G.Kadereit, comb. nov. ≡ Dissotis arborescens
A.Fern. & R.Fern. in Bol. Soc. Brot., sér. 2, 29: 51. 1955
– Holotype: Tanzania, Iringa District, Sao Hill, Greenway
6176 (EA barcode EA000001983!; isotypes: K barcode
K000313066!, PRE barcode PRE0341343-0!).
Dissotidendron bussei (Gilg ex Engl.) Veranso-Libalah &
G.Kadereit, comb. nov. ≡ Dissotis bussei Gilg ex Engl.,
Veg. Erde 9(3,2): 749. 1921 – Neotype (designated here):
Tanzania, Dodomo, Mpwapwa District, Kibakwe division, Mbuga yard, Mbuga village, Mafermera subvillage,
Rubeho mountains, Langangulu river ca. 200–300 m S
and N of foot track to Dibulilo subvillage, Kayombo & al.
4914 (MO!; isoneotype: EA!).
Note. – In the protologue of Dissotis bussei, Engler (1921:
749) cited a specimen collected from Tanzania, Mpwapwa
District, Kiboriani Mts. by Busse with no collection number
and housed in B as the type. Unfortunately, this specimen was
likely destroyed during World War II since there was no trace
of it in B and we were unable to locate any specimen from the
same field collection in other herbaria. We selected Kayombo
& al. 4914 housed in MO as the neotype because it is complete
and well preserved.
Dissotidendron caloneurum (Gilg ex De Wild.) Veranso-Libalah
& G.Kadereit, comb. nov. ≡ Dissotis caloneura Gilg ex De
Dissotidendron caloneurum var. confertiflorum (A.Fern. &
R.Fern.) Veranso-Libalah & G.Kadereit, comb. nov. ≡
Dissotis caloneura var. confertiflora A.Fern. & R.Fern.
in Bol. Soc. Brot., sér. 2, 34: 59, fig. 1. 1960 – Holotype:
Tanzania, Ufipa District, Kito (Keto) Mt., Richards 6176
(K barcode K000313068!).
Dissotidendron caloneurum var. pilosum (A.Fern. & R.Fern.)
Veranso-Libalah & G.Kadereit, comb. nov. ≡ Dissotis
caloneura var. pilosa A.Fern. & R.Fern. in Kirkia 1: 73,
fig. 7. 1961 – Holotype: Zambia, Luanshya, Fanshawe 3118
(K barcode K000313391!).
Dissotidendron caloneurum var. setosius (A.Fern. & R.Fern.)
Veranso-Libalah & G.Kadereit, comb. nov. ≡ Dissotis
caloneura var. setosior A.Fern. & R.Fern. in Bol. Soc.
Brot., sér. 2, 30: 171, fig. 5. 1956 – Holotype: Tanzania,
Kigoma District, Kafulu, Eggeling 6172 (EA barcode
EA000001984!; isotype: K barcode K000313067!).
Dissotidendron cordatum (Gilg) Veranso-Libalah & G.Kadereit, comb. nov. ≡ Dissotis cordata Gilg in Engler, Monogr.
Afrik. Pflanzen-Fam. 2: 17. 1898 – Neotype (designated
here): Democratic Republic of the Congo, Kakalalwe,
Malaisse & al. 533 (BR barcode BR0000017293129!; isoneotype: BR barcode BR0000017293112!).
= Dissotis derriksiana P.A.Duvign. in Bull. Soc. Roy.
Bot. Belgique 90: 257. 1958, syn. nov. – Holotype:
Version of Record
605
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
Democratic Republic of the Congo, Haut-Katanaga,
Dikuluwe, Duvigneaud & Damblon 3088D (BRLU barcode BRLU0000090023040!; isotype: BRLU barcode
BRLU0000090023248!).
Note. – In the protologue of Dissotis cordata, Gilg (1898:
17) cited a specimen collected from “Oberes Congogebiet,
Baschilange, am Luluafluss” by Pogge with collection number
134. Unfortunately, this specimen was likely destroyed during
World War II since there was no trace of it in B and we were
unable to locate any specimen from the same field collection
in other herbaria. We selected 0000017293129 in BR as the
neotype because it is complete and well preserved.
Also, the original description of Dissotis derriksiana
by Duvigneaud (1958: 257) clearly matches that of D. cordata. Hence, Dissotis derriksiana is treated as a synonym of
Dissotidendron cordatum.
= Dissotis whytei Baker in Bull. Misc. Inform. Kew 1897: 267.
1897 – Holotype: Malawi, Zomba Mt., Whyte s.n. (K barcode K000313389!).
Dissotidendron dichaetantheroides (Wickens) VeransoLibalah & G.Kadereit, comb. nov. ≡ Dissotis dichaetantheroides Wickens in Kew Bull. 29(1): 141. 1974 – Holotype: Tanzania, Morogoro District, Nguru Mts., Drummond
& Hemsley 1960 (K barcode K000313063!; isotypes: EA
barcodes EA000001980! & EA000003753!, K barcode
K000313062!).
Dissotidendron glandulicalyx (Wickens) Veranso-Libalah &
G.Kadereit, comb. nov. ≡ Dissotis glandulicalyx Wickens
in Kew Bull. 29(1): 142. 1974 – Holotype: Tanzania,
Mpanda District, Kungwe Mt., Harley 9542 (K barcode
K00313065!).
Dissotidendron johnstonianum (Baker f.) Veranso-Libalah &
G.Kadereit, comb. nov. ≡ Dissotis johnstoniana Baker f. in
Trans. Linn. Soc. London, Bot. 4(1): 14. 1894 – Holotype:
Malawi, Mlanje Mt., Whyte 74 (K barcode K000313388!).
Dissotidendron johnstonianum var. strigosum (Brenan)
Veranso-Libalah & G.Kadereit, comb. nov. ≡ Dissotis
johnstoniana var. strigosa Brenan in Mem. New York
Bot. Gard. 8: 440. 1954 – Holotype: Malawi, Mlanje Mt.,
Chambe Plateau, Brass 16757 (K barcode K000313387!;
isotypes: BM barcode BM000580060!, BR barcode
BR0000006494551!; PRE barcode PRE0601890-0!; SRGH
barcode SRGH0106508-0!).
Dissotidendron lanatum (A.Fern. & R.Fern.) VeransoLibalah & G.Kadereit, comb. nov. ≡ Dissotis lanata
A.Fern. & R.Fern. in Bol. Soc. Brot., sér. 2, 43: 294.
1969 – Holotype: Malawi, Northern Prov., Mafinga Mts.,
Robson 533 (K barcode K000049527!; isotypes: BM barcode BM000902405!, LISC barcode LISC 002950!, SRGH
barcode SRGH0106556-0!).
Dissotidendron melleri (Hook.f.) Veranso-Libalah & G.Kadereit,
comb. nov. ≡ Dissotis melleri Hook.f. in Oliver, Fl. Trop.
Afr. 2: 451. 1871 – Holotype: Malawi, Mt. Chiradzura,
Manganja range, Meller s.n. (K barcode K000313390!).
606
Dissotidendron melleri var. greenwayi (A.Fern. & R.Fern.)
Veranso-Libalah & G.Kadereit, comb. nov. ≡ Dissotis
greenwayi A.Fern. & R.Fern. in Bol. Soc. Brot., sér. 2,
30: 172. 1956 ≡ Dissotis melleri var. greenwayi (A.Fern. &
R.Fern.) A.Fern. & R.Fern. in Bol. Soc. Brot., sér. 2, 46: 68
1972 – Holotype: Tanzania, Rungwe District, R. Kiwira,
Greenway 8410 (EA; isotype: K barcode 00313064!).
= Dissotis rubroviolacea Gilg in Bot. Jahrb. Syst. 30: 365. 1901
– Lectotype (designated here): Tanzania, N. Mt. Rungwe,
Goetze 1136 (BM barcode BM000902402!: isolectotype:
P barcode P00412542!).
Note. – In the protologue of Dissotis rubroviolacea, Gilg
(1901: 365) cited a specimen collected from Tanzania, N. Mt.
Rungwe by Goetze with collection number 1136 and housed
in B as the type. This specimen was likely destroyed during
World War II since there was no trace of it in B but we were
able to locate two specimens from the same field collection
housed in BM and P. As a result we selected BM000902402,
as the lectotype and P00412542 as the isolectotype.
Dissotidendron polyanthum (Gilg) Veranso-Libalah & G.Kadereit, comb. nov. ≡ Dissotis polyantha Gilg in Engler, Monogr.
Afrik. Pflanzen-Fam. 2: 16. 1898 – Neotype (designated
here): Tanzania, Morogoro, Kilosa district, T6, Ukwiva forest reserve, Kifuluma area, Festo & al. 1901 (MO barcode
MO-2721881!; isoneotype: C!).
Note. – In the protologue of Dissotis polyantha, Gilg (1898:
16) cited a specimen collected from Tanzania, Usagara, Kifuru
(Central Uluguru) by Stuhlmann with collection number 9081
and housed in B as the type. Unfortunately, this specimen was
likely destroyed during World War II since there was no trace
of it in B and we were unable to locate any specimen from the
same field collection in other herbaria. As a result, we selected
MO-2721881 as the neotype.
4. Dupineta Raf., Sylva Tellur.: 101. 1838 ≡ Dissotis subg.
Dupineta (Raf.) A.Fern. & R.Fern. in Bol . Soc. Brot., sér.
2, 43: 288. 1969 – Type: Dupineta multiflora (Sm.) Raf.
Morphological diagnosis. – Dupineta differs from Pseudosbeckia A.Fern. & R.Fern. by the following combination
of characters: perennial shrublets, inflorescence a panicle of
cymes, flowers 5-merous with small caducous bracts, calyxlobes early caducous, intersepalar appendages absent, stamens
dimorphic, appendages bifid, calyx-tube accrescent in fruit
and developing distinct longitudinal ribs. Pseudosbeckia is
a shrub with flowers 5-merous, sepals tardily caducous, stamens isomorphic, fruiting capsule widely campanulate and
not accrescent.
Molecular diagnosis. – Dupineta also differs from the
other African Melastomateae genera in the following nrITS
nucleotide positions: GT/AT-to-AC substitution at positions
325–326, TT insertion at positions 703–704, CA/CG-to-TG
substitution at positions 705–706; five psbK-psbL nucleotide
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
substitutions: a G-to-T substitution at position 241, T-to-G substitution at position 259, C-to-T substitution at position 293,
T-to-G substitution at position 342, CG/AT/CT-to-TT substitution at positions 446–447; and four accD-psaI nucleotide
substitutions: C-to-A substitution at position 128, G/indel-to-C
substitution at position 435, a T-to-G substitution at position
450, AT/AG-to-GT substitution at positions 984–985 (summarised in Table 4).
= Afzeliella Gilg in Engler, Monogr. Afr. Pflanzen-Fam. 2: 4.
1898 – Type: Afzeliella ciliata (Hook.f.) Gilg.
= Heterotis sect. Cyclostemma Benth. in Hooker, Niger Fl.: 349.
1849 – Type: Heterotis antennina Benth.
Morphological diagnosis. – Guyonia differs from
Argyrella and Anaheterotis by the following characters: mostly
herbaceous geophytes rooting at the nodes, except G. jacquesii
(A.Chev.) Veranso-Libalah & R.D.Stone which is lignified; no
hypanthial emergences; calyx-tube glabrous (glandular hairs
in G. jacquesii); intersepalar appendages present or absent;
flowers solitary or in cymes of 1–3 flowers. Argyrella and
Anaheterotis are erect herbs and the inflorescence a panicle.
Molecular diagnosis. – Guyonia also differs from the
other African Melastomateae genera in the following nrITS
nucleotide positions: G-to-T substitution at position 127, C-to-T
substitution at positions 711 and 713 and G-to-A substitution
at position 739; and two accD-psaI nucleotide substitutions:
deletion of CATTATTT at positions 1032–1039 and deletion of
CA at positions 1045–1046 (summarised in Table 4).
Dupineta brazzae (Cogn.) Veranso-Libalah & G.Kadereit, comb.
nov. ≡ Dissotis brazzae Cogn. in Candolle & Candolle,
Monogr. Phan. 7: 372. 1891 – Lectotype (designated here):
Gabon, Franceville, de Brazza 43 (P barcode P00412522!;
isolectotypes: BR barcode BR0000006493585!, P barcodes
P00412521! & P00412523!).
= Dissotis tanganjikae Kraenzl. in Vierteljahrsschr. Naturf.
Ges. Zürich 76: 150. 1931 – Holotype: Democratic Republic
of the Congo, Katanga, Kibandu, Kassner 3055a (Z barcode Z-000015029!).
Note. – In the protologue of Dissotis brazzae, Cogniaux
(1891: 372) cited de Brazza 43, housed in P as the type specimen. We located four specimens, three housed in P and one
in BR, all part of the same field collection but none of them
was designated as the holotype. We selected P00412522 as the
lectotype.
Dupineta hensii (Cogn.) Veranso-Libalah & G.Kadereit, comb.
nov. ≡ Dissotis hensii Cogn. in Candolle & Candolle,
Monogr. Phan. 7: 372. 1891 – Holotype: Democratic
Republic of the Congo (Zaire), Bangala, Hens 129 (BR barcode BR0000006493509!; isotypes: P barcode P00412536!;
Z barcodes Z-000015020! & Z-000015021!).
Dupineta loandensis (Exell) Veranso-Libalah & G.Kadereit,
comb. nov. ≡ Dissotis loandensis Exell in J. Bot. 67
(Suppl. 1): 179. 1929 – Lectotype (designated here):
Angola, Luanda, Cazengo District, Gossweiler 697
(BM barcode BM000902404!; isolectotypes: K barcode
K00313356!; P barcode P00412541!).
Note. – In the protologue of Dissotis loandensis, Exell
& al. (1929: 179) cited Gossweiler 697 as the type specimen.
We located three specimens housed in BM, K and P from the
same field collection but none of them was designated as the
holotype. Since Exell worked in BM, we think he likely studied
BM000902404 and therefore it was selected as the lectotype.
Dupineta pauwelsii (Jacq.-Fél.) Veranso-Libalah & G.Kadereit,
comb. nov. ≡ Dissotis pauwelsii Jacq.-Fél. in Bull. Jard.
Bot. Natl. Belg. 44(1–2): 162. 1974 – Holotype: Democratic
Republic of the Congo, District du Bas-Congo, environs de
Kimvula, Pauwels 2719 (BR barcode BR0000006422837!;
isotype: P barcode P06602990!).
Other included species. – Dupineta multiflora (Sm.) Raf.
5. Guyonia Naudin in Ann. Sci. Nat., Bot., sér. 3, 14: 149. 1850
– Type: Guyonia tenella Naudin.
Guyonia antennina (Sm.) Veranso-Libalah & R.D.Stone,
comb. nov. ≡ Osbeckia antennina Sm. in Rees, Cycl. 25.
1813 ≡ Heterotis antennina (Sm.) Benth. in Hooker, Niger
Fl.: 349. 1849 ≡ Dissotis antennina (Sm.) Hook.f. in Oliver,
Fl. Trop. Afr. 2: 449. 1871 – Lectotype (designated here):
Sierra-Leone, Afzelius s.n. (BM barcode BM000902302!;
isolectotype: K barcode K000313180!).
Note. – In the protologue of Osbeckia antennina, Smith
(1813) cited Afzelius s.n. as the type specimen. We located two
specimens housed in BM and K likely from the same field
collection but neither of them was designated as the holotype.
Since Smith worked in BM, he likely studied BM000902302
and therefore it was selected as the lectotype.
Guyonia arenaria (Jacq.-Fél.) Veranso-Libalah & R.D.Stone,
comb. nov. ≡ Heterotis arenaria Jacq.-Fél. in Bull.
Mus. Natl. Hist. Nat., B, Adansonia, sér. 4, 9(3): 255.
1988 (“1987”) – Holotype: Gabon, 15 km environ au
Sud de Mayoumba, De Wilde & al. 629 (WAG barcode
WAG0002339!; isotype: WAG barcode WAG0002338!).
Guyonia ciliata Hook.f. in Oliver, Fl. Trop. Afr. 2: 443. 1871
≡ Afzeliella ciliata (Hook.f.) Gilg in Engler, Monogr. Afr.
Pflanzen-Fam. 2: 5. 1898 – Neotype (designated here):
Liberia, Lofa, East of Nekebuzu, Jongkind & al. 11859
(WAG barcode WAG.1488062!; isoneotypes: BR barcode
BR00000016915534!, MO barcode MO-2970207!).
= Guyonia intermedia Cogn. in De Wildeman & Durand,
Pl. Thonn. Congol.: 30. 1900 ≡ Afzeliella intermedia
(Cogn.) Gilg in Mildbraed, Wiss. Erg. Deut. Zentr.Afr. Exped., Bot.: 582. 1913 – Holotype: Democratic
Republic of the Congo (Zaire), Ngali, Thonner 21 (BR
barcode BR0000008953339!; isotype: BR barcode
BR0000008953667!).
Note. – In the protologue of Afzeliella ciliata, Gilg (1898:
5) cited a specimen collected from Sierra-Leone by Afzelius
with no collection number and housed in B as the type.
Version of Record
607
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
Unfortunately, this specimen was probably destroyed during
World War II since there was no trace of it in B and we were
unable to locate any specimen from the same field collection
in other herbaria. We selected WAG.1488062 as the neotype.
Guyonia pygmaea (A.Chev. & Jacq.-Fél.) Veranso-Libalah
& R.D.Stone, comb. nov. ≡ Dissotis pygmaea A.Chev.
& Jacq.-Fél. in Bull. Mus. Natl. Hist. Nat., sér. 2, 4: 685.
1932 ≡ Heterotis pygmaea (A.Chev. & Jacq.-Fél.) Jacq.Fél. in Adansonia, sér. 2, 20: 419. 1981 – Lectotype (designated here): Guinea, environs de Kindia, JacquesFélix 195 (P barcode P00412585!; isolectotype: P barcode
P00412586!).
Note. – In the protologue of Dissotis pgymeae, Chevalier
& Jacques-Félix (Jacques-Félix, 1932: 685) cited Jacques-Félix
195 as the type specimen. We located two specimens housed
in P, both part of the same field collection but neither of them
was designated as the holotype. We selected P00412585 as the
lectotype because it has Jacques-Félix’s original label with
his collection number, is in agreement with the original description, is complete, well preserved and likely the specimen
studied by Chevalier and Jacques-Félix.
Guyonia cinerascens (Sm.) Veranso-Libalah & R.D.Stone,
comb. nov. ≡ Dissotis cinerascens Hutch. in Bull. Misc.
Inform. Kew 1921: 372. 1921 ≡ Heterotis cinerascens
(Hutch.) Jacq.-Fél. in Adansonia, sér. 2, 20: 419. 1981 –
Holotype: Northern Nigeria, Neill’s Valley, 2 miles east of
Government Station, Naraguta, H.V. Lely 259 (K barcode
K000313164!).
Guyonia entii (J.B.Hall) Veranso-Libalah & R.D.Stone, comb.
nov. ≡ Dissotis entii J.B.Hall in Kew Bull. 24: 346. 1970
≡ Heterotis entii (J.B.Hall) Jacq.-Fél. in Adansonia, sér. 2,
20: 419. 1981 – Holotype: Ghana, West Region, Aiyinasi,
edge of small stream in forest, Hall GC35533 (K barcode
K000313173!; isotypes: GC barcode GC000007519!, LISC
barcode LISC 002431!, P barcode P00412575!, WAG barcode WAG0002331!).
Guyonia glandulosa (A.Fern. & R.Fern.) Veranso-Libalah &
R.D.Stone, comb. nov. ≡ Dissotis glandulosa A.Fern. &
R.Fern. in Bol. Soc. Brot., sér. 2, 43: 293. 1969 – Holotype:
Zambia, Mwinilunga Distr., sources of R. Zambesi,
dry sandy woodland, Robinson 5990 (SRGH barcode
SRGH0106559-0!).
Guyonia humilis (A.Chev. & Jacq.-Fél.) Veranso-Libalah &
R.D.Stone, comb. nov. ≡ Dissotis humilis A.Chev. & Jacq.Fél. in Bull. Mus. Natl. Hist. Nat., sér. 2, 4: 684. 1932 –
Holotype: Guinea, Kindia, Jacques-Félix 220 (P barcode
P00412576!).
Guyonia jacquesii (A.Chev.) Veranso-Libalah & R.D.Stone,
comb. nov. ≡ Dissotis jacquesii A.Chev. in Bull. Mus.
Natl. Hist. Nat., sér. 2, 4: 686. 1932 ≡ Heterotis jacquesii
(A.Chev.) Aké Assi in Bull. Mus. Natl. His. Nat., sér. 4,
9: 459. 1987 – Lectotype (designated here): Guinea,
Kindia and Mont Gangan, Jacques-Félix 240 (P barcode
P00412577!; isolectotypes: P barcodes P00412578! &
P00412579!).
Note. – In the protologue of Dissotis jacquesii, Chevalier in
Jacques-Félix (1932: 686) cited Jacques-Félix 240, housed in P
as the type specimen. We located three specimens in P, all part
of the same field collection but none of them was designated as
the holotype. We selected P00412577 as the lectotype because
it has Jacques-Félix’s original label with his collection number,
the specimen he probably studied, is in agreement with the
original description, is complete and well preserved.
Guyonia obamae (Lejoly & Lisowki) Veranso-Libalah & R.D.
Stone, comb. nov. ≡ Heterotis obamae Lejoly & Lisowki in
Syst. & Geogr. Pl. 69(2): 185. 1999 – Holotype: Equatorial
Guinea, Rio Muni, Ndote Sud, près du village Etembue,
Lejoly & Elad 98/112 (BRLU barcode BRLU0010836!).
608
Guyonia rupicola (Gilg ex Engl.) Veranso-Libalah & R.D.
Stone, comb. nov. ≡ Dissotis rupicola Gilg ex Engl., Veg.
Erde 9(3,2): 748. 1921 ≡ Heterotis rupicola (Gilg ex Engl.)
Jacq.-Fél. in Adansonia, sér. 2, 20: 419. 1981 – Lectotype
(designated here): Liberia, Dinklage 2139 (B barcode
B 10 0159359!; isolectotypes: B barcodes B 10 0159360!,
B 10 0159361! & B 10 0159362!).
= Dissotis glauca Keay in Kew Bull. 7: 545. 1953 (“1952”) ≡
D. rupicola Hutch. & Dalziel, Fl. W. Trop. Afr. 1: 211. 1927;
in Bull. Misc. Inform. Kew 1928: 221. 1928, nom. illeg.,
non D. rupicola Gilg ex Engl. – Holotype: Sierra Leone,
Sugar Loaf Mt., Scott-Elliot 3977 (K barcode K000313166!;
isotype: K barcode K000313168!).
Note. – In the protologue of Dissotis rupicola, Hutchinson
& Dalziel (1927: 748) cited Dinklage 2139, housed in B as the
type specimen. We located four specimens in B, all part of the
same field collection but none of them was designated as the
holotype. We selected B 10 0159359 as the lectotype because
it has Dinklage’s original label with his collection number, is
in agreement with the original description, is complete and
well preserved.
Guyonia seretii (De Wild.) Veranso-Libalah & R.D.Stone,
comb. nov. ≡ Dissotis seretii De Wild. in Ann. Mus. Congo
Belge, Bot., sér. 5, 2(3): 328. 1908 ≡ Heterotis seretii (De
Wild.) Jacq.-Fél. in Adansonia, sér. 2, 20: 419. 1981 –
Holotype: Democratic Republic of the Congo, Bokoyo,
Seret 587 (BR barcode BR0000006421205!).
Guyonia seretii var. gracilifolia (Wickens) Veranso-Libalah
& R.D.Stone, comb. nov. ≡ Dissotis seretii var. gracilifolia Wickens in Kew Bull. 29(1): 141. 1974 – Holotype:
Tanzania, Kigoma Distr., Kabogo Mts., Azuma 1014 (EA
barcode EA000001979!).
Guyonia sylvestris (Jacq.-Fél.) Veranso-Libalah & R.D.Stone,
comb. nov. ≡ Dissotis sylvestris Jacq.-Fél. in Bull. Mus.
Natl. Hist. Nat., sér. 2, 10: 632. 1939 (“1938”) ≡ Heterotis
sylvestris (Jacq.-Fél.) Jacq.-Fél. in Adansonia, sér. 4, 16
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
(2–4): 272. 1995 – Lectotype (designated here): Guinea,
Macenta, forêt du Ziama, Jacques-Félix 2088 (P barcode
P00412591!; isolectotype: K barcode K000313189!).
Note. – In the protologue of Dissotis sylvestris, JacquesFélix (1939: 632) cited Jacques-Félix 2088 as the type specimen. We located two specimens housed in P and K, both part of
the same field collection but neither of them was designated as
the holotype. We selected P00412591 as the lectotype because it
has Jacques-Félix’s original label with his collection number, is
in agreement with the original description, likely the specimen
he studied, is complete and well preserved.
7. Melastomastrum Naudin in Ann. Sci. Nat., Bot., sér. 3, 13:
296, t. 5. 1850 – Type: Melastomastrum capitatum (Vahl)
A.Fern. & R.Fern.
= Heterotis sect. Leiocalyx Planch. ex Benth. in Hooker, Niger
Fl.: 350. 1849 – Type: Heterotis segregata Benth.
= Heterotis sect. Wedeliopsis Planch. ex Benth. in Hooker, Niger
Fl.: 351. 1849 – Type: Heterotis theifolia (G.Don) Benth.
Morphological diagnosis. – Melastomastrum can be differentiated from Tristemma Juss., its closest relative, by the following combination of characters: subshrubs or woody herbs with
large 5-merous flowers surrounded by an involucre of persistent bracts, calyx-tube campanulate, stamens dimorphic (except
in M. porteresii (Jacq.-Fél.) Veranso-Libalah & G.Kadereit) and
fruits capsular with cochleate seeds. Tristemma has calyx-tubes
with a ring of bristles midway, stamens isomorphic (except
T. cornifolium (Benth.) Triana) and fruits baccate.
Molecular diagnosis. – Melastomastrum also differs from
the other African Melastomateae genera by two psbK-psbL nucleotide substitutions: C-to-T substitution at positions 163 and
261; and one accD-psaI nucleotide substitutions: CC/TG-to-CA
substitution at positions 1181–1182 (summarised in Table 4).
Other included species. – Guyonia tenella Naudin.
6. Heterotis Benth. in Hooker, Niger Fl.: 347. 1849 ≡ Dissotis
sect. Heterotis (Benth.) Hook.f. in Oliver, Fl. Trop. Afr.
2: 447. 1871 ≡ Dissotis subg. Heterotis (Benth.) A.Fern.
& R.Fern. in Bol. Soc. Brot., sér. 2, 43: 286. 1969 – Type:
Heterotis rotundifolia (Sm.) Jacq.-Fél.
Morphological diagnosis. – Heterotis can be distinguished
from the other African Melastomateae genera by the following
combination of characters: decumbent herbs, leaves orbicular to ovate-lanceolate, hypanthium with stalked emergences,
flowers large, 5-merous, calyx-lobes persistent, apex of intersepalar appendages with stellate emergences, seeds cochleate
and visibly arillate.
Molecular diagnosis. – Heterotis also differs from the
other African Melastomateae genera in the following nrITS
nucleotide positions: C/A-to-T substitution at position 64, tandem duplication of CAA at positions 203–205, TG/CG-to-TT
substitution at positions 235–236; two psbK-psbL nucleotide
substitutions: A-to-T substitution at position 253, A-to-C substitution at position 323; and one accD-psaI nucleotide substitution: T-to-G substitution at position 1154 (summarised in
Table 4).
Heterotis cogniauxiana (A.Fern & R.Fern.) Veranso-Libalah
& G.Kadereit, comb. nov. ≡ Osbeckia welwitschii Cogn.
in Candolle & Candolle, Monogr. Phan. 7: 333. 1891 ≡
Dissotis cogniauxiana A.Fern. & R.Fern. in Garcia de Orta
2: 175. 1954 (in obs.); in Bol. Soc. Brot., sér. 2, 28: 67. 1954
– Holotype: Angola, Pungo Andongo, rio Cuanza, pr. de
Sansamanda, Welwitsch 907 (G-DC barcode G00319429!;
isotypes: BM barcode BM000902297!, COI barcode
COI00005417!, K barcode K000313363!, LISU barcode
LISU209417!, P barcode P000412573!)
Heterotis fruticosa (Brenan) Veranso-Libalah & G.Kadereit,
comb. nov. ≡ Dissotis rotundifolia var. fruticosa Brenan in
Kew Bull. 5: 227. 1950 ≡ Dissotis fruticosa (Brenan) Brenan
& Keay in Kew Bull. 7: 547. 1953 (“1952”) – Holotype:
Nigeria, Ondo, Keay 22569 (K barcode K000313153!).
Other included species. – Heterotis buettneriana (Cogn.
ex Büttner) Jacq.-Fél., H. decumbens (P.Beauv.) Jacq.-Fél.,
H. prostrata (Thonn.) Benth., H. rotundifolia (Sm.) Jacq.-Fél.
Melastomastrum porteresii (Jacq.-Fél.) Veranso-Libalah &
G.Kadereit, comb. nov. ≡ Osbeckia porteresii Jacq.-Fél. in
Bull. Mus. Natl. Hist. Nat., sér. 2, 10: 630. 1939 (“1938”) –
Lectotype (designated here): Guinea (Guinée française),
Mont Nimba, Portères 3175 (P barcode P00412494!; isolectotype: P barcode P00412494!).
Note. – In the protologue of Osbeckia porteresii, JacquesFélix (1939: 630) cited Portères 3175 as the type specimen. We
located two specimens housed in P, both part of the same field
collection but neither of them was designated as the holotype.
We selected P00412494 as the lectotype because it has Portères’
original label with his collection number, the specimen probably studied by Jacques-Félix, is in agreement with the original
description, is complete and well preserved.
Other included species. – Melastomastrum afzelii
(Hook.f.) A.Fern. & R.Fern., M. afzelii var. dialonkeanum
Jacq.-Fél., M. afzelii var. lecomteanum (Hutch. & Dalziel)
Jacq.-Fél., M. afzelii var. paucistellatum (Stapf) Jacq.-Fél.,
M. autranianum (Cogn.) A.Fern. & R.Fern., M. autranianum
var. latibracteatum (De Wild.) Jacq.-Fél., M. capitatum (Vahl)
A.Fern. & R.Fern., M. capitatum var. silvaticum Jacq.-Fél.,
M. segregatum (Benth.) A.Fern. & R.Fern., M. theifolium
(G.Don) A.Fern. & R.Fern.
8. Tristemma Juss., Gen. Pl.: 329. 1789 – Type: Tristemma mauritianum J.F.Gmel.
Morphological diagnosis. – Tristemma differs from
Melastomastrum, its closest African Melastomateae relative,
by the following combination of characters: fruit indehiscent,
calyx with 1 to several discrete rings of bristles, stamens isomorphic except in T. cornifolium.
Molecular diagnosis. – Tristemma also differs from the
other African Melastomateae genera by one nrITS nucleotide substitution: C insertion at position 71; four psbK-psbL
Version of Record
609
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
nucleotide substitutions: a T-to-C substitution at positions 47
and 130, deletion of AT at positions 67–68, CC/GC -to-CG
substitution at positions 78–79; and four accD-psaI nucleotide substitutions: T-to-A substitution at position 86, C-to-T
substitution at position 1045 and CA/TG-to-CC substitution
at positions 1174–1175 (summarised in Table 4).
CABI 2014 (last modified 5 Mar 2014). Heterotis rotundifolia (pink
lady). In: Invasive Species Compendium. http://www.cabi.org/isc/
datasheet/120226 (accessed 20 Jun 2016).
Clausing, G. & Renner, S.S. 2001. Molecular phylogenetics of
Melastomataceae and Memecylaceae: Implications for character
evolution. Amer. J. Bot. 88: 486–498.
https://doi.org/10.2307/2657114
Cogniaux, C.A. 1891. Melastomaceae. In: Candolle, A. de & Candolle,
C. de (eds.), Monographiae phanerogamarum, vol. 7. Parisiis
[Paris]: sumptibus G. Masson.
https://doi.org/10.5962/bhl.title.45961
Darriba, D., Taboada, G.L., Doallo, R. & Posada, D. 2012. jModelTest 2: More models, new heuristics and parallel computing.
Nature, Meth. 9: 772. https://doi.org/10.1038/nmeth.2109
De Wildeman, E. 1913. Études sur la Flore du Katanga. Ann. Mus.
Congo Belge. Bot., sér. 4, 1: 1–207.
Drummond, A.J. & Rambaut, A. 2007. BEAST: Bayesian evolutionary
analysis by sampling trees. B. M. C. Evol. Biol. 7: 214.
https://doi.org/10.1186/1471-2148-7-214
Drummond, A.J., Suchard, M.A., Xie, D. & Rambaut, A. 2012.
Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molec.
Biol. Evol. 29: 1969–1973. https://doi.org/10.1093/molbev/mss075
Duvigneaud, P. 1958. Etudes sur la vegetation Katanga et des ses
sols metalliferes. Réalisées sous l’égide du Fonds National de la
Recherche Scientifique, avec l’aide de l’Union Minière du HautKatanga. Communication No I. La végétation du Katanga et de
ses sols métallifèrs. Bull. Soc. Roy. Bot. Belgique 90(2): 127–286.
Engler, A. 1921. Die Pflanzenwelt Afrikas insbesondere seiner tropischen Gebiete: Grundzüge der Pflanzenverbreitung in Afrika
und die Charakterpflanzen Afrikas, vol. 3(2), Charakterpflanzen
Afrikas (insbesondere des tropischen): Die Familien der afrikanischen Pflanzenwelt und ihre Bedeutung in derselben, 2,
Die dikotyledonen Angiospermen Euphorbiaceae, Sapindales –
Umbelliflorae (Schluß). Die Vegetation der Erde 9. Leipzig:
Engelmann. https://doi.org/10.5962/bhl.title.50144
Exell, A.W., Good, R., Norman, C., Greves, S. & Moore, S. 1929. Mr.
John Gossweiler’s Portuguese West African plants (part 1). J. Bot.
67 (Suppl. 1): 177–203.
Farron, C. & Favarger, C. 1983–84. Contribution à la cytotaxonomie
des Melastomatacées africaines. Garcia de Orta, Sér. Bot. 6(1–2):
83–88.
Favarger, C. 1952. Recherches sur quelques Mélastomacées d’Afrique
occidentale. Ber. Schweiz. Bot. Ges. 62: 5–65.
Favarger, C. 1962. Nouvelles recherches cytologiques sur les Mélastomatacées. Ber. Schweiz. Bot. Ges. 72: 290–305.
Feissly, C. 1964. Sur l’ornementation du tube calicinal de quelques
Osbeckiées africaines. Bull. Soc. Neuchâtel. Sci. Nat. 87: 137–170.
Felsenstein, J. 1985. Confidence limits on phylogenies: An approach
using the bootstrap. Evolution 39: 783–791.
https://doi.org/10.2307/2408678
Fernandes, A. & Fernandes, R. 1954a. Sur la position systématique
de la section Pseudodissotis Cogn. du genre Osbeckia L. (Note
préliminaire). Bol. Soc. Brot., sér. 2, 28: 65–76.
Fernandes, A. & Fernandes, R. 1954b. Contribuição para o conhecimento das Melastomatáceas da Guiné Portuguesa. Garcia de
Orta 2: 273–285.
Fernandes, A. & Fernandes, R. 1954c. Contribution to the knowledge
of the Melastomataceae of Moçambique. Bol. Soc. Brot., sér. 2,
28: 205–214, 6 pl.
Fernandes, A. & Fernandes, R. 1956a. Revisão das Melastomatoideae
do East African Herbarium e do Southern Rhodesia Government
Herbarium. Mem. Soc. Brot. 11: 65–96.
Fernandes, A. & Fernandes, R. 1956b. Melastomataceae africanae
novae vel minus cognitae – III. Bol. Soc. Brot., sér. 2, 30: 167–186,
25 pl.
Fernandes, A. & Fernandes, R. 1962. O que é Osbeckia tubulosa Sm.?
Pp. 1–9 in: Estudos científicos oferecidos em homenagem ao Prof.
Other included species. – Tristemma akeassii Jacq.Fél., T. albiflorum (G.Don) Benth., T. camerunense Jacq.Fél., T. cornifolium (Benth.) Triana, T. coronatum Benth.,
T. demeusei De Wild., T. hirtum P.Beauv., T. involucratum
Benth., T. leiocalyx Cogn., T. littorale Benth., T. mauritianum J.F.Gmel., T. oreophilum Gilg, T. oreothamnos Mildbr.,
T. rubens A.Fern. & R.Fern., T. schliebenii Markgr., T. vestitum Jacq.-Fél.
ACKNOWLEDGEMENTS
We would like to thank the curators and personnel of the herbaria
BR, BRLU, C, EA, K, MO, NHN, and UPS for providing loans and for
assistance during visits. We are also grateful to J. Burrows (BNRH),
N. Barker and R. Clark (GRA), D. Goyder (K), R. Gereau (MO),
Olivier Maurin (UJ) and B. Wursten (BR) and to the herbaria BR,
BRLU, C, MJG, MO, NHN for additional DNA samples. We would
also like to thank F. Almeda, E. Bidault & P.B. Phillipson (http://
www.tropicos.org/ImageSearch.aspx), R. von Blittersdorff (http://
www.eastafricanplants.org), V. Droissart, D. Goyder and B. Wursten
for providing us with additional photographs. We thank J.W. Kadereit
for critically reading through the manuscript and providing valuable
comments on an earlier draft. We would like to thank Ricardo Kriebel
and an anonymous reviewer for their help in improving this article.
The field study was supported by the International Association for
Plant Taxonomy (IAPT), herbarium visits by Equal Opportunity travel
grant of Mainz University, Germany, and laboratory work by German
Science Foundation (DFG), project number KA1816/10-1. The first
author also acknowledges the German Academic Exchange Service
(DAAD) for a Doctoral scholarship and Idea Wild for assisting in the
purchase of equipment.
LITERATURE CITED
Andrews, F.W. 1950. The flowering plants of the Anglo-Egyptian
Sudan, vol. 1. Scotland: Buncle.
APG 2009. An update of the Angiosperm Phylogeny Group classification
for the orders and families of flowering plants: APG III. Bot. J. Linn.
Soc. 161: 105–121. https://doi.org/10.1111/j.1095-8339.2009.00996.x
Baillon, H. 1877 (“1880”). Histoire des plantes, vol. 7, Mélastomacées,
Cornacées, Ombellifères, Rubiacées, Valérianacées, Dipsacacées.
Paris: Librairie Hachette. https://doi.org/10.5962/bhl.title.40796
Bartling, F.G. 1830. Ordines naturales plantarum. Gottingae [Göt tingen]: sumtibus Dieterichianis.
http://bibdigital.rjb.csic.es/ing/Libro.php?Libro=5595
Bentham, G. 1849. Melastomaceae. Pp. 344–358 in: Hooker, W.J. (ed.),
Niger flora. London: Hippolyte Baillière.
https://doi.org/10.5962/bhl.title.594
Bentham, G. & Hooker, J.D. 1867. Genera plantarum, vol. 1(3).
Londini [London]: venit apud Reeve & Co.
https://doi.org/10.5962/bhl.title.747
610
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Doutor J. Carrington da Costa por ocasião do seu 70º aniversário
– Abril de 1961. Lisbon: Junta de Investigações do Ultramar.
Fernandes, A. & Fernandes, R. 1969. Melastomataceae africanae novae
vel minus cognitae – V. Bol. Soc. Brot., sér. 2, 43: 285–306, 17 pl.
Fernandes, R. & Fernandes, A. 1970. Melastomataceae. Pp. 114–165 in:
Exell, A.W., Fernandes, A. & Mendes, E.J. (eds.), Conspectus florae Angolensis, vol. 4. Lisbon: Junta de Investigações do Ultramar
e Instituto de Investigação Científica de Angola.
Fernandes, R. & Fernandes, A. 1978. Melastomataceae. Pp. 220–
276 in: Launert, E. (ed.), Flora Zambesiaca, vol. 4. Richmond:
Royal Botanic Gardens, Kew.
Fritsch, P.W., Almeda, F., Renner, S.S., Martins, A.B. & Cruz,
B.C. 2004. Phylogeny and circumscription of the near-endemic
Brazilian tribe Microlicieae (Melastomataceae). Amer. J. Bot. 91:
1105–1114. https://doi.org/10.3732/ajb.91.7.1105
Gilg, E. 1898. Monographieen afrikanischer Pflanzen-Familien und
-Gattungen, vol. 2, Melastomataceae. Leipzig: Engelmann.
https://doi.org/10.5962/bhl.title.53505
Gilg, E. 1901. Melastomataceae. Pp. 365–367 in: Engler, A. Berichte
über die botanischen Ergebnisse der Nyassa-See und KingaGebirgs-Expedition der Hermann- und Elise- geb. HeckmannWentzel-Stiftung. IV. Die von Goetze am Rukwa-See und NyassaSee sowie in den zwischen beiden Seen gelegenen Gebirgsländern,
insbesondere dem Kinga-Gebirge gesammelten Pflanzen, nebst
einigen Nachträgen (durch * bezeichnet) zu Bericht III. Bot. Jahrb.
Syst. 30: 239–445.
Goldenberg, R., Penneys, D.S., Almeda, F., Judd, W.S. & Michelangeli, F.A. 2008. Phylogeny of Miconia (Melastomataceae):
Patterns of stamen diversification in a megadiverse neotropical
genus. Int. J. Pl. Sci. 169: 963–979. https://doi.org/10.1086/589697
Grimm, D. & Almeda, F. 2013. Systematics, phylogeny and biogeography of Chaetolepis (Melastomataceae). J. Bot. Res. Inst. Texas
7: 217–263.
Guillemin, J.-A., Perrottet, G.S. & Richard, A. 1833 (“1830–1833”).
Florae Senegambiae tentamen. Parisiis [Paris]: apud Treuttel et
Würtz. https://doi.org/10.5962/bhl.title.595
Hooker, J.D. 1871. Melastomaceae. Pp. 439–464 in: Oliver, D. (ed.),
Flora of tropical Africa, vol. 2. London: L. Reeve.
https://doi.org/10.5962/bhl.title.42
Hutchinson, J. & Dalziel, J.M. 1927. Flora of west tropical Africa,
vol. l(1). London: The Crown Agents for the Colonies.
Hutchinson, J. & Dalziel, J.M. 1928. Tropical African plants: III. Bull.
Misc. Inform. Kew 1928: 211–229. https://doi.org/10.2307/4107692
Ionta, G.M., Judd, W.S., Williams, N.H. & Whitten, W.M. 2007.
Phylogenetic relationships in Rhexia (Melastomataceae): Evidence
from DNA sequence data and morphology. Int. J. Pl. Sci. 168:
1055–1066. https://doi.org/10.1086/518837
Jacques-Félix, H. 1932. Sur quelques Melastomacées nouvelles ou
peu connues de l’Afrique occidentale. Bull. Mus. Natl. Hist. Nat.,
sér. 2, 6: 678–687.
Jacques-Félix, H. 1939 (“1838”). Sur quelques Melastomacées africaines. Bull. Mus. Natl. Hist. Nat., sér. 2, 10: 630–642.
Jacques-Félix, H. 1953. Sur quelques Melastomaceae d’Afrique. Bull.
Inst. Franç. Afrique Noire 15: 972–1001.
Jacques-Félix, H. 1975 (“1974”). Le genre Melastomastrum Naudin
(Melastomataceae). Bull. Mus. Natl. Hist. Nat., Sér. 3, Bot. 17: 49–84.
Jacques-Félix, H. 1976. Le genre Tristemma Jussieu (Melastomataceae).
Bull. Mus. Natl. Hist. Nat., Sér. 3, Bot. 28: 137–207.
Jacques-Félix, H. 1981. Observations sur les caractères staminaux et
la classification des Osbeckieae (Melastomataceae) capsulaires
africaines. Adansonia, sér. 2, 20: 405–429.
Jacques-Félix, H. 1983. Mélastomatacées. In: Satabié, B. & Leroy, J.-F.
(eds.), Flore du Cameroun, vol. 24. Yaounde: Délégation Générale
à la Recherche Scientifique et Technique.
Jacques-Félix, H. 1986. Description d’un Tristemma (Melastomataceae)
nouveau du Gabon. Bull. Mus. Natl. Hist. Nat., B, Adansonia 8:
191–193.
Jacques-Félix, H. 1995. Histoire des Melastomataceae d’Afrique. Bull.
Mus. Natl. Hist. Nat., B, Adansonia 16: 235–311.
Judd, W.S., Ionta, G.M., Skean, J.D., Campbell, K., Jr. & Penneys,
D.S. 2011. Noteworthy collections of Heterotis rotundifolia in
Jamaica and Dominica. Castanea 76: 311–312.
https://doi.org/10.2179/11-011.1
Keay, R.W.J. 1954. Flora of west tropical Africa, ed. 2, vol. 1(1). London:
Crown Agents for Oversea Governments and Administrations.
Krasser F. 1893. Melastomataceae. Pp. 130–199 in: Engler, A. &
Prantl, K. (eds.), Die natürlichen Pflanzenfamilien III(7). Leipzig:
Engelmann.
Kriebel, R. 2016. Phylogenetic placement of the monotypic genus
Schwackaea (Melastomeae: Melastomataceae) and the evolution
of its unique fruit. Int. J. Pl. Sci. 177: 440–448.
https://doi.org/10.1086/685689
Kriebel, R., Michelangeli, F.A. & Kelly, L.M. 2015. Discovery of
unusual anatomical and continuous characters in the evolutionary history of Conostegia (Miconieae: Melastomataceae). Molec.
Phylogen. Evol. 82: 289–313.
https://doi.org/10.1016/j.ympev.2014.09.021
Leistner, O.A. 2005. Seed plants of southern tropical Africa: Families
and genera. Pretoria: South African Botanical Diversity Network.
Maddison, W.P. & Maddison, D.R. 2015. Mesquite: A modular system
for evolutionary analysis, version 3.04 http://mesquiteproject.org
(accessed 15 Dec 2015).
Majure, L.C., Neubig, K.M., Skean, J.D., Bécquer, E.R. & Judd,
W.S. 2015. Evolution of the sandpaper clade (Miconieae, Melastomataceae). Int. J. Pl. Sci. 176: 607–626.
https://doi.org/10.1086/682148
Meyer, K. 2001. Revision of the Southeast Asian genus Melastoma
(Melastomataceae). Blumea 46: 351–398.
Michelangeli, F.A., Penneys, D.S., Giza, J., Soltis, D., Hils, M.H. &
Skean, J.D. 2004. A preliminary phylogeny of the tribe Miconieae
(Melastomataceae) based on nrITS sequence data and its implications on inflorescence position. Taxon 53: 279–290.
https://doi.org/10.2307/4135608
Michelangeli, F.A., Judd, W.S., Penneys, D.S., Skean, J.D.,
Bécquer-Granados, E.R., Goldenberg, R. & Martin, C.V. 2008.
Multiple events of dispersal and radiation of the tribe Miconieae
(Melastomataceae) in the Caribbean. Bot. Rev. (Lancaster) 74:
53–77. https://doi.org/10.1007/s12229-008-9004-x
Michelangeli, F.A., Guimaraes, P.J.F., Penneys, D.S., Almeda, F. &
Kriebel, R. 2013. Phylogenetic relationships and distribution of
New World Melastomeae (Melastomataceae). Bot. J. Linn. Soc. 171:
38–60. https://doi.org/10.1111/j.1095-8339.2012.01295.x
Naudin, C. 1850. Melastomacearum monographicae descriptionis. Ann.
Sci. Nat., Bot., sér. 3, 13: 273–303.
Naudin, C. 1853 (“1849–1853”). Melastomacearum monographicae
descriptionis. Parisiis [Paris]: apud Victor Masson. [Consolidated
reprint of a series of papers originally published in Ann. Sci. Nat.,
Bot., sér. 3]. https://doi.org/10.5962/bhl.title.12248
Penneys, D.S. & Judd, W.S. 2013. Combined molecular and morphological phylogenetic analyses of the Blakeeae (Melastomataceae).
Int. J. Pl. Sci. 174: 802–817. https://doi.org/10.1086/670011
Perrier de la Bâthie, H. 1932. Les Mélastomacées de Madagascar.
Mém. Acad. Malgache 12: 1–292, 10 pl.
Perrier de la Bâthie, H. 1951. Flore de Madagascar et des Comores
(plantes vasculaires), vol. 153, Mélastomatacées. Paris: FirminDidot. https://doi.org/10.5962/bhl.title.6600
Rafinesque, C.S. 1838. Sylva telluriana. Philadelphia: printed for the
author and publisher. https://doi.org/10.5962/bhl.title.248
Rambaut, A. & Drummond, A.J. 2009. Tracer, version 1.5. http://tree.
bio.ed.ac.uk/software/tracer/ (accessed 01 Dec 2009).
Reginato, M. & Michelangeli, F.A. 2016. Untangling the phylogeny
of Leandra s.str. (Melastomataceae, Miconieae). Molec. Phylogen.
Evol. 96: 17–32. https://doi.org/10.1016/j.ympev.2015.11.015
Reginato, M., Michelangeli, F.A. & Goldenberg, R. 2010. Phylogeny
Version of Record
611
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
of Pleiochiton (Melastomataceae, Miconieae): Total evidence. Bot.
J. Linn. Soc. 162: 423–434.
https://doi.org/10.1111/j.1095-8339.2009.01022.x
Rendle, A.B., Baker, E.G., Moore, S. & Gepp, A. 1911. A contribution to our knowledge of the flora of Gazaland: Being an account
of collections made by C. F. M. Swynnerton, Esq., F.L.S. J. Linn.
Soc. London, Bot. 40: 1–245.
Renner, S.S. 1993. Phylogeny and classification of the Melastomataceae
and Memecylaceae. Nordic J. Bot. 13: 519–540.
https://doi.org/10.1111/j.1756-1051.1993.tb00096.x
Renner, S.S. & Meyer, K. 2001. Melastomeae come full circle: Biogeographic reconstruction and molecular clock dating. Evolution
55: 1315–1324. https://doi.org/10.1111/j.0014-3820.2001.tb00654.x
Renner, S.S., Triebel, D., Almeda, F., Stone, R.D., Ulloa, C.,
Michaelangeli, F.A., Goldenberg, R. & Mendoza, H. (eds.)
2007–. MEL names – A database with names of Melastomataceae.
http://www.melastomataceae.net/MELnames/ (accessed Feb 2014).
Reveal, J.L. 2011–. Indices Nominum Supragenericorum Plantarum
Vascularium. http://www.plantsystematics.org/reveal/pbio/fam/
allspgnames.html)
Small, R.L., Ryburn, J.A., Cronn, R.C., Seelanan, T. & Wendel,
J.F. 1998. The tortoise and the hare: Choosing between noncoding
plastome and nuclear Adh sequences for phylogeny reconstruction
in a recently diverged plant group. Amer. J. Bot. 85: 1301–1315.
https://doi.org/10.2307/2446640
Smith, J.E. 1813. Osbeckia. In: Rees, A., The Cyclopaedia, vol. 25.
London: printed for Longman, etc.
https://doi.org/10.5962/bhl.title.59683
Stamatakis, A. 2014. RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30:
1312–1313. https://doi.org/10.1093/bioinformatics/btu033
Stevens, P.F. 2001–. Angiosperm Phylogeny Website, version 12, July
2012 [and more-or-less continuously updated since]. http://www.
mobot.org/MOBOT/research/APweb/ (accessed 27 Dec 2015).
Stone, R.D. 2014. The species-rich, paleotropical genus Memecylon
(Melastomataceae): Molecular phylogenetics and revised infrageneric classification of the African species. Taxon 63: 539–561.
https://doi.org/10.12705/633.10
Stone, R.D. & Andreasen, K. 2010. The Afro-Madagascan genus
Warneckea (Melastomataceae): Molecular systematics and revised
infrageneric classification. Taxon 59: 83–92.
Stone, R.D. & Veranso-Libalah, M.C. 2017. (2527) Proposal to reject
the name Kadalia (Melastomataceae: Melastomateae). Taxon 66:
758–759. https://doi.org/10.12705/663.26
Swofford, D.L. 2002. PAUP*: Phylogenetic analysis using parsimony
(*and other methods), version 4. Sunderland, Massachusetts:
Sinauer. http://paup.csit.fsu.edu/ (accessed 10 Dec 2014).
Triana, J.J. 1865. Dispositio Melastomacearum. Bull. Congr. Int. Bot.
Hort. Amsterdam 1865: 457–461.
Triana, J.J. 1872 (“1871”). Les Mélastomacées. Trans. Linn. Soc.
London 28(1): 1–188.
https://doi.org/10.1111/j.1096-3642.1871.tb00222.x
Van Tieghem, P. 1891. Classification anatomique des Mélastomacées.
Bull. Soc. Bot. France 38: 114–124.
https://doi.org/10.1080/00378941.1891.10828542
Van Vliet, G.J.C.M., Koek-Noorman, J. & Ter Welle, B.J.H.
1981. Wood anatomy, classification and phylogeny of the
Melastomataceae. Blumea 27: 463–473.
Whiffin, T. & Tomb, A.S. 1972. The systematic significance of seed
morphology in the neotropical capsular-fruited Melastomataceae.
Amer. J. Bot. 59: 411–422. https://doi.org/10.2307/2441552
Wickens, G.E. 1975. Melastomataceae. In: Polhill, R.M. (ed.), Flora
of tropical East Africa. London: Crown Agents for Oversea
Governments and Administrations.
Appendix 1. Voucher information (Species name, Synonym, Country, Region, Collector number, herbarium code, herbarium barcode or sheet number and
Lab code if available). GenBank accessions numbers are given for all nrITS, psbK-psbL, accD-psaI sequences included in this study. Missing sequence data
is indicated by a dash (–). DNA sequences newly generated for this study are marked with an asterisk (*).
OUTGROUP: Aciotis indecora (Bonpl.) Triana, French Guiana, C.V. Martin 411 (NY), JQ730039, JQ730250, JQ730459; Aciotis paludosa (Mart. ex DC.)
Triana, Brazil, P. Guimarães 317 (RB), JQ730040, JQ730251, JQ730460; Aciotis rubricaulis (Mart. ex DC.) Triana, Brazil, R. Goldenberg 850 (NY),
JQ730042, JQ730253, JQ730462; Acisanthera hedyotidea (C.Presl.) Triana, Guyana, K.J. Wurdack 4145 (NY), JQ730044, JQ730255, JQ730464;
Acisanthera quadrata Pers., Venezuela, F.A. Michelangeli 826 (BH), JQ730045, JQ730256, JQ730465; Brachyotum incrassatum E.Cotton, Ecuador, J.L.
Clark 8896 (NY), JQ730056, JQ730267, JQ730475; Brachyotum ledifolium (Desr.) Triana, Ecuador, D.S. Penneys 1554 (FLAS), JQ730057, JQ730268,
JQ730476; Cambessedesia espora (A.St.Hil. ex Bonpl.) DC., Brazil, P. Guimarães 397 (RB), JQ730062, JQ730273, JQ730481; Cambessedesia hilariana
(A.St.Hil. ex Bonpl.) DC., Brazil, P. Guimarães 405 (RB), JQ730063, JQ730274, JQ730482; Chaetolepis cufodontisii Standl., Costa Rica, F.A. Michelangeli
1160 (NY), JQ730067, JQ730278, JQ730485; Chaetolepis microphylla (Bonpl.) Miq., Brazil, F.A. Michelangeli 1268 (NY), JQ730068, JQ730279, JQ730486;
Comolia microphylla Benth., Guyana, K.M. Redden 454 (NY), JQ730070, JQ730281, JQ730488; Dinophora spenneroides Benth., Cameroon, SW,
Mundemba, M.C. Veranso 107 (MJG), 20950, DINOSPEN21, KX889227*, KY284746*, –; Ernestia glandulosa Gleason, French Guiana, C.V. Martin 471
(NY), JQ730080, JQ730293, JQ730499; Graffenrieda latifolia (Naudin) Triana, Venezuela, F.A. Michelangeli 794 (BH), AY460450, JQ730297, JQ730503;
Graffenrieda moritziana Triana, Venezuela, F.A. Michelangeli 832 (BH), AY460451, JQ730298, JQ730504; Heterocentron elegans (Schltdl.) Kuntze,
U.S.A. (collected in cultivation outside native range), cultivated in NYBG (NY), JQ730085, JQ730299, JQ730506; Lavoisiera imbricata (Thunb.) DC.,
Brazil, R. Goldenberg 820 (NY), JQ730091, JQ730304, JQ730511; Lavoisiera mucorifera Mart. ex Schrank ex DC., Brazil, P. Guimarães 345 (RB),
JQ730092, JQ730305, JQ730512; Leandra cf. kleinii Brade, R. Goldenberg 728 (UPCB), EU055693, GQ139258, GQ139342; Meriania longifolia (Naudin)
Cogn., Venezuela, F.A. Michelangeli 610 (BH), AY460454, JQ730316, JQ730526; Miconia dodecandra (Desr.) Cogn., Dominican Republic, W.S. Judd
8093 (FLAS), AY460506, JQ730317, JQ730527; Miconia tomentosa (Rich.) D.Don ex DC., Brazil, F.A. Michelangeli 1344 (NY), EF418905, JQ730318,
JQ730528; Microlepis oleifolia (DC.) Triana, Brazil, R. Goldenberg 1036 (NY), JQ730107, JQ730319, JQ730529; Monochaetum discolor H.Karst. ex
Triana, Venezuela, I. Capote 811 (NY), JQ730109, JQ730321, JQ730531; Monochaetum meridense (Klotzsch ex H.Karst.) Naudin, Venezuela, I. Capote
862 (NY), JQ730111, JQ730323, JQ730533; Monochaetum volcanicum Cogn., Costa Rica, F.A. Michelangeli 718 (BH), JQ730114, JQ730326, JQ730536;
Physeterostemon tomasii Amorim, Michelangeli & Goldenb. Brazil, A. Amorim 5054 (NY), JQ730121, JQ730332, JQ730542; Pilocosta nana (Standl.)
Almeda & Whiffin, Costa Rica, R.C. Moran 6928 (NY), JQ730123, JQ730334, JQ730544; Pilocosta nubicola Almeda, Costa Rica, D.S Penneys 1775
(FLAS), JQ730124, JQ730335, JQ730545; Pterogastra divaricata (Bonpl.) Naudin, Venezuela, F.A. Michelangeli 540 (BH), JQ730126, JQ730337, JQ730547;
Pterogastra minor Naudin, Brazil, G.A. Romero 1682 (CAS), JQ730127, JQ730338, JQ730548; Pterolepis glomerata (Rottb.) Miq., French Guiana, C.V.
Martin 419 (NY), JQ730129, JQ730340, JQ730550; Pterolepis repanda (DC.) Triana, Brazil, F. Almeda 7731 (CAS), JQ730131, JQ730342, JQ730552;
Rhexia aristosa Britton, U.S.A., R.F.C. Naczi 12065 (NY), JQ730134, JQ730345, JQ730555; Rhexia virginica L., U.S.A., F.A. Michelangeli 1448 (NY),
JQ730136, JQ730346, JQ730557; Rousseauxia andringitrensis (H.Perrier) Jacq.-Fél., Madagascar, F. Almeda 9390 (CAS), JQ730139, JQ730350, JQ730561;
Rousseauxia minimifolia (Jum. & H.Perrier) Jacq.-Fél., Madagascar, F. Almeda 8704 (CAS), JQ730140, JQ730351, JQ730562; Svitramia pulchra Cham.,
Brazil, P. Guimarães 388 (RB), JQ730145, JQ730356, JQ730567; Tibouchina arborea (Gardn.) Cogn., Brazil, P. Guimarães 324 (RB), JQ730152, JQ730363,
JQ730575; Tibouchina arenaria Cogn., Brazil, P. Guimarães 352 (RB), JQ730153, JQ730364, JQ730576; Tibouchina candolleana (DC.) Cogn., Brazil,
J. Lima 711 (NY), JQ730164, JQ730375, JQ730585; Tibouchina cardinalis (Bonpl.) Cogn., Brazil, P. Guimarães 407 (RB), JQ730165, JQ730376, JQ730586;
612
Version of Record
TAXON 66 (3) • June 2017: 584–614
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
Appendix 1. Continued.
Tibouchina citrina (Naudin) Cogn., Bolivia, M. Nee 55308 (NY), JQ730171, JQ730380, JQ730592; Tibouchina clavata (Pers.) Wurdack, Brazil, A.K.A.
Santos 696 (UFB), JQ730172, JQ730381, JQ730593; Tibouchina confertiflora (Naudin) Cogn., Bolivia, M. Nee 55377 (NY), JQ730175, JQ730384,
JQ730595; Tibouchina cristata Brade, Brazil, J.F.A. Baumgratz 819 (R), JQ730177, JQ730386, JQ730597; Tibouchina cryptadena Gleason, Venezuela,
F.A. Michelangeli 708 (NY), JQ730178, JQ730387, JQ730598; Tibouchina dubia (Cham.) Cogn., Brazil, I.G. Varassin 101 (UFCB), JQ730179, JQ730388,
JQ730599; Tibouchina gracilis (Bonpl.) Cogn., Brazil, P. Guimarães 336 (RB), JQ730190, JQ730398, JQ730610; Tibouchina granulosa (Desr.) Cogn.,
Brazil, P. Guimarães 378 (RB), JQ730191, JQ730399, JQ730611; Tibouchina pulchra (Cham.) Cogn., Brazil, J.F.A. Baumgratz 1068 (R), JQ730222,
JQ730430, JQ730642; Tibouchina radula Markgr., Brazil, R. Goldenberg 1281 (NY), JQ730223, JQ730431, JQ730643; Trembleya parviflora Cogn.,
Brazil, R. Goldenberg 824 (NY), JQ730242, JQ730451, JQ730663. — INGROUP: Amphorocalyx multiflorus Baker, Madagascar, F. Almeda 8669 (CAS),
JQ730046, JQ730257, JQ730466; Amphorocalyx rupestris H.Perrier, Madagascar, F. Almeda 8723 (CAS), JQ730047, JQ730258, JQ730467; Anaheterotis
pobeguinii (Hutch. & Dalziel) Veranso-Libalah & G.Kadereit (Heterotis pobeguinii), Guinea, A.S. Goman 259 (BR), BR0000017294690, HETEPOBE132,
KX889284*, KY284809*, KY284710*; Antherotoma angustifolia (A.Fern. & R.Fern.) Jacq.-Fél., Zambia, M.G. Bingham 9607 (WAG), WAG.1091744,
ANTHANGU140, KX889210*, KY284727*, KY284649*; Antherotoma debilis (Sond.) Jacq.-Fél., Angola, Cuando Cubango, D. Goyder 8094 (K),
ANTHDEBI185, –, KY284728*, KY284650*; Antherotoma debilis (Sond.) Jacq.-Fél., Mozambique, Bilene, J.E. Burrows 14526 (BNRH), ANTHDEBI59,
–, KY284729*, KY284651*; Antherotoma densiflora (Gilg) Jacq.-Fél., Tanzania, S. Bidgood & al. 2728 (WAG), WAG.1091979, ANTHDENS143,
KX889211*, KY284730*, –; Antherotoma gracilis (Cogn.) Jacq.-Fél., Angola, Bamps & al. 4165 (BR), BR0000018229899, ANTHGRAC43, KX889212*,
KY284731*, –; Antherotoma irvingiana (Hook.f.) Jacq.-Fél., Cameroon, NW, Sabga, M.C. Veranso 226 (MJG), 20985, ANTHIRVI33, KX889213*,
KY284732*, KY284652*; Antherotoma irvingiana (Hook.f.) Jacq.-Fél., Cameroon, NW, Mbiame, M.C. Veranso 199 (MJG), 20954, ANTHIRVI9,
KX889214*, KY284733*, KY284653*; Antherotoma naudinii Hook.f., Zimbabwe, Chimanimanis, B. Wursten 2271 (BR), BR-SIL-SP-8448, ANTHNAUD3,
KX889215*, KY284734*, KY284654*; Antherotoma naudinii Hook.f., Cameroon, NW, Mbiame, M.C. Veranso 200 (MJG), 20955, ANTHNAUD38,
KX889216*, KY284735*, KY284655*; Antherotoma naudinii Hook.f., Madagascar, F. Almeda 8624 (CAS), JQ730048, JQ730259, –; Antherotoma phaeotricha (Hochst.) Jacq.-Fél. var. phaeotricha, Tanzania, Kagera, Bukoba, L. Festo 708 (MO), MO-2721884, ANTHPHAE119, KX889217*, KY284736*,
KY284656*; Antherotoma phaeotricha (Hochst.) Jacq.-Fél. var. phaeotricha, Tanzania, Kagera, Bukoba, D. Sitoni & al. 1032 (MO), MO-2721889,
ANTHPHAE120, KX889218*, KY284737*, KY284657*; Antherotoma senegambiensis (Guill. & Perr.) Jacq.-Fél., Tanzania, Kagera, Bukoba, Festo 1184
(MO), MO-2721879, ANTHSENE113, KX889219*, KY284738*, KY284658*; Antherotoma senegambiensis (Guill. & Perr.) Jacq.-Fél., Zimbabwe, B.
Wursten 2269 (BR), BR-SIL-SP-8328, ANTHSENE60, KX889220*, KY284739*, –; Argyrella amplexicaulis (Jacq.-Fél.) Veranso-Libalah & G.Kadereit
(Heterotis amplexicaulis), Guinea, Nzérékoré, G. Traore 29 (BR), BR0000016915527, HETEAMPL54, KX889273*, KY284798*, KY284702*; Argyrella
bambutorum (Gilg & Ledermann ex Engl.) Veranso-Libalah & G.Kadereit (Heterotis angolensis var. bambutorum), Cameroon, NW, Mbiame, M.C.
Veranso 202 (MJG), 20956, HETEANGO39, KX889274*, KY284799*, KY284703*; Argyrella bambutorum (Gilg & Ledermann ex Engl.) Veranso-Libalah
& G.Kadereit (Heterotis angolensis var. bambutorum), Cameroon, NW, Bui, M.C. Veranso 188 (MJG), 20984, HETEANGO8, KX889275*, KY284800*,
KY284704*; Argyrella canescens (E.Mey. ex Graham) Harv. (Heterotis canescens), Zimbabwe, Chimanimani, B. Wursten 2880 (BR), BR-SIL-SP-8409,
HETECANE1, KX889279*, KY284804*, KY284706*; Argyrella canescens (E.Mey. ex Graham) Harv. (Heterotis canescens), Cameroon, Adamawa,
Ngaoundere, M.C. Veranso 154 (MJG), 20962, HETECANE36, KX889280*, KY284805*, KY284707*; Argyrella sp., Tanzania, S. Bidgood, D. Sitoni &
al. 3935 (C), HETERICH94, KX889285*, KY284810*, KY284711*; Dichaetanthera africana (Hook.f.) Jacq.-Fél., Gabon, Wieringa & al. 3817 (BR),
DICHAFRI170, KX889221*, KY284740*, –; Dichaetanthera africana Jacq.-Fél., Gabon, S. Smith 1885 (US), JQ730074, JQ730285, JQ730492;
Dichaetanthera arborea Baker, Madagascar, Toamasina, Atsinanana, A. Razanatsima 823 (MO), MO-2303613, DICHARBO172, KX889222*, KY284741*,
KY284659*; Dichaetanthera arborea Baker, Madagascar, G. Clausing & al. 281 (MJG), DICHARBO176, KX889223*, KY284742*, –; Dichaetanthera
articulata Endl., Madagascar, Toamasina, D. Ravelonarivo & F. Edmond 4320 (MO), MO-2698044, DICHARTI173, KX889224*, KY284743*, –;
Dichaetanthera erici-rosenii (R.E.Fr)A.Fern. & R.Fern., Tanzania, Rukwa, Sumbawanga, H.H. Schmidt & al. 1190 (MO), 4627710, DICHERIC156,
KX889225*, KY284744*, –; Dichaetanthera oblongifolia Baker, Madagascar, F. Almeda 7926a (CAS), JQ730075, JQ730286, JQ730493; Dichaetanthera
strigosa (Cogn.) Jacq.-Fél., Gabon, Haut-Ogooue, A.F. Bradley & al. 1016 (MO), MO-275871, DICHSTRI174, KX889226*, KY284745*, –; Dionycha
bojeri Naudin, Madagascar, F. Almeda 8626 (CAS), JQ730076, JQ730287, JQ730494; Dissotidendron caloneurum (Gilg ex Engl.)Veranso-Libalah &
G.Kadereit (Dissotis caloneura), Tanzania, S. Bidgood & al. 4484 (C), DISSCALO86, KX889230*, KY284749*, KY284662*; Dissotidendron caloneurum
(Gilg ex Engl.) Veranso-Libalah & G.Kadereit var. caloneurum (Dissotis caloneura var. caloneura), Burundi, S. Ntore 119 (BR), BR0000017295017,
DISSCALOC87, KX889232*, KY284751*, –; Dissotidendron caloneurum var. pilosum (A.Fern. & R.Fern.) Veranso-Libalah & G.Kadereit (Dissotis
caloneura var. pilosa), N. Rhodesia (Zimbabwe), Luanshya, D.B. Fanshawe 3123 (BR), BR0000013426354, DISSCALO93, KX889231*, KY284750*,
KY284663*; Dissotidendron cordatum (P.A.Duvign) Veranso-Libalah & G.Kadereit, (Dissotis derriksiana), D.R. Congo, F. Malaisse, E. Semereab & G.
Handjila 533 (BR), BR0000017293129, DISSDERR163, KX889239*, KY284759*, –; Dissotidendron cordatum (P.A.Duvign) Veranso-Libalah & G.Kadereit
(Dissotis derriksiana), D.R. Congo, Fungurume, I. Parmentier, Kisimba & Sahato 4433 (BR), BR000000972923, DISSDERR81, KX889240*, KY284760*,
–; Dissotidendron melleri (Hook.f.) Veranso-Libalah & G.Kadereit (Dissotis melleri), Tanzania, Iringa, Njombe, O.A. Kibure 745 (MO), MO-2721895,
DISSMELL106, KX889248*, KY284769*, –; Dissotidendron polyanthum Gilg (Dissotis polyantha), Tanzania, L. Festo, M.A. Mwangoka & R. Mbalo
1901 (C), DISSPOLY89, KX889254*, KY284775*, –; Dissotis chevalieri Gilg, Cameroon, Adamawa, Lac Tison, M.C. Veranso 130 (MJG), 20982,
DISSPERK28, KX889252*, KY284773*, KY284679*; Dissotis congolensis (Cogn. ex Büttner) Jacq.-Fél., Cameroon, Adamawa, Tibati, M.C. Veranso 151
(MJG), 20973, DISSCONG18, –, KY284753*, KY284664*; Dissotis congolensis (Cogn. ex Büttner) Jacq.-Fél., Gabon, Haut-Ogooue, G.M. Walters 1004
(MO), MO-228453, DISSCONG41, KX889234*, KY284754*, KY284665*; Dissotis congolensis (Cogn. ex Büttner) Jacq.-Fél., Gabon, L.J.T. White 693
(MO), MO-2721893, DISSCONG58, KX889235*, KY284755*, KY284666*; Dissotis crenulata Cogn., Angola, B. Goyder, Baker & Clark 7476 (GRA),
DISSCREN165, KX889236*, KY284756*, KY284667*; Dissotis cryptantha Baker, Tanzania, Rukwa, Mpanda, W.R.Q. Luke & al. 16225 (MO), MO2721885, DISSCRYP98, KX889237*, KY284757*, KY284668*; Dissotis denticulata A.Fern. & R.Fern., Zambia, D.K. Harder & al. 2249 (WAG),
WAG.1091980, DISSDENT122, KX889238*, KY284758*, –; Dissotis facipila Gilg, Zambia, Lisombo, J. Loveridge 859 (BR), BR0000013426262,
DISSFALC127, KX889241*, KY284761*, KY284669*; Dissotis gilgiana De Wild., D.R. Congo, R. Desenfans 5537 (BRLU), DISSCFGILG161, KX889233*,
KY284752*, –; Dissotis gilgiana De Wild., D.R. Congo, M. Schaijes 2005 (BR), BR0000017295055, DISSGILG52, KX889242*, KY284762*, –; Dissotis
grandiflora Benth., Guinea, Phillipson 6387 (MO), MO-2721314 & MO-2721315, DISSGRAN105, KX889243*, KY284763*, KY284670*; Dissotis longisetosa Gilg & Ledermann ex Engl., Cameroon, NW, Mbiame, M.C. Veranso 203 (MJG), 20977, DISSLONG11, KX889245*, KY284766*, KY284673*;
Dissotis longisetosa Gilg & Ledermann ex Engl., Cameroon, NW, Mbiame, M.C. Veranso 205 (MJG), 20958, DISSLONG23, KX889246*, KY284767*,
KY284674*; Dissotis longisetosa Gilg & Ledermann ex Engl., Cameroon, NW, Bui, M.C. Veranso 178 (MJG), 20969, DISSLONG25, KX889247*,
KY284768*, KY284675*; Dissotis perkinsiae, Cameroon, NW, Bui, M.C. Veranso 179 (MJG), 20968, DISSPERK7, KX889253*, KY284774*, KY284680*;
Dissotis perkinsiae Gilg, Cameroon, NW, Mbiame, M.C. Veranso 196 (MJG), 20952, DISSSCHW160, KX889261*, KY284783*, KY284688*; Dissotis
princeps (Bonpl.) Triana, Cameroon, Adamawa, Ngaoundere, M.C. Veranso 156 (MJG), 20961, DISSPRIN20, KX889255*, KY284776*, KY284681*;
Dissotis princeps (Bonpl.) Triana, Cameroon, NW, Santa, M.C. Veranso 231 (MJG), 20992, DISSPRIN27, KX889256*, KY284777*, KY284682*; Dissotis
princeps (Bonpl.) Triana, Cameroon, NW, Bui, M.C. Veranso 181 (MJG), 20953, DISSPRIN42, KX889257*, KY284778*, KY284683*; Dissotis princeps
(Bonpl.) Triana, Zimbabwe, Chimanimanis, B. Wursten 2256 (BR), BR-SIL-SP-8385, DISSPRIN6, KX889258*, KY284779*, KY284684*; Dissotis princeps
(Bonpl.) Triana, Mozambique, O. Maurin 3393 (UJ), UJ09328, DISSPRIN61, KX889259*, KY284780*, KY284685*; Dissotis pulchra A.Fern & R.Fern.,
Zimbabwe, Chimanimanis, B. Wursten 2277 (BR), BR-SIL-SP-8445, DISSPULC2, –, KY284781*, KY284686*; Dissotis romiana De Wild., D.R. Congo,
Version of Record
613
Veranso-Libalah & al. • African Melastomateae (Melastomataceae)
TAXON 66 (3) • June 2017: 584–614
Appendix 1. Continued.
Oso de Nzilo, M. Schaijes 2348 (BR), BR0000013426477, DISSROMI45, KX889260*, KY284782*, KY284687*; Dissotis sizenandii Cogn., D.R. Congo,
S. Ntore 232 (BR), BR0000016153400, DISSSIZE137, KX889262*, KY284784*, KY284689*; Dissotis thollonii Cogn. ex Büttner var. elliotii, Cameroon,
NW, Mbiame, M.C. Veranso 208 (MJG), 20967, DISSTHOL12, KX889263*, KY284785*, KY284690*; Dissotis thollonii Cogn. ex Büttner var. elliotii,
Cameroon, Adamawa, Tibati, M.C. Veranso 144 (MJG), 20959, DISSTHOL24, KX889264*, KY284786*, KY284691*; Dissotis thollonii Cogn. ex Büttner
var. elliotii, Cameroon, Adamawa, Tibati, M.C. Veranso 145 (MJG), 20960, DISSTHOL26, KX889265*, KY284787*, KY284692*; Dissotis thollonii Cogn.
ex Büttner var. thollonii, Congo, L. Devred 509 (BR), BR0000016908123, DISSTHOLT44, KX889266*, KY284788*, KY284693*; Dissotis trothae Gilg,
Tanzania, Rukwa, Nkasi, General collector - SHCP 322 (MO), MO-2454118, DISSTROT102, KX889267*, KY284789*, –; Dissotis trothae Gilg, Tanzania,
Iringa, Makete, J. Lovett 4716 (MO), MO-2721846, DISSTROT117, KX889268*, KY284790*, KY284694*; Dissotis trothae Gilg, Tanzania, Rukwa,
Mpanda, Luke 16260 (MO), MO-2721886, DISSTROT118, KX889269*, KY284791*, KY284695*; Dissotis trothae Gilg, Burundi, S. Ntore 142 (BR),
BR0000016993273, DISSTROT90, –, KY284793*, KY284697*; Dissotis tubulosa (Sm.) Triana, Cameroon, NW, Donga-Mantung, Akweto, M.C. Veranso
221 (MJG), 20957, DISSTUBU22, KX889270*, KY284794*, KY284698*; Dissotis tubulosa (Sm.) Triana, Cameroon, NW, Bambili, M.C. Veranso 217
(MJG), 20979, DISSTUBU31, KX889271*, KY284795*, KY284699*; Dissotis weltwitschii Cogn., Angola, Moxico Province, D. Goyder 8240 (K),
DISSWELT184, –, KY284796*, KY284700*; Disstotis trothae Gilg, Burundi, S. Ntore 224 (BR), BR0000016993280, DISSTROT84, –, KY284792*,
KY284696*; Dupineta brazzae (Cogn.) Veranso-Libalah & G.Kadereit (Dissotis brazzae), Cameroon, NW, Santa, M.C. Veranso 234 (MJG), 20975,
DISSBRAZ19, KX889228*, KY284747*, KY284660*; Dupineta brazzae (Cogn.) Veranso-Libalah & G.Kadereit (Dissotis brazzae), Cameroon, NW,
Donga-Mantung, Akweto, M.C. Veranso 215 (MJG), 20981, DISSBRAZ29, KX889229*, KY284748*, KY284661*; Dupineta hensii (Cogn.) VeransoLibalah & G.Kadereit (Dissotis hensii), D.R. Congo, Itindi, E. Boyekoli 110 (BR), BR-SIL-SP-3226, DISSHENS57, KX889244*, KY284764*, KY284671*;
Dupineta hensii (Cogn.) Veranso-Libalah & G.Kadereit (Dissotis hensii), D.R. Congo, Bomane, Aruwimi river, E. Boyekoli 595 (BR), BR-SIL-SP-3871,
DISSHENS62, –, KY284765*, KY284672*; Dupineta multiflora (Sm.) Veranso-Libalah & G.Kadereit (Dissotis multiflora), Gabon, Ogoouè-Ivindo, L.
White 481 (MO), 6454998, DISSMULT108, KX889249*, KY284770*, KY284676*; Dupineta multiflora (Sm.) Veranso-Libalah & G.Kadereit (Dissotis
multiflora), Cameroon, Littoral, M.C. Veranso 174 (MJG), 20951, DISSMULT30, KX889250*, KY284771*, KY284677*; Dupineta multiflora (Sm.)
Veranso-Libalah & G.Kadereit (Dissotis multiflora), Cameroon, SW, Mundemba, M.C. Veranso 104 (MJG), 20964, DISSMULT32, KX889251*, KY284772*,
KY284678*; Dupineta multiflora (Sm.) Raf. (Dissotis multiflora), Gabon, S. Smith 1886 (US), JQ730077, JQ730288, JQ730495; Guyonia antennina
(Benth.) Veranso-Libalah & R.D.Stone (Heterotis antennina), Guinea, Nimba Botanic Team JR1031 (WAG), WAG.1092926, HETEANTE148, KX889276*,
KY284801*, KY284705*; Guyonia arenaria (Jacq.-Fél.) Veranso-Libalah & R.D.Stone (Heterotis arenaria), Gabon, Ogoouè-Maritime, J.C.M. Mbembo
& al. 171 (BR), BR000000946821, HETEAREN46, KX889277*, KY284802*, –; Guyonia ciliata Hook.f., Liberia, C.C.H. Jongkind 11859 (BR),
BR0000016915534, GUYOCILI131, KX889272*, KY284797*, KY284701*; Guyonia ciliata Hook.f., Cameroon, P. Mambo 248 (NY), JQ730084, –,
JQ730505; Guyonia jacquesii (A.Chev.) Veranso-Libalah & R.D.Stone (Heterotis jacquesii), Guinea, C.C.H. Jongkind 7437 (WAG), WAG0237101,
HETEJACQ144, KX889283*, KY284808*, –; Guyonia seretii (De Wild.) Veranso-Libalah & R.D.Stone (Heterotis seretii), D.R. Congo, M. Schaijes 2080
(BR), BR0000016904798, HETESERE48, KX889289*, KY284814*, –; Heterotis buettneriana (Cogn. ex Büttner) Jacq.-Fél., Gabon, Estuaire, E.L.A.N.
Simons & R. Westerduijn 291 (MO), MO-2454117, HETEBUET121, KX889278*, KY284803*, –; Heterotis cogniauxiana (A.Fern. & R.Fern.) VeransoLibalah & G.Kadereit, Angola, B. Goyder, Baker & Clark 7373 (GRA), HETECOGN164, KX889281*, KY284806*, KY284708*; Heterotis decumbens
(P.Beauv.) Jacq.-Fél., Gabon, S. Smith 1705 (US), JQ730088, JQ730302, JQ730509; Heterotis decumbens (P.Beauv.) Jacq.-Fél., Cameroon, South, Kribi,
M.C. Veranso 170 (MJG), 20978, HETEDECU37, KX889282*, KY284807*, KY284709*; Heterotis prostrata (Thonn.) Benth., Cameroon, Littoral, Edea,
PK37, M.C. Veranso 177 (MJG), 20963, HETEROTU34, KX889286*, KY284811*, KY284712*; Heterotis prostrata (Thonn.) Benth., Cameroon, SW,
Mundemba, M.C. Veranso 119 (MJG), 20991, HETEROTU35, KX889287*, KY284812*, KY284713*; Heterotis prostrata (Thonn.) Benth., cult. Botanic
Garden Meise (BR), 19851681, HETEROTU5, KX889288*, KY284813*, KY284714*; Heterotis rotundifolia (Sm.) Jacq.-Fél., Dominica, D.S. Penneys
1304 (FLAS), JQ730089, JQ730290, –; Melastoma candidum D.Don, Taiwan, C.H. Chen 6443 (TAIE), JQ730103, JQ730312, JQ730522; Melastoma
denticulatum Labill., New Caledonia, M.P. Simmons 1842 (BH), JQ730104, JQ730313, JQ730523; Melastoma malabathrichum L., China, D.S. Penneys
1998 (CAS), JQ730105, JQ730314, JQ730524; Melastoma sanguineum Sims, China, D.S Penneys 2000 (CAS), JQ730106, JQ730315, JQ730525;
Melastomastrum afzeli (Hook.f.) A.Fern & R.Fern., Guinea, C.C.H. Jongkind 10424 (BR), BR0000013189273, MELAAFZE155, KX889290*, KY284815*,
KY284715*; Melastomastrum capitatum (Vahl) A.Fern & R.Fern., Tanzania, Kigoma, C.J. Kayombo & Frank M. Mbago 1158 (MO), MO-04127,
MELACAPI162, KX889291*, KY284816*, KY284716*; Melastomastrum porteresii (Jacq.-Fél.) Veranso-Libalah & G.Kadereit (Osbeckia porteresii),
Guinea, G. Traore 48 (BR), BR0000016974258, OSBEPORT129, KX889299*, KY284825*, KY284721*; Melastomastrum porteresii (Jacq.-Fél.) VeransoLibalah & G.Kadereit (Osbeckia porteresii), Guinea, Nimba Mt., C.C.H. Jongkind 7439 (MO), 6177076, OSBEPORT17, KX889300*, KY284826*, –;
Melastomastrum segregatum (Benth.) A.Fern. & R.Fern., Gabon, F.J. Breteler 13067 (BR), BR0000007000271, MELASEGR135, KX889294*, KY284819*,
KY284717*; Melastomastrum theifolium (G.Don) A.Fern. & R.Fern., Cameroon, NW, Mbiame, M.C. Veranso 197 (MJG), 20988, MELATHEI10, –,
KY284820*, KY284718*; Melastomastrum theifolium (G.Don) A.Fern. & R.Fern., Cameroon, NW, G.F. De Wilde 7304 (BR), BR0000016884199,
MELATHEI83, KX889295*, KY284821*, KY284719*; Melastomastrum theifolium (G.Don) A.Fern. & R.Fern., Cameroon, M.C. Veranso 204 (MJG),
20987, MELATHEI95, KX889296*, KY284822*, KY284720*; Osbeckia australiana Naudin, Australia, K.G. Brennan 7008 (NY), JQ730116, JQ730328,
JQ730538; Osbeckia decandra DC., Guinea, Bidault & al. 153 (BR), BR0000016974180, OSBEDECA124, KX889297*, KY284823*, –; Osbeckia decandra
DC., Liberia, J.G. Adam 28699 (BR), BR0000016974197, OSBEDECA125, KX889298*, KY284824*, –; Osbeckia nepalensis Hook., China, D.S. Penneys
1986 (CAS), JQ730118, JQ730329, JQ730539; Osbeckia stellata Buch.-Ham. ex Ker Gawl., China, D.S. Penneys 1969 (CAS), JQ730119, JQ730330,
JQ730540; Osbeckia togoensis Leuenb., Benin, Akoeginou & al. 4268 (BR), BR0000016974401, OSBETOGO130, KX889301*, KY284827*, –; Pseudosbeckia
swynnertonii (Baker f.) A.Fern & A.Fern., Mozambique, Chimanimani, B. Wursten 2235 (BR), BR-SIL-SP-8403, PSEUSWYN4, KX889302*, KY284828*,
KY284722*; Tristemma akeasii Jacq.-Fél., Guinea, C.C.H. Jongkind 10696 (BR), BR000001697668, TRISAKEA147, KX889303*, KY284829*, –;
Tristemma albiflorum (G.Don) Benth., Guinea, C.C.H. Jongkind 10355 (BR), BR0000016975804, TRISALBI158, KX889304*, –, KY284723*; Tristemma
cornifolium (Benth.) Triana (Melastomastrum cornifolium), Liberia, J.G. Adam 27383 (WAG), WAG0049684, MELACORN152, KX889292*, KY284817*,
–; Tristemma corniflolium (Benth.) Triana (Melastomastrum corniflolium), Liberia, J.G. Adam 27383 (WAG), WAG0049685, MELACORN166, KX889293*,
KY284818*, –; Tristemma coronatum Benth., Ghana, H.H. Schmidt 3386 (CAS), JQ730243, JQ730452, JQ730664; Tristemma demeusei De Wild., Congo
(Brazzaville), Sangha, D.W. Thomas & al. 8838 (MO), TRISDEMEW159, KX889306*, KY284831*, KY284724*; Tristemma hirtum P. Beauv., Gabon, S.
Smith 1879 (US), JQ730244, JQ730453, JQ730665; Tristemma involucratum Benth., Liberia, C.C.H. Jongkind 12379 (BR), BR0000016976337,
TRISINVO154, KX889307*, KY284832*, KY284725*; Tristemma littorale Benth., Gabon, S. Smith 1725 (US), JQ730245, JQ730454, JQ730666; Tristemma
mauritianum J.F.Gmel., Madagascar, F. Almeda 8021 (CAS), JQ730246, JQ730455, JQ730667; Tristemma oreophilium Gilg, Gabon, F.J. Breteler 14156
(BR), BR0000017420655, TRISOREO153, KX889308*, KY284833*, KY284726*; Tristemma sp., Gabon, F.J. Breteler & al. 15253 (BR), BR0000016975965,
TRISDEMEP146, KX889305*, KY284830*, –; Tristemma vestitum Jacq.-Fél., Gabon, Champulvier 6169 (BR), TRISVEST157, KX889309*, KY284834*,
–.
614
Version of Record