Academia.eduAcademia.edu
Download Free PDF
paper cover icon
Download Free PDF
paper cover thumbnail

Phylogeny and systematics of African Melastomateae (Melastomataceae)

Phylogeny and systematics of African Melastomateae (Melastomataceae)

Profile image of Robert Douglas StoneRobert Douglas Stone

2017, Taxon

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).

Related Papers

Taxon

Towards a complete phylogeny of African Melastomateae: systematics of Dissotis and allies (Melastomataceae).

2020 •

Robert Douglas Stone

Dissotis has long been regarded as the most species-rich genus of African Melastomataceae, yet its diagnostic characters have never been examined in an explicitly phylogenetic context. In a previous study, we recovered a large clade consisting of "Dissotis and allies" but with poorly understood generic limits. Here we present a nearly complete phylogeny of African Melastomateae with focus on Dissotis and allies using sequences generated from two nuclear (nrETS, nrITS) and two plastid markers (ndhF, psbK-psbL), sampling 94 accessions representing 69 species including outgroups. We infer phylogenetic relationships using maximum likelihood, parsimony and Bayesian approaches and propose a revised generic classification of Dissotis and allies based on our molecular trees. In addition, we reconstruct the ancestral state of 20 important morphological characters of African Melastomateae. Dissotis and Antherotoma as previously circumscribed are polyphyletic, while all four of the previously recognised sections of Dissotis subg. Dissotis are monophyletic (i.e., sections Dissotis, Macrocarpae, Sessilifoliae and Squamulosae). We also recovered a clade consisting of D. congolensis, A. sene-gambiensis, A. clandestina and Nerophila gentianoides plus the species earlier treated in African Osbeckia. Ancestral character reconstruction indicates there are no unambiguous morphological synapomorphies for Dissotis s.l. In particular, the character of staminal dimorphism is shown to be homoplasious. However, the individual clades making up Dissotis and allies are each diagnosed by one to several morphological characters. In view of these results, we propose the following changes for the classification of African Melasto-mateae: (1) five new genera are described, Almedanthus (for Dissotis pachytricha), Eleotis (for D. sect. Sessilifoliae), Feliciotis (for D. sect. Macrocarpae), Pyrotis (for D. gilgiana) and Rosettea (for D. sect. Squamulosae); (2) the genus Nerophila is reinstated with an expanded circumscription to include Antherotoma senegambiensis, A. clandestina, D. congolensis, D. pauciflora and three species earlier treated in African Osbeckia; (3) circumscription of Dissotis s.str. is reduced to six species; (4) the monospecific genus Derosiphia is reinstated; and (5) two species previously placed in Dissotis are transferred to Antherotoma. To complete our generic realignment, 12 new synonyms and 58 new combinations are effected at or below the species level. A total of 40 lecto-or neotypes are designated. A revised key is provided for African Melastomateae, including all nine genera presently recognised in Dissotis and allies. Dissotis leo-nensis (type and only species of D. subg. Paleodissotis) together with the monospecific Cailliella and Dionychastrum remain unsampled.

View PDF

TAXON

Molecular phylogenetics of the Dissochaeta alliance (Melastomataceae): Redefining tribe Dissochaeteae

2021 •

Abdulrokhman Kartonegoro

The Dissochaeta alliance (= subtribe Dissochaetinae) is a group of woody climbing taxa within tribe Dissochaeteae (Melastomataceae) and comprises 90 species mainly distributed in Southeast Asia. The circumscription of the alliance and its genera has been problematic when based on only morphological characters. With a broad sampling of relevant taxa, this study aims at a natural circumscription of and revised generic delimitation within this alliance. A molecular phylogeny based on two nuclear (nrETS, nrITS) and four chloroplast (ndhF, psbK‐psbL, rbcL, rpl16) markers confirmed the non‐monophyly of Dissochaeteae s.l. Taxa traditionally placed in Dissochaeteae are found in three lineages in distant positions within Melastomataceae. The Medinilla alliance is resolved within Sonerileae + Oxysporeae, while the monophyletic Dinophora/Ochthocharis lineage forms a polytomy with Marcetieae, Melastomateae, Microlicieae, and Rhexieae; and the well‐supported Dissochaeta alliance is sister to Cam...

View PDF

South African Journal of Botany

Warneckea parvifolia (Melastomataceae-Olisbeoideae), a new "sand-forest" endemic from northeastern KwaZulu-Natal (South Africa) and southernmost Mozambique, and a phylogenetic analysis of eastern and southern African representatives of W. section Warneckea

2013 •

Robert Douglas Stone

Warneckea populations from “sand-forest” or “sand-thicket” habitats in Tembe Elephant Park, South Africa, and Licuati Forest Reserve in adjacent southern Mozambique were previously thought to be a small-leaved formof W. sousae, which typically includes larger-leaved plants ranging from central Mozambique northward to Tanzania. We examine this hypothesis using molecular and morphological evidence. Maximum-likelihood phylogenetic analysis of combined nrDNA ETS and ITS sequence data failed to resolve W. sousae and the Maputaland populations as an exclusively monophyletic group. Instead, the Kenyan endemic W. mouririifolia was strongly supported as the sister species of W. sousae, and the Maputaland plants were resolved in a separate, strongly supported clade together with populations of an as-yet undetermined Warneckea species from northern Mozambique. A hypothesis of exclusive monophyly for the plants from Tembe and Licuati had moderate support in separate ETS and ITS1 analyses (bootstrap proportions of 88% and 81%, respectively). Statistically significant differences in leaf dimensions and internode length were found between the Maputaland plants and typical W. sousae. We conclude that the populations from Tembe and Licuati represent a distinct species, which we describe as W. parvifolia. The species differs from W. sousae in having shorter internodes (mostly 5–25 mm not 10–60 mm long), smaller leaves (mostly 14–32 × 8–19 mm not 40–76 × 22–52 mm), shorter petioles (mostly 1–1.5 mm not 1.5–6 mm long), smaller flowers (hypanthium 1 × 1.5–1.75 mm not 1.5–2 × 2 mm; calyx lobes 0.5 mm not 0.75 mm long; staminal filaments 3–4 mm not 5 mm long; style 4–5 mm not 9 mm long), and globose fruit (not obovoid). An IUCN conservation status of Endangered (EN) B1a, b(ii, iii) is indicated for W. parvifolia, due to its limited distribution and projected declines in its habitat quality and area of occupancy.

View PDF

Botanical Journal of the Linnean Society

Phylogenetic relationships and distribution of New World Melastomeae (Melastomataceae)

2013 •

Frank Almeda

View PDF

South African Journal of Botany

Phylogenetic analysis of East and southern African Memecylon section Buxifolia (Melastomataceae): Insights on patterns and processes of diversification

2017 •

Robert Douglas Stone

Molecular phylogenetics is a valuable approach not only for discovering previously undetected plant diversity at the species level, but also for inferring the underlying processes of diversification. In southern Africa, two species of the woody genus Memecylon have been recognised, viz. M. natalense and M. bachmannii, with the range of M. natalense reportedly extending to northern Mozambique and southern Malawi. We investigated the evolutionary relationships of southern African Memecylon with population-level sampling and comparative sequencing of two fast-evolving DNA regions (nuclear rDNA ITS and 5′ ETS). Within M. natalense sensu lato, analyses revealed three geographically outlying populations representing distinct, semi-cryptic lineages (in forests of Maputaland and on the Soutpansberg and Mount Mulanje, respectively). Another sample from the Lindi region (southeastern Tanzania) produced an ITS sequence close to that of Kenyan M. fragrans while the ETS sequence was placed nearer to those of Mozambican M. torrei and M. natalense sensu stricto. This case of phylogenetic incongruence is taken to indicate the presence of an undescribed, homoploid hybrid species in the Rovuma region of southern Tanzania and northern Mozambique, although this hypothesis needs additional study. In KwaZulu-Natal and the Eastern Cape, rDNA spacer sequences were found to be identical or nearly so amongst samples of M. natalense and M. bachmannii, a surprising result given the clear morphological and ecological differences between these species. The minimal sequence variation would seem to indicate recent divergence or perhaps introgressive hybridisation. Further study is needed of the species boundary between M. natalense and M. bachmannii.

View PDF

Taxon

The Afro-Madagascan genus Warneckea (Melastomataceae): molecular systematics and revised infrageneric classification

2010 •

Robert Douglas Stone

Warneckea consists of shrubs and small trees endemic to tropical forests in Africa, Madagascar, and Mauritius. Phylogenetic analyses of sequence data from the transcribed spacers of nuclear ribosomal DNA (ETS, ITS) indicate that W. sect. Carnosae Jacq.-Fél. (1 sp., East Africa and Madagascar) is a divergent element that is best treated at subgeneric level. The analyses recovered three major lineages in W. subg. Warneckea, together forming a basal trichotomy. The three lineages represent W. sect. Strychnoides (western and central Africa), sect. Warneckea (East Africa, Madagascar, Mauritius), and a third group with three West African species (W. fascicularis, W. guineensis, W. mangrovensis) comprising the newly proposed W. sect. Guineenses.

View PDF

Taxon

The species-rich, paleotropical genus Memecylon (Melastomataceae): molecular phylogenetics and revised infrageneric classification of the African species

2014 •

Robert Douglas Stone

The internal and external transcribed spacers of nuclear ribosomal DNA were sequenced in 167 samples of Memecylon s.str. and 22 outgroup samples of Lijndenia, Mouriri, Spathandra, Votomita, and Warneckea. Maximum-likelihood analyses of ETS, ITS1, 5.8S and ITS2 alignments yielded tree topologies that are not significantly incongruent, with one exception involving the Tanzanian sample Luke 9741. Monophyly of Memecylon s.str. is strongly supported in the separate ETS and the combined ETS + ITS1 + 5.8S + ITS2 analyses. Also supported in all analyses except 5.8S is a sister-group relationship between a small species-group from western and central Africa (Memecylon subg. Mouririoidea, ovary 4-loculed) and the remaining taxa (M. subg. Memecylon, ovary unilocular). In the combined analysis, internal branches at the base of M. subg. Memecylon are short and weakly supported, yet within this large subgenus one finds a series of monophyletic groups representing different parts of the widespread paleotropical distribution (one group in western and central Africa; two separate groups in East Africa, one of these extending to southern Africa and disjunctly to western and northern Madagascar; one species-rich group occurring exclusively on Madagascar and the neighboring Comoro and Mascarene islands; and three distinct groups in Indo-Malesia, one of these also including the Seychelles endemic M. elaeagni). Within the western and central African clade, M. sect. Polyanthema sensu Jacques-Félix is paraphyletic with respect to M. sect. Afzeliana, and at the morphological level it seems to be a group defined by symplesiomorphies (ovary unilocular, fruit globose). Three East African species (M. fragrans, M. greenwayi, M. semseii) are returned to Memecylon s.str. after being erroneously transferred to Lijndenia by Borhidi. T he Madagascan endemic M. mocquerysii is no longer considered a taxonomic synonym of the distantly related, Tanzanian M. cogniauxii. In addition to elevating M. sect. Mouririoidea to subgeneric rank, the following changes are proposed in the infrageneric classification of African Memecylon: (1) the circumscription of sect. Polyanthema is narrowed to comprise only the members of the “M. polyanthemos complex” sensu Jacques-Félix; (2) Engler’s sections Tenuipedunculata, Cauliflora, and Obtusifolia are re-instated with emended descriptions (in the case of sect. Cauliflora with an expanded circumscription); (3) seven new sections, Buxifolia, Diluviana, Felixiocylon, Germainiocylon, Magnifoliata, Montana, and Sitacylon, are described; (4) the purported occurrence of M. sect. Pseudonaxiandra in East Africa is rejected. A key is provided to the two subgenera and twelve sections currently recognized in African Memecylon. Further study is needed toward a sectional classification of Indo-Malesian Memecylon, and for revision of the seven Madagascan sections recognized by Jacques-Félix.

View PDF
A Guide to Curating New World Melastomataceae Collections with a Linear Generic Sequence to World-Wide Melastomataceae

Frank Almeda

The following guide is aimed at aiding in the curation of herbarium collections in Melastomataceae, with an emphasis on the New World species. It contains a summary of the taxonomic realignments at the tribal and generic level within Neotropical taxa of Melastomataceae, as well as some general comments for other groups. A table with a generic linear sequence is also provided, as well as tables with new and synonymized New World genera since 2005 and all currently accepted species. Lastly, a table with the synonyms of over 1000 accepted Neotropical species that have been impacted by these generic realignments is also provided.

View PDF

Willdenowia

Phylogenetic placement of Cailliella praerupticola (Melastomataceae), a rare, monospecific lineage from Guinea, West Africa

2021 •

Robert Douglas Stone, pepe Marc Haba

The genus Cailliella includes a single species, C. praerupticola, restricted to a sandstone plateau region in Coyah, Dubreka, Forécariah and Kindia Préfectures, Guinea, West Africa. Until its rediscovery in 2016, this species was known only by the type gathering made in 1937. We provide an emended description, photographs of living material, locality and habitat data, and a discussion of its conservation status. We further designate a lectotype for C. praerupticola. On the basis of morphology, Cailliella had been considered an isolated member of the tribe Melastomateae, possibly sister to Melastomastrum. We tested the classification of Cailliella by means of a phylogenetic analysis of C. praerupticola and 40 other samples of African Melastomateae, based on molecular sequence data from the external and internal transcribed spacer regions of nuclear ribosomal DNA (nrETS and nrITS) as well as the ndhF, psbK-psbL and accD-psaI regions of chloroplast DNA. Our results corroborate the placement of Cailliella in the Melastomateae, where it was resolved as sister to a clade consisting of both Melastomastrum and Tristemma. The hypothesis of C. praerupticola being an evolutionary relict or “palaeoendemic” is supported by our findings.

View PDF
Ortiz & al. • Phylogeny of African Mutisieae

Santiago Ortiz

View PDF
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

RELATED PAPERS

Integration of statistics and food process engineering: Assessing the uncertainty of thermal processing and shelf-life estimations

2010 •

Concepción Pérez-Lamela

View PDF

Material Culture Review Revue De La Culture Materielle

A Different Way of Sitting: American Patent Folding Chairs of the 19th Century

2012 •

Tania Batley

View PDF

PLOS ONE

Mismatch repair deficiency and aberrations in the Notch and Hedgehog pathways are of prognostic value in patients with endometrial cancer

Georgia Karayannopoulou

View PDF
Microfinance Market Diagnoses and Beneficiaries Impact- Case of Albania

2014 •

Dr. Forcim Kola

View PDF
Outcomes Based Education, PBL and Kolb's Learning Cycle Help Forest Engineering Students Learn Forest Operations Planning

2004 •

E. Robak

View PDF

Journal of the American Academy of Child and Adolescent Psychiatry

Ode to the Brits

2018 •

julie chilton

View PDF

JualObatPenghilangBuluPalingAmpuhDiMedan

WA 0852-7060-4590 Jual Obat Penghilang Bulu Paling Ampuh Di Medan

2022 •

Jual Obat P E R O N T O K Bulu

View PDF

British Dental Journal

Trigeminal nerve injuries in relation to the local anaesthesia in mandibular injections

2010 •

Tara Renton

View PDF

Investigative Ophthalmology & Visual Science

Mitochondrial Sequence Changes in Keratoconus Patients

2014 •

Tahira Sultan

View PDF

Trabajo Y Sociedad

A comparative experience on classification of unemployed people and receptors of conditional income transfer programs

2010 •

Pilar Arcidiacono

View PDF

JURNAL PANGAN

Teknologi Pengolahan Dodol dan Peningkatan Kandungan Gizinya

2019 •

Adil Ahza

View PDF

Catalysts

CO2 Methanation on Supported Rh Nanoparticles: The combined Effect of Support Oxygen Storage Capacity and Rh Particle Size

2020 •

Ioannis Yentekakis

View PDF

Developing Leadership and Employee Health Through the Arts

Erratum: Developing Leadership and Employee Health Through the Arts

2016 •

Julia Romanowska

View PDF

Cartografías del Sur. Revista de Ciencias, Artes y Tecnología

La Planificación como herramienta de gestión

2019 •

Silvina Pane

View PDF

Archives of Virology

Novel smacoviruses identified in the faeces of two wild felids: North American bobcat and African lion

2019 •

Laurel Serieys

View PDF

Journal of Nanoparticle Research

Nanostructure, solvation dynamics, and nanotemplating of plasmonically active SERS substrate in reverse vesicles

2013 •

Ranajay Saha

View PDF

Jurnal Penelitian Enjiniring

Strategi Pengembangan Infrastruktur Energi Listrik untuk Mewujudkan Makassar Sustainable City

2019 •

Ichsan Caesar Pratama

View PDF
Bone histology and growth of chasmosaurine ceratopsid dinosaurs from the late Campanian Kaiparowits Formation, Southern Utah

2013 •

Carolyn Levitt

View PDF
Calcaterra 08A0328 Estratti Miscellania DG

Emanuela Calcaterra

View PDF

Journal of Nuclear Materials

Materials research under ITER-like divertor conditions at FOM Rijnhuizen

2011 •

R Al

View PDF

Neuroscience Letters

High sensitivity quantitative analysis of cobalt uptake in rat cerebellar granule cells with and without excitatory amino acids

1998 •

Didier Perret

View PDF

Salud Pública de México

Historical distribution of central nervous system tumors in the Mexican National Institute of Neurology and Neurosurgery

2016 •

Julio Sotelo

View PDF

Tourism Management

Breaking boundaries: Exploring gendered challenges and advancing equality for Iranian women careers in tourism

2024 •

Fernando Almeida García, Seyedasaad Hosseini

View PDF
Aktivitas Antioksidan Senyawa Asam 5(E)-Eikosenoat dari Batang Tanaman Polygonum pulchrum Bl

2015 •

Rini Hamsidi

View PDF

IIUM Medical Journal Malaysia

Prediction of Clinical Symptoms of Schizophrenia Based on Comt Methylation Marker

2017 •

Norlelawati Talib

View PDF

RELATED TOPICS

AfricaPlant Taxonomy (Taxonomy)Plant SystematicsMelastomataceaeMolecular phylogenetics