Phylogeny of the complex Vanguerieae (Rubiaceae) genera Fadogia, Rytigynia, and Vangueria with close relatives and a new circumscription of Vangueria

Birgitta Bremer

Pl. Syst. Evol. (2005) DOI 10.1007/s00606-005-0313-9 Phylogeny of the complex Vanguerieae (Rubiaceae) genera Fadogia, Rytigynia, and Vangueria with close relatives and a new circumscription of Vangueria H. Lantz1 and B. Bremer2 1 2 Department of Systematic Botany, Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden The Bergius Foundation at the Royal Swedish Academy of Sciences, Stockholm, Sweden Received August 14, 2004; accepted February 9, 2005 Published online: May 25, 2005 Springer-Verlag 2005 Abstract. The phylogeny of the Vanguerieae genera Fadogia, Rytigynia, Vangueria, and closely related genera is investigated using nuclear rDNA ITS sequences and plastid trnT-F and rpl16 sequences. Individual and combined analyses reveal several strongly supported clades. There are indications that Fadogia, Rytigynia, Tapiphyllum, and Vangueria are para- or polyphyletic and only Multidentia is strongly supported as monophyletic. Several taxa are found to have incongruent positions in the ITS and chloroplast phylogenies, and possible reasons behind these incongruencies are discussed. For Ancylanthos rubiginosus a chloroplast capture event can explain the incongruent position. In the Fadogia-Rytigynia group the incongruence is more widespread and cannot be attributed to a single taxon or a few taxa, but hybridization and introgression is the most likely explanation for the incongruence. It is concluded that the genera Ancylanthos, Lagynias, Pachystigma, and Tapiphyllum and the three species Fadogia agrestis, Rytigynia fuscosetulosa, and Rytigynia induta should be transferred to Vangueria. Key words: Fadogia, hybridization, phylogeny, Rubiaceae, Rytigynia, taxonomy, Vangueria, Vanguerieae. Introduction Until recently, the tribe Vanguerieae of the Rubiaceae remained one of the poorest understood larger groups in the family concerning the phylogeny of the group. Recent studies (Lantz et al. 2002, Lantz and Bremer 2004) have improved our knowledge and now there is a good general knowledge of the phylogeny. The monophyly of the tribe has rarely been doubted (Verdcourt 1958). A very distinct and easily recognized character, a pollen presenter at the apex of the style is of great diagnostic value and makes it easy to distinguish between members of Vanguerieae and other Rubiaceae tribes also in the ﬁeld. Eastern and southern tropical African taxa have so far received the most attention (Verdcourt and Bridson 1991, Bridson 1998). Around 600 species are currently recognized, but the delimitation of many of these is uncertain, especially for the poorly known species from Madagascar and South East Asia. An earlier study (Lantz and Bremer 2004) identiﬁed one subclade of the tribe in which the generic limits were especially unclear. Due to H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera the slightly longer corolla tubes and lobes present in this group compared to other members of Vanguerieae, the group was given the informal name ‘the large-ﬂowered group’. This group is the focus of the present study. The large-ﬂowered group is with few exceptions restricted to Africa south of Sahara, but a few Rytigynia and Vangueria species are found on Madagascar. Two subgroups of the large-ﬂowered group had earlier been identiﬁed (Lantz et al. 2002), i.e. the Vangueria group and the Fadogia-Rytigynia group. Included in the Vangueria group is the type genus of the tribe, Vangueria, together with Ancylanthos (tentatively included), Lagynias, Pachystigma, and Tapiphyllum (Lantz et al. 2002, Lantz and Bremer 2004). These genera are considered some of the most problematic in the tribe, concerning delimitation of species as well as genera (Bridson 1998). The four genera Fadogia, Fadogiella, Hutchinsonia, and Rytigynia form the species-rich Fadogia-Rytigynia group (Lantz et al. 2002) with just above 100 species. Cuviera, Multidentia, Pygmaeothamnus, Robynsia, Vangueriopsis and a single species of Canthium, C. oligocarpum, also belong to the large-ﬂowered group, but their phylogenetic aﬃnities are more uncertain (Lantz and Bremer 2004). In total, over 180 species are currently recognized in the group, but a number of these, especially in Rytigynia, are known from very few collections. The work of Robyns (1928) has been especially important for the classiﬁcation of the group. With the exclusion of Cuviera and the large genus Canthium, Robyns revised all species of Vanguerieae, and made over a hundred new combinations and descriptions of species in the large-ﬂowered group. He also described several new genera, of which Fadogiella, Hutchinsonia, Pygmaeothamnus, Tapiphyllum, and Vangueriopsis are known to belong to the large-ﬂowered group (Lantz and Bremer 2004). Later authors (e.g. Verdcourt 1981, 1987; Bridson 1996, 1998) have added new species to the genera in the group, and also performed several synonymizations. Over 180 species representing thirteen genera are currently recognized in the largeﬂowered group. Most of the genera are hard to delimit morphologically, good diagnostic characters are usually lacking. In earlier studies covering the whole tribe (Lantz et al. 2002, Lantz and Bremer 2004), but with a rather small sample of the large-ﬂowered group, we were unable to resolve the relationships in the group. With this study we aim to resolve the phylogeny with strong support by greatly increasing the sample size and adding an extra data set to the already existing data. The new data set consists of sequences of the rpl16 intron, one of the fastest evolving chloroplast regions (Small et al. 1998) and is likely to add support for the close relationships in this diﬃcult group. We intend to use the phylogeny to test the monophyly of the genera as currently conceived and if necessary base a new classiﬁcation on the results. Some incongruence between chloroplast and nuclear phylogenies was noted in an earlier publication (Lantz and Bremer 2004), and we also aim to quantify the extent to which the diﬀerent genera are aﬀected by this. Material and methods Taxon sampling. Our strategy was to include as many species as possible from all genera known to belong in the large-ﬂowered clade (Lantz and Bremer 2004). Sixty-six ingroup (of 180) taxa were sampled. The availability of material largely determined which species were included. A majority of the in-group species sampled occur in eastern and southern Africa, but ﬁve species from West Africa, one from Central Africa, and two from Madagascar are also included (see Fig. 1) . The distributions of the taxa detailed only to major geographical regions are given in Fig. 1. This information is added only to give an indication of the distribution of the taxa. The taxa are in many cases restricted to smaller areas within these regions and can also in some small amount occur outside of the mentioned areas. Eight representatives of Canthium s. str., shown to be the closest relative of the ingroup (Lantz et al. 2002, Lantz and Bremer 2004), were chosen as outgroup. Several of these were only H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Table 1. Voucher information and EMBL accession numbers. Herbarium abbreviations according to Holmgren et al. (1990). - indicates sequences not included due to sequencing diﬃculties. Sequences new for this study are underlined. Species Voucher EMBL number ITS/trnT-F/ rpl16 Canthium coromandelicum (Burm. f.) Alston Canthium glaucum Hiern ssp. glaucum Canthium inerme (L.f.) Kuntze Canthium tetraphyllum (Schweinf. ex Hiern) Baill. Canthium armatum (K. Schum.) Lantz Pygmaeothamnus cf. chamaedendrum (Kuntze) Robyns Canthium bugoyense (K. Krause) Lantz Canthium stenosepalum Lantz Andreasen 36 (UPS) AJ315081/AJ620122/AJ876810 Kuchar 17410 (UPS) AJ617752/AJ620124/AJ876811 Bremer & Bremer 3686 (UPS) Bremer 3074 (UPS) AJ315120/AJ620125/AJ876812 AJ315083/AJ620149/AJ876813 Bremer & Bremer 3790 (UPS) AJ315082/AJ620155/AJ876814 Bremer & Bremer 3800 (UPS) AJ315119/AJ620165/AJ876815 Rwburindore 3536 (UPS) AJ315084/AJ620172/AJ876816 Merello et al. 1494 (K) AJ315085/AJ620182/AJ876817 Ancylanthos rubiginosus Desf. Canthium oligocarpum Hiern ssp. captum (Bullock) Bridson Cuviera angolensis Welw. ex K. Schum. Fadogia agrestis Schweinf. ex Hiern Fadogia ancylantha Hiern Fadogia arenicola K. Schum. & K. Krause Fadogia cienkowskii Schweinf. Fadogia elskensii De Wild. Fadogia stenophylla Hiern Fadogia tetraquetra K. Krause Fadogia triphylla Baker Fadogia verdcourtii Tennant var. verdcourtii Fadogiella stigmatoloba (K. Schum) Robyns Hutchinsonia barbata Robyns Lagynias dryadum (S. Moore) Robyns Lagynias lasiantha (Sond.) Bullock Lagynias monteiroi (Oliv.) Bridson Lagynias pallidiﬂora Bullock Zimba et al. 776 (UPS) Borhidi et al. 85449 (UPS) AJ617747/AJ620119/AJ617755/AJ620129/AJ876818 McPherson 16297 (MO) AJ315088/AJ620134/- Madsen 5495 (S) AJ874980/AJ874942/AJ876819 Chapman & Chapman 9109 (UPS) AJ315103/AJ620136/AJ876820 AJ874981/AJ874943/AJ876821 Gereau et al. 6011 (UPS) Lantz 101 (UPS) Taylor et al. 8318 (UPS) Mwangoka 1742 (UPS) Bremer & Bremer 3799 (UPS) Bidgood et al. 589 (UPS) Gereau et al. 6010 (UPS) -/AJ620137/AJ876822 AJ315118/AJ719191/AJ876823 -/AJ875117/AJ876824 AJ315099/AJ620139/AJ876825 AJ874982/AJ874944/AJ876826 AJ315116/AJ620140/AJ876827 Lawton 1318 (S) AJ315100/AJ620141/AJ876828 Adam 20599 (UPS) Bremer & Bremer 3811 (UPS) AJ315102/AJ620142/AJ876829 AJ315090/AJ620146/AJ876830 Bremer & Bremer 3792 (UPS) AJ315089/AJ620147/AJ876831 Bremer et al. 4297 (UPS) AJ874983/AJ874945/AJ876832 Ntemi Sallu et al. 309 (UPS) AJ874984/AJ874946/AJ876833 H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Table 1. (Continued) Species Voucher EMBL number ITS/trnT-F/ rpl16 Multidentia concrescens Bidgood et al. 845 (K) AJ315086/AJ620150/AJ876834 (Bullock) Bridson & Verdc. AJ874985/AJ874947/AJ876835 Multidentia crassa (Hiern) Lantz 150 (UPS) Bridson & Verdc. Multidentia fanshawei Lovett et al. 3311 (K) AJ315087/AJ620151/AJ876836 (Tennant) Bridson AJ874986/AJ874948/AJ876837 Multidentia sclerocarpa Luke 9442 (UPS) (K. Schum.) Bridson Pachystigma gillettii Gilbert & Thulin 128 (UPS) AJ874987/AJ874949/AJ876838 (Tennant) Verdc. Pachystigma latifolium Bremer et al. 4304 (UPS) -/AJ874950/AJ876839 Sond. Pachystigma loranthifolium Luke 9464 (UPS) AJ874988/AJ874951/AJ876840 (K. Schum.) Verdc. Pachystigma pygmaeum Pawek 12335 (BR) AJ315091/AJ620152/AJ876841 Robyns Pachystigma schumannianum Mlangwa & Mbuso 1274 (UPS) AJ874989/AJ874952/AJ876842 (Robyns) Bridson & Verdc. Pygmaeothamnus zeyheri (Sond.) Bremer et al. 4356 (UPS) AJ617773/AJ620166/AJ876843 Robyns var. zeyheri Robynsia glabrata Hutchinson Hall & Amponsah 46545 (K) AJ617774/AJ620170/Rytigynia adenodonta (K. Schum.) Robyns var. reticulata (Robyns) Verdc. Lantz 108 (UPS) AJ874990/AJ874953/AJ876844 Rytigynia bagshawei Borhidi et al. 84439 (UPS) AJ315101/AJ620171/AJ876845 (S. Moore) Robyns var. bagshawei Rytigynia beniensis Festo 1151 (UPS) AJ874991/AJ874954/AJ876846 (De Wild.) Robyns Rytigynia celastroides Mwangoka & Maingo 1514 (UPS) AJ874992/AJ874955/AJ876847 (Baillon) Verdc. AJ874993/AJ874956/AJ876848 Rytigynia decussata Rulangaranga et al. 95 (MO) (K. Schum.) Robyns Rytigynia eickii (K. Schum. & Borhidi et al. 84003 (UPS) AJ874994/AJ874957/AJ876849 K. Krause) Bullock Rytigynia fuscosetulosa Verdc. Frimodt Moller et al. NG117 (K) AJ315097/AJ620173/Rytigynia hirsutiﬂora Verdc. Luke et al. 9189 (UPS) AJ874995/AJ874958/AJ876850 AJ874996/AJ874959/AJ876851 Rytigynia induta (Bullock) Verdc. Luke 9051 (UPS) Rytigynia lichenoxenos Luke et al. 9104 (UPS) AJ874997/AJ874960/AJ876852 (K. Schum.) Robyns Rytigynia longicaudata Verdc. Borhidi & Pocs 85206 (UPS) -/AJ874961/AJ876853 Rytigynia longipedicellata Verdc. Bidgood et al. 1577 (UPS) -/AJ874962/AJ876854 Rytigynia monantha (K. Schum.) Hedrén et al. 308 (UPS) -/AJ874963/AJ876855 Robyns Rytigynia neglecta (Hiern) Robyns Gilbert et al. 7937 (UPS) -/AJ874964/AJ876856 H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Table 1. (Continued) Species Voucher EMBL number ITS/trnT-F/ rpl16 Rytigynia parvifolia Verdc. Rytigynia pseudolongicaudata Verdc. Rytigynia sambavensis Cavaco Rytigynia senegalensis Blume Rytigynia seyrigii Cavaco Rytigynia sp. L of F.T.E.A. Rytigynia uhligii (K. Schum. & K. Krause) Verdc. Rytigynia umbellulata (Hiern) Robyns Rytigynia xanthotricha (K. Schum.) Verdc. Tapiphyllum cinerascens (Welw. ex Hiern) Robyns var. cinerascens Tapiphyllum obtusifolium (K. Schum.) Robyns Tapiphyllum velutinum (Hiern) Robyns Vangueria apiculata K. Schum. Vangueria esculenta S. Moore Vangueria infausta Burch. Vangueria madagascariensis J. F. Gmelin Vangueria parvifolia Sond. Vangueria praecox Verdc. Vangueria proschii Briq. Vangueria randii S. Moore Vangueria soutpansbergensis N. Hahn Vangueria volkensii K. Schum. Vangueriopsis cf. longiﬂora Verdc. Luke 8345 (UPS) Kayombo 1953 (UPS) Davis 2157 (K) van den Berghen 8746 (BR) Gereau et al. 5731 (MO) Luke 9039 (UPS) Borhidi et al. 84458 (UPS) AJ874998/AJ874965/AJ876857 AJ874999/-/AJ875000/AJ874966/AJ315104/AJ620175/AJ876858 AJ875001/AJ874967/AJ875002/AJ874968/AJ875003/AJ874969/AJ876859 Gobbo & Gilagiza 860 (UPS) AJ875004/AJ874970/AJ876860 Borhidi et al. 87353 (UPS) AJ875005/AJ874971/AJ876861 Milne-Redhead 3292 (BR) AJ315096/AJ620177/AJ876862 Gereau et al. 6669 (UPS) AJ875006/AJ874972/AJ876863 Emanuelsson 672 (S) AJ315098/AJ620178/AJ876864 Kårehed & Odhult 161 (UPS) Lantz 145 (UPS) Bremer et al. 4254 (UPS) Bremer 3077 (UPS) AJ315095/AJ620179/AJ876865 AJ875007/AJ874973/AJ876866 AJ617777/AJ620180/AJ876867 AJ224839/AJ620184/AJ876868 Bremer & Bremer 3771 (UPS) Luke 9357 (UPS) Blomberg et al. BMP 466 (UPS) Luke 5161 (EA) Bremer et al. 4324 (UPS) AJ315092/AJ620181/AJ876869 AJ875008/AJ874974/AJ876870 AJ875009/AJ874975/AJ876871 AJ875010/AJ874976/AJ876872 AJ875011/AJ874977/AJ876873 Borhidi et al. 85162 (UPS) Luke 8316 (UPS) AJ875012/AJ874978/AJ876874 AJ617778/AJ620183/AJ876875 recently transferred to Canthium (Lantz and Bremer 2004). Molecular methods. DNA was extracted using the CTAB method (Doyle and Doyle 1987) from herbarium material, fresh, or silica-dried material. The DNA was puriﬁed using either QiaQuick PCR puriﬁcation kit (QIAGEN) or caesium chloride/ ethidium bromide gradient centrifugation. PCR ampliﬁcation and sequencing for the ITS and trnTF region were performed as in Lantz and Bremer (2004). Ampliﬁcation and sequencing of rpl16 were performed as for the trnT-F region, but used the following primers: L16exon1 (Downie et al. 2000) and 1067F (C. Asmussen; pers. comm.) for the PCR ampliﬁcation and primers1067F and rpl1618R (Asmussen 1999) for sequencing. Phylogenetic methods. Sequences were aligned manually. Informative insertion/deletion events (indels) were coded manually using the simple method of Simmons and Ochoterena (2000). The phylogenies were constructed using parsimony with the program PAUP* (Swoﬀord 2002). All three data sets (ITS, trnT-F, and rpl16) were run together in a combined analysis as well as separately. A chloroplast data set (trnT-F + rpl16) was also analyzed. Incongruence between the ITS and chloroplast data sets was assessed by the comparison of bootstrap values from the separate analyses. H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera 63 88 94 79 68 97 100 * 51 62 97 77 98 91 95 79 71 100 75 100 99 84 87 100 75 100 75 63 D 51 99 83 100 * 57 100 * 96 93 95 C 98 100 100 Canthium coromandelicum India Canthium glaucum E Canthium tetraphyllum E Canthium armatum S, Z Outgroup Pygmaeothamnus cf. chamaedendrum S Canthium inerme S, Z Canthium bugoyense C, E, Z Canthium stenosepalum W Ancylanthos rubiginosus A, Z Fadogia agrestis W Pachystigma gillettii E Pachystigma loranthifolium E Pachystigma schumannianum E Lagynias dryadum Z, S Lagynias lasiantha S Lagynias monteiroi S Lagynias pallidiflora E Pachystigma latifolium S Pachystigma pygmaeum E, S, Z Tapiphyllum cinerascens C, E, Z Vangueria apiculata E, Z Vangueria group Vangueria volkensii E Vangueria esculenta Z Vangueria infausta E, S, Z Vangueria proschii A, Z Vangueria madagascariensis Afr. Vangueria parvifolia S, Z Vangueria soutpansbergensis S Vangueria randii E, Z Rytigynia fuscosetulosa E Rytigynia induta E Vangueria praecox E Tapiphyllum obtusifolium E Tapiphyllum velutinum Z Canthium oligocarpum E, Z Cuviera angolensis W Multidentia concrescens E, Z Multidentia crassa E, Z Multidentia fanshawei E, Z Multidentia sclerocarpa E Pygmaeothamnus zeyheri C, E, Z, S Robynsia glabrata W Vangueriopsis cf. longiflora E Fadogia ancylantha C, E, W, Z Fadogia triphylla Afr. Fadogia verdcourtii E Fadogia arenicola E, Z Fadogiella stigmatoloba E, Z Fadogia cienkowskii Afr. Fadogia elskensii C, E, Z Fadogia stenophylla A, C, E, Z Fadogia tetraquetra Afr. Rytigynia umbellulata Afr. Rytigynia adenodonta E, Z Rytigynia celastroides E Rytigynia parvifolia E Rytigynia lichenoxenos E Rytigynia sp. L E Fadogia-Rytigynia Rytigynia uhligii E, Z group Rytigynia beniensis C, E Rytigynia neglecta E Hutchinsonia barbata W Rytigynia senegalensis W Rytigynia bagshawei C, E Rytigynia hirsutiflora E Rytigynia longicaudata E Rytigynia monantha E, Z Rytigynia xanthotricha E Rytigynia pseudolongicaudata E Rytigynia sambavensis M Rytigynia seyrigii M Rytigynia decussata E Rytigynia eickii E Rytigynia longipedicellata E H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Table 2. Data set characteristics Data set Number of taxa ITS 67 trnT-F Number of characters Informative characters (including indels) Informative indel characters 712 157 20 73 2096 111 22 rpl16 66 854 60 7 trnT-F + rpl16 73 2950 171 29 ITS + trnT-F+ rpl16 74 3662 328 49 ITS + trnT-F+ rpl16, incongruent taxa removed 69 3662 311 46 When strongly incongruent taxa had been identiﬁed (see results), all data sets were analyzed again but with the incongruent taxa removed. A heuristic search with 100 random addition replicates and TBR branch-swapping was performed to ﬁnd the most parsimonious trees. Support was measured by the bootstrap option as implemented by PAUP* using 10000 bootstrap replicates, each with 5 random addition replicates, TBR branch-swapping and MULTREES oﬀ. Results The number of taxa, total number of characters, number of informative characters, and number of indel-characters for each of the six data sets are presented in Table 2. We were unable to get ITS sequences for seven species, trnT-F sequences for one species, and rpl16 sequences for eight species (Table 1). The ITS data set has the highest number of informative characters followed by trnT-F, and rpl16. The comparison is, however, likely inﬂuenced by the uneven sample sizes of the diﬀerent data sets. About half of the informative characters stem from the ITS data set. For all of the analyses, the number of most parsimonious trees was too large to be saved and the analyses did not run to completion. We therefore choose to present the majority rule consensus trees from the bootstrap analyses. Results from four of the analyses are presented; the combined analysis (ITS + trnT-F + rpl16; Fig. 1), ITS (Fig. 2), the chloroplast tree (trnTF + rpl16; Fig. 3), and a combined analysis with strongly incongruent taxa removed (Fig. 4 ). We found no cases of topological incongruence between the phylogenies based on the trnT-F and rpl16 data sets (results not b Fig. 1. Majority rule consensus tree from the bootstrap analysis of the combined (ITS + trnT-F + rpl16) data set. Bootstrap support ‡50 is included above nodes. Letters in the tree refer to clades discussed in the text. Letters added after the taxa names indicate geographical distribution (see material and methods), i.e. Afr.=Subsaharan Africa, A=Angola with Namibia, C=Central Africa, E=East Africa, M=Madagascar, S=South Africa, Z=Zambia with Botzwana, Malawi, Mozambique, and Zimbabwe (Zambesian distribution), W=West Africa. Types of genera are indicated in bold. Clades which receive strong support both from the plastid and nuclear data sets are indicated by an * H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera 77 57 57 92 100 55 64 84 96 95 67 54 64 65 53 82 F 100 62 98 92 E 63 G 52 96 50 58 77 100 100 72 99 54 92 92 Canthium coromandelicum Canthium glaucum Canthium tetraphyllum Canthium armatum Canthium inerme Pygmaeothamnus cf. chamaedendrum Canthium bugoyense Canthium stenosepalum Ancylanthos rubiginosus Fadogia agrestis Lagynias pallidiflora Pachystigma loranthifolium Pachystigma gillettii Pachystigma schumannianum Vangueria apiculata Vangueria volkensii Vangueria randii Lagynias dryadum Lagynias lasiantha Lagynias monteiroi Pachystigma pygmaeum Rytigynia fuscosetulosa Rytigynia induta Vangueria praecox Tapiphyllum cinerascens Vangueria esculenta Vangueria infausta Vangueria proschii Vangueria madagascariensis Vangueria parvifolia Vangueria soutpansbergensis Tapiphyllum obtusifolium Tapiphyllum velutinum Canthium oligocarpum Multidentia concrescens Multidentia crassa Multidentia fanshawei Multidentia sclerocarpa Cuviera angolensis Robynsia glabrata Fadogia ancylantha Fadogia triphylla Fadogia verdcourtii Fadogia arenicola Fadogiella stigmatoloba Rytigynia adenodonta Rytigynia beniensis Rytigynia celastroides Rytigynia parvifolia Rytigynia sp. L Rytigynia lichenoxenos Rytigynia uhligii Fadogia elskensii Fadogia tetraquetra Rytigynia umbellulata Rytigynia decussata Rytigynia eickii Hutchinsonia barbata Rytigynia bagshawei Rytigynia hirsutiflora Rytigynia pseudolongicaudata Rytigynia xanthotricha Rytigynia sambavensis Rytigynia seyrigii Rytigynia senegalensis Pygmaeothamnus zeyheri Vangueriopsis cf. longiflora Outgroup Vangueria group Fadogia-Rytigynia group H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera shown). Since both genes are from the plastid genome they are likely to share the same evolutionary history. We therefore combined the two data sets into a single chloroplast data set and only present the results from the bootstrap analysis of this data set. Compared to the phylogenies of the individual (trnT-F and rpl16) data sets, the combined chloroplast tree has both increased resolution and support. There are strongly supported clades in all presented phylogenies. Comparing the ITS and chloroplast phylogenies, both include a similar number of strongly supported clades (above 90% bootstrap), but only in three instances are clades supported strongly from both data sets (see Fig. 1). Comparisons are, however, made diﬃcult by the uneven sample sizes of the chloroplast and ITS data sets. This is especially true in the Fadogia-Rytigynia group in which seven taxa are present in only one of the data sets (Table 1). There are also several cases of incongruence in the comparison between the ITS and chloroplast phylogenies. Most incongruencies are weakly supported (e.g. the position of Rytigynia induta), but there are some that are more strongly supported, i.e. situations where a strongly supported node in one of the phylogenies is contradicted by an at least moderately supported node in the other. In the Fadogia-Rytigynia group the positions of Hutchinsonia barbata, Rytigynia beniensis, and the clade consisting of R. decussata and R. eickii are strongly incongruent. In the Vangueria group, no taxon is strongly incongruent as deﬁned above, but Ancylanthos rubiginosus has signiﬁcantly diﬀerent positions in the two trees. In the ITS tree, Ancylanthos rubiginosus has an internal position in the group, but is in the chloroplast tree sister to a clade consisting not only of the other members of the Vangueria group but also Multidentia, Pygmaeothamnus, Robynsia, and Vangueriopsis. Cases where taxa have an unresolved position in one of the phylogenies and a resolved in the other (e.g. the positions of Cuviera angolensis and Robynsia glabrata) could possibly be inﬂuenced by incongruence, but the unresolved position in one of the phylogenies makes any comparisons uninformative. Of more importance are strongly supported nodes which collapse when the data sets are combined. In the Vangueria group, no strongly supported clades present in the chloroplast or ITS phylogenies collapse in the combined phylogeny, although clade A (Figs. 3 and 4) is more strongly supported (bp=96) in the chloroplast phylogeny than in the combined phylogeny (bp=79), indicating some conﬂict. Of greater concern are the strongly supported clades present in the ITS phylogeny for a clade in the Fadogia-Rytigynia group (clade E; Fig. 2 ) and the moderately supported clade of Cuviera and Robynsia (clade F; Fig. 2 ) that collapses in the combined phylogeny ( Fig. 1 ). These clades are not present in the chloroplast phylogeny (Fig. 3) and the incongruence can therefore not be attributed to a single taxon or a single clade such as in the above mentioned examples. After we identiﬁed Ancylanthos rubiginosus, Hutchinsonia barbata, Rytigynia beniensis, R. decussata, and R. eickii as strongly incongruent, these taxa were removed from further analyses (results only shown for the combined data set with incongruent taxa removed; Fig. 4). Comparing the results from the combined analyses with incongruent taxa included or excluded ( Fig. 1 and Fig. 4 resp.), little was changed in the phylogeny for the Vangueria group, although the support for the node separating the clade consisting of Tapiphyllum obtusifolium and T. velutinum from the other taxa in the group was increased from 62 to 85 and the support for the Vangueria group from 91 to 99. In the Fadogia-Rytigynia group the b Fig. 2. Majority rule consensus tree from the bootstrap analysis of the ITS data set. Bootstrap support ‡ 50 is included above nodes. Letters in the tree refer to clades discussed in the text H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera 72 69 76 96 89 78 100 96 71 64 86 57 62 92 54 57 87 100 73 100 95 87 69 87 100 76 98 67 73 98 67 99 100 90 82 100 62 99 59 Canthium coromandelicum Canthium glaucum Canthium tetraphyllum Canthium armatum Pygmaeothamnus cf. chamaedendrum Canthium inerme Canthium bugoyense Canthium stenosepalum Ancylanthos rubiginosus Fadogia agrestis Pachystigma gillettii Lagynias dryadum Lagynias lasiantha Lagynias monteiroi Lagynias pallidiflora Pachystigma latifolium Pachystigma pygmaeum Tapiphyllum cinerascens Vangueria apiculata Vangueria esculenta Vangueria infausta Vangueria madagascariensis Vangueria parvifolia Vangueria proschii Vangueria randii Vangueria soutpansbergensis Vangueria volkensii Pachystigma loranthifolium Pachystigma schumannianum Rytigynia fuscosetulosa Rytigynia induta Tapiphyllum obtusifolium Tapiphyllum velutinum Vangueria praecox Vangueriopsis cf. longiflora Pygmaeothamnus zeyheri Robynsia glabrata Multidentia concrescens Multidentia crassa Multidentia fanshawei Multidentia sclerocarpa Canthium oligocarpum Cuviera angolensis Fadogia ancylantha Fadogia triphylla Fadogia verdcourtii Rytigynia beniensis Fadogia arenicola Fadogia cienkowskii Fadogia elskensii Fadogia stenophylla Fadogia tetraquetra Fadogiella stigmatoloba Rytigynia adenodonta Rytigynia celastroides Rytigynia parvifolia Rytigynia lichenoxenos Rytigynia sp. L Rytigynia uhligii Rytigynia neglecta Rytigynia senegalensis Rytigynia umbellulata Hutchinsonia barbata Rytigynia bagshawei Rytigynia hirsutiflora Rytigynia longicaudata Rytigynia monantha Rytigynia xanthotricha Rytigynia decussata Rytigynia eickii Rytigynia longipedicellata Rytigynia sambavensis Rytigynia seyrigii Outgroup Vangueria group Fadogia-Rytigynia group H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera changes in support and resolution were more signiﬁcant to the point of making comparisons diﬃcult. However, a clade of Rytigynia species where all but R. seyrigii have long appendages on the corolla lobes (clade C; Figs. 1 and 4) is supported in both phylogenies, but the support is lowered from 95 to 57 when the incongruent taxa are removed. The large clade in which Hutchinsonia is included (clade D; Fig. 1) has weak support (63), but this is increased to 95 (clade D; Fig. 4) when the incongruent taxa are removed, although these two clades are not identical and the results are possibly inﬂuenced not only by incongruence but also by sampling eﬀects. Comparing the phylogenies based on the individual data sets with and without incongruent taxa (results only shown for phylogenies with incongruent taxa included; Figs. 2 and 3), only minor changes in support occur by the removal of taxa for nodes at which the incongruent taxa were not attached, and comparisons of support can equally well be done with the incongruent taxa included (Figs. 2 and 3). The Vangueria group as circumscribed by Lantz and Bremer (2004) is supported in the ITS and combined phylogenies but not in the chloroplast phylogeny due to the incongruent position of Ancylanthos rubiginosus. The Fadogia-Rytigynia group is strongly supported in all trees. In the chloroplast phylogeny (Fig. 3 ), the clade is sister to a clade consisting of all other taxa, and this result is also favoured by the combined analyses, but is unresolved in the ITS phylogeny ( Fig. 2 ). Rytigynia and Fadogia are polyphyletic in all four phylogenies and there are indications that this is true also for Pachystigma, Tapiphyllum and Vangueria. Pygmaeothamnus is present both in the outgroup and in the ingroup and is thus also polyphyletic, but this has been discussed elsewhere (Lantz and Bremer 2004) and will not be investigated in detail here. Within the Fadogia-Rytigynia group the situation is complex. The taxa with incongruent positions and the low supported or unresolved nodes hamper our ability to draw conclusions, but Fadogia and Rytigynia are not supported as monophyletic in any of the trees. Multidentia is the only genus which is monophyletic in all four phylogenies, although with weak support in the ITS phylogeny. Discussion We will divide the discussion into two major parts. In the ﬁrst part we discuss the results of the phylogenetic analyses. Morphological characters of importance for the delimitation of the clades and genera are also discussed, but we restrict the discussion to characters found to be informative in Lantz and Bremer (2004). In the second part we try to explain the incongruencies noticed in the comparisons between the individual analyses. We base our discussion on the phylogeny in which the strongly incongruent taxa have been removed (Fig. 4), unless otherwise noted. This phylogeny exhibits increased support and resolution in comparison to the analyses based on the individual data sets. Although the taxa with most strongly incongruent positions have been removed, it is possible that the phylogeny still is inﬂuenced by incongruence. Indeed, in two cases did the combination of data sets with taxa removed result in signiﬁcantly lowered bootstrap support for a node, indicating conﬂicting phylogenetic signals. With this in mind, we refrain from making changes to the classiﬁcation in situations not unequivocally supported by all data sets, unless support from morphology can corroborate any of the phylogenies. We often choose to use the term b Fig. 3. Majority rule consensus tree from the bootstrap analysis of the chloroplast (trnT-F + rpl16) data set. Bootstrap support ‡ 50 is included above nodes. Letters in the tree refer to clades discussed in the text. Ancylanthos rubiginosus is supported as a member of the Vangueria group in the ITS and combined trees (Figs. 1, 2, and 4) H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera 65 86 93 82 67 B 97 100 60 97 79 A 85 76 98 95 78 99 78 52 100 100 98 84 100 87 100 78 72 99 95 D 71 100 51 96 G 100 83 93 75 57 C 94 Canthium coromandelicum Canthium glaucum Canthium tetraphyllum Canthium armatum Pygmaeothamnus cf. chamaedendrum Canthium inerme Canthium bugoyense Canthium stenosepalum Canthium oligocarpum Cuviera angolensis Fadogia agrestis Pachystigma gillettii Pachystigma loranthifolium Pachystigma schumannianum Lagynias dryadum Lagynias lasiantha Lagynias monteiroi Lagynias pallidiflora Pachystigma latifolium Pachystigma pygmaeum Tapiphyllum cinerascens Vangueria apiculata Vangueria volkensii Vangueria esculenta Vangueria infausta Vangueria proschii Vangueria madagascariensis Vangueria parvifolia Vangueria soutpansbergensis Vangueria randii Rytigynia fuscosetulosa Rytigynia induta Vangueria praecox Tapiphyllum obtusifolium Tapiphyllum velutinum Vangueriopsis cf. longiflora Multidentia concrescens Multidentia crassa Multidentia fanshawei Multidentia sclerocarpa Pygmaeothamnus zeyheri Robynsia glabrata Fadogia ancylantha Fadogia triphylla Fadogia verdcourtii Fadogia arenicola Fadogiella stigmatoloba Fadogia cienkowskii Fadogia elskensii Fadogia stenophylla Fadogia tetraquetra Rytigynia adenodonta Rytigynia celastroides Rytigynia parvifolia Rytigynia lichenoxenos Rytigynia sp. L Rytigynia uhligii Rytigynia neglecta Rytigynia umbellulata Rytigynia senegalensis Rytigynia bagshawei Rytigynia hirsutiflora Rytigynia longicaudata Rytigynia monantha Rytigynia xanthotricha Rytigynia pseudolongicaudata Rytigynia sambavensis Rytigynia seyrigii Rytigynia longipedicellata Outgroup Vangueria group Fadogia-Rytigynia group H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera ‘incongruent taxon/taxa’ rather than the bulkier but more correct ‘taxon/taxa with incongruent positions’ fully aware of the fact that a taxon (when represented by a single terminal) cannot be incongruent in itself, but prefer the shorter, less cumbersome term. The Vangueria group. This group was ﬁrst identiﬁed in Lantz et al. (2002) and also investigated in Lantz and Bremer (2004). Lagynias, Pachystigma, Tapiphyllum, Vangueria, and one species of Rytigynia, R. fuscosetulosa, were identiﬁed as belonging to this clade, and also Ancylanthos rubiginosus, but with some hesitation due to the incongruent positions for this taxon supported by the plastid and nuclear data sets. All of these genera have plurilocular ovaries, a character traditionally used to distinguish between Vangueria and the bilocular genus Canthium, the two ﬁrst described genera in the tribe. The group is identical to ‘Gattungskomplex Vangueria’ suggested by Igersheim (1989) with the exception of R. fuscosetulosa. In the present study, the ITS and chloroplast phylogenies support a similar circumscription of the Vangueria group, with the exception of Ancylanthos rubiginosus. Of the two possible positions for A. rubiginosus suggested by the ITS and chloroplast phylogenies, the internal position in the Vangueria group seen in the ITS phylogeny is better supported by morphology. Also in the Vangueria group are Pachystigma pygmaeum and Tapiphyllum cinerascens. All three are geofrutices and sometimes have ternate leaves, and both A. rubiginosus and T. cinerascens have short, stiﬀ external indumentum on the corollas. Geofrutescent habit and ternate leaves are otherwise rare features in the Vangueria group, but external indumentum on the corollas is more common (e.g. Bridson 1998). The inclusion of Ancylanthos in the Vangueria group is thus supported by morphology and by the ITS phylogeny. To conclude, we consider the Vangueria group to comprise the genera Ancylanthos, Lagynias, Pachystigma, Tapiphyllum, Vangueria, and the three species Fadogia agrestis, Rytigynia fuscosetulosa, and R. induta. Nine of the included Vangueria species, all in subgenus Vangueria, form a moderately supported clade (A; Fig. 4) together with four species of Lagynias, Pachystigma latifolium, P. pygmaeum, and Tapiphyllum cinerascens. A tenth species, V. praecox, tentatively included in Vangueria and the single member of subgenus Itigi (Verdcourt 1981), instead forms a clade with Rytigynia fuscosetulosa and R. induta at a more basal position in the Vangueria group. Vangueria subgenus Vangueria has traditionally been recognized by a combination of large leaves and lax manyﬂowered inﬂorescences, but is in the most recent circumscription (Verdcourt 1981, Bridson 1998) morphologically heterogeneous. Vangueria parvifolia and V. soutpansbergensis both have small leaves and few-ﬂowered, shortly pedunculate and dense inﬂorescences and form a moderately supported clade in the combined phylogeny (Fig. 4). Vangueria parvifolia was earlier included in Tapiphyllum but was found to be atypical of that genus and was transferred (Bridson 1998) shortly after the morphologically similar South African endemic V. soutpansbergensis was described (Hahn 1997). Two other clades formed by the Vangueria species are well delimited by type of calyx lobes, a character earlier suggested to be of importance in the classiﬁcation of Vangueria (Bridson 1998). The type species V. madagascariensis forms a strongly supported clade together with V. esculenta, V. infausta, and V. proschii, all of which have triangular to narrowly oblong calyx lobes. Vangueria apiculata and V. volkensii, both with narrowly oblong to linear calyx lobes, group together with strong support. Surprisingly V. randii, a b Fig. 4. Majority rule consensus tree from the bootstrap analysis of the combined data set (ITS + trnT-F + rpl16) with putative hybrids removed. Bootstrap support ‡ 50 is included above nodes. Letters in the tree refer to clades discussed in the text H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera third species with similar elongate calyx lobes, does not group with these species but has an unresolved position in the same clade. Lagynias is a genus of ﬁve species distributed in eastern and southern Africa characterized by well-developed often oblong or narrowly obovate calyx lobes and usually reﬂexed corolla lobes in the mature ﬂower (Bullock 1931). Two species of Lagynias, L. dryadum and L. lasiantha, form a strongly supported group, but Lagynias as a genus is not supported as monophyletic. Tapiphyllum cinerascens is a highly variable species (Verdcourt and Bridson 1991) which in the phylogeny (Fig. 4) is distanced from the two other included species of Tapiphyllum. The genus is considered ill-deﬁned (Bridson 1998) and is diagnosed principally by a thick indumentum, dense inﬂorescences, and mostly elongate calyx lobes. Sister to the clade including most of the Vangueria species (clade A; Fig. 4) is a weakly supported clade (B; Fig. 4) consisting of three species of Pachystigma and one representative of Fadogia, F. agrestis. Pachystigma is thus represented in two clades, a division supported by type of habit. Pachystigma latifolium and P. pygmaeum are both geofrutices while the three other species are shrubs or small trees. Fadogia agrestis is a West African species obviously misplaced in Fadogia, diﬀering from the species of Fadogia present in the FadogiaRytigynia group by its large ovate to triangular calyx lobes and consistently paired leaves. As shown by the phylogeny and also supported by morphology, Pachystigma contains smaller groups of closely related species, but is as a genus united by little but an absence of the characters used to delimit Tapiphyllum and Vangueria. The problematic circumscription of the genus has also been discussed elsewhere (e.g. Verdcourt 1981, Bridson 1996). The already mentioned clade of Rytigynia fuscosetulosa, R. induta, and Vangueria praecox is interesting in that all three species were tentatively included in their genera, but a close relationship between the three has never been proposed. Rytigynia fuscosetulosa is a member of the small (two species) subgenus Sali (Verdcourt 1987, Verdcourt and Bridson 1991) and was when described mentioned as intermediate between several genera and with pyrenes similar to Vangueria (Verdcourt 1981). Rytigynia induta has been suggested to be misplaced in the genus (D. Bridson, pers. comm.) and Vangueria praecox was tentatively included in Vangueria (Verdcourt 1981). All three species have few-ﬂowered and shortly pedunculate inﬂorescences, but this is not unknown from other species in the Vangueria group (e.g. Vangueria parvifolia and V. soutpansbergensis) and we can ﬁnd no unique morphological synapomorphy for the clade. Two species of Tapiphyllum form a strongly supported clade as sister to the other taxa of the Vangueria group. Tapiphyllum obtusifolium is restricted to Tanzania and T. velutinum has a Zambesian distribution. These species are both shrubs or small trees in contrast to the type species of Tapiphyllum, the geofrutex T. cinerascens (clade B; Fig. 4). They otherwise share a number of features with T. cinerascens such as a thick indumentum on the leaves, dense inﬂorescences, and elongate often linear calyx lobes, and would based on morphology be suspected to be more closely related. None of the genera in the Vangueria group are supported as monophyletic, and indeed, the weak internal nodes in the clade do not either enable us to tell with certainty if the genera are paraphyletic or polyphyletic. However, there are indications that Pachystigma, Tapiphyllum, and Vangueria are not monophyletic. Moreover, there is little morphological support for the current circumscriptions of the genera. The genera are not homogeneous in their current circumscriptions. All genera include groups of species clearly closely related but there are also a number of species often tentatively included and clearly dissimilar from the other species. These are also quite often intermediate in generic characters. The here included Vangueria parvifolia and V. soutpansbergensis, or Lagynias monteiroi, are examples of such species but there are more we have not H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera been able to include such as Tapiphyllum schliebenii Verdc., a species of Tapiphyllum with lax inﬂorescences. With the addition of two species currently included in Rytigynia and one in Fadogia, as discussed earlier, the generic characters of the genera in the Vangueria group break down even further. The problematic circumscription of these genera has also been noted by other authors. Verdcourt (1981) wrote (about Pachystigma) ‘There are no constant technical characters to distinguish it from Vangueria save facies.’. He goes on to state that ‘Tapiphyllum is no more than a velvety Pachystigma and some South African species are certainly intermediate in character…’. Bridson (1996) noted in the discussion on Ancylanthos monteiroi Oliv. (now Lagynias monteiroi (Oliv.) Bridson) that ‘The distinction between Pachystigma and Lagynias is in need of clariﬁcation…’. We agree with these views and consider a new classiﬁcation for the genera in the Vangueria group to be desirable. The best solution is to consider the whole Vangueria group as one genus; a classiﬁcation that receives support both from the phylogeny and from morphological characters. Vangueria is the oldest name in the group (Jussieu 1789), and would have priority. With the exception of Ancylanthos, this clade is present in all phylogenies. An alternative solution would be to recognize the smaller clades present (i.e. A, B etc; Fig. 4 ) as genera, but these clades are likely to change with the addition of new species and we are also unable to ﬁnd morphological support for any of the clades. Vangueria in this new sense is recognized by a combination of the following characters: domatia rarely present, inﬂorescences usually borne at nodes from which the leaves have fallen, bracteoles small (present on secondary branches, usually not more than 1 mm in length), smooth retrorse hairs in the corolla (Lantz and Bremer 2004), and large fruits (above 1 cm in length) with 3 to 5 locules. This combination of characters is not known from any other currently recognised genus in the tribe and will greatly facilitate the identiﬁcation of Vanguerieae genera. Below, we make combinations at the genus level and also at the species level for all currently recognised species. The Fadogia-Rytigynia group. This group was identiﬁed by Lantz et al. (2002), and is identical in circumscription to ‘Gattungskomplex Rytigynia’ suggested by Igersheim (1989) on morphological grounds. For this group there exists no conﬂict as regards the circumscription, it is strongly supported in all phylogenies (Figs. 1, 2, 3, and 4). Internally however, the phylogeny is strongly aﬀected by incongruence between the chloroplast and ITS data sets as enumerated above. The great extent of these incongruencies greatly hinders our ability to draw conclusions about the phylogeny of the group. We can, however, make comparisons with morphology and see to which extent the phylogenies support the current classiﬁcation. As a general rule, the Fadogia-Rytigynia group can be distinguished from the Vangueria group or any of the other genera in the large-ﬂowered clade by a presence of domatia and a calyx without or with poorly developed calyx lobes, although this rule is not without exceptions. None of the included genera are supported as monophyletic in any of the phylogenies, but the small genera Fadogiella (three species) and Hutchinsonia (two species) are both sampled by one species each and we are thus unable to test the monophyly of these genera. Fadogiella stigmatoloba was in an earlier analysis with a smaller sample (Lantz et al. 2002) nested within Fadogia, but this result is not corroborated here. It groups with Fadogia arenicola with weak support in the ITS and combined analysis, a position supported by the thick indumentum of both species. Hutchinsonia is morphologically close to Rytigynia and is mainly distinguished by its long and slender corolla tubes. Fadogia and Rytigynia are the two largest genera in the group, both comprising at least 50 species. Fadogia is diagnosed by a geofrutescent habit and leaves at least sometimes in whorls of three (or more). Rytigynia is much harder to diagnose and is recognized mainly by a shrubby habit and H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera absence of the characters typical of the genera in the Vangueria group or Fadogia. As shown here, and earlier (Lantz et al. 2002, Lantz and Bremer 2004), Rytigynia is polyphyletic with species included also in the spiny group (the present outgroup) and in the Vangueria-group. The majority of the currently recognized species in Rytigynia belong in the FadogiaRytigynia group and we consider these to constitute Rytigynia s. str. The characters most useful to distinguish between Rytigynia s.str. and the species incorrectly included in Rytigynia (such as the here sampled R. fuscosetulosa and R. induta) is the development of the calyx lobes and the size of the fruits. Rytigynia s. str. has a truncate or very shortly toothed calyx and fruits rarely more than 10 mm in diameter. Rytigynia induta has short calyx lobes and is in many ways morphologically similar to Rytigynia s. str. but has much larger fruits (20 mm in diameter compared to usually below 10 mm). R. fuscosetulosa has elongate calyx lobes. As often is the case in Vanguerieae, even distantly related species can be closely similar morphologically and it is only with the aid of a phylogeny the distinguishing characters can be found. A clade consisting of Fadogia ancylantha, F. triphylla, and F. verdcourtii is consistently recovered in all phylogenies although with the addition of Rytigynia beniensis in the chloroplast phylogeny. All three Fadogia species have corollas of 5 mm or more and have in contrast to the other included species of Fadogia sometimes paired leaves, especially on the younger parts of the branches. This character has not been considered of importance earlier, but is strongly supported here. Fadogiella stigmatoloba is the only geofrutex sampled here which otherwise has both paired or whorled leaves. Rytigynia beniensis does not share any of these characters or any other characters used to delimit Fadogia. The ITS phylogeny groups R. beniensis with several species of Rytigynia in a weakly supported clade (clade G; Fig. 1 ), a position much better supported by morphology. Also included in the same clade is R. celastroides, a species which shares the rare character branches sometimes in whorls of three with R. beniensis. However, R. umbellulata, considered very close to R. beniensis (Verdcourt 1987), instead forms a moderately supported clade with Fadogia elskensii and F. tetraquetra in the same phylogeny. The two Fadogia species are considered close (Verdcourt 1981) but the relationship with R. umbellulata is quite unexpected. In the combined analysis (Fig. 4) F. tetraquetra groups with F. stenophylla with strong support, F. elskensii forms a moderately supported clade with F. cienkowskii, and R. umbellulata has unresolved position in a moderately supported clade. Fadogia cienkowskii and F. stenophylla were, however, not sequenced for ITS. The clade in which R. beniensis was supported to belong by the ITS phylogeny receives strong support in both of the combined analyses (clade G; Figs. 1 and 4), and the support is strongest when the incongruent taxa are not included (Fig. 4). None of the taxa in the clade have long appendages (over 2 mm) on the corolla lobes, a character otherwise common in another clade (C; Fig. 4). Appendages are also missing from R. longipedicellata, R. neglecta, R. senegalensis, R. umbellulata, and two of the taxa identiﬁed as strongly incongruent, R. decussata and R. eickii. Rytigynia longipedicellata is morphologically similar to R. eickii, and forms a strongly supported clade with this species in the chloroplast phylogeny (Fig. 3), but was not sequenced for ITS and the position in the combined phylogeny (Fig. 4) as sister to the clade with long appendages is thus only based on chloroplast data. Hutchinsonia barbata also has appendages on the corolla lobes and the position indicated by the ITS phylogeny ( Fig. 2 ) is thus better supported by morphology. The two Malagasy specimens R. sambavensis and R. seyrigii group together in all phylogenies, but only with weak support in the chloroplast phylogeny. The clade is strongly supported as sister to the other taxa in the appendaged clade by the combined phylogeny with incongruent taxa included (Fig. 1), but this support is greatly decreased when the incongruent taxa H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera are removed (Fig. 4). In the combined analysis with incongruent taxa removed, the FadogiaRytigynia group is divided into two clades, one of which is strongly supported (clade D; Fig. 4. Apart from the partial congruence with presence or absence of appendages on the corolla lobes, we see little support for this division from morphology. We can thus ﬁnd morphological support for clades present in both the chloroplast and ITS phylogenies and none of the two phylogenies agree with morphology better than the other. The widespread incongruence for these genera puts all phylogenies in some doubt, and we are not prepared to make any changes to the classiﬁcation. The phylogenies do not enable us to present any classiﬁcation which would result in monophyletic and unequivocally supported genera. In addition, several of the strongly supported clades lack morphological support. The inclusion of the whole clade in one genus would result in a monophyletic but morphologically heterogeneous genus. At this point, when a number of the currently recognized species are poorly known, especially those occurring in West and Central Africa, we prefer to stress the morphological and geographical variation present within the group by retaining the genera rather than hide it within one very large genus. Other genera in the large-flowered group. In the chloroplast and combined phylogenies, Canthium oligocarpum, Cuviera, Multidentia, Pygmaeothamnus zeyheri, Robynsia, and Vangueriopsis form a weakly to moderately supported clade together with the Vangueria group as sister to the Fadogia-Rytigynia group. Their relationships to the Vangueria group or to each other are however weakly supported. The only relationship which receives support here is the moderately supported clade formed by Cuviera and Robynsia in the ITS phylogeny (clade F, Fig. 2 ). Cuviera is a morphologically distinct genus centred in West Africa but with some few representatives also in central and East Africa. Diagnostic for the genus are large and many-ﬂowered inﬂorescences with linear to lanceolate or even elliptic bracts and bracteoles, and many species also have hairs on the styles, sometimes in combination with peculiar swellings (Hallé 1959). The calyx lobes are also often large and foliose, usually similar to the bracts and bracteoles. Today the genus includes around 20 species (Bridson 1998). The monotypic genus Robynsia (Hutchinson 1931) has very similar inﬂorescences to Cuviera, but has long and slender corolla tubes not known in Cuviera. A relationship between Cuviera and Robynsia is thus supported by morphology, but the clade collapses in the combined phylogenies, possibly due to incongruence. Canthium oligocarpum is a member of the small (three species) subgenus Lycioserissa. It is clearly distanced both from the type species of Canthium, C. coromandelicum and the type species of subgenus Lycioserissa, C. inerme, both of which are represented in the outgroup, but is of uncertain position in the large-ﬂowered group. It is similar in leaf morphology to some species of Multidentia (e.g. M. sclerocarpa), and is in the ITS phylogeny ( Fig. 2 ) weakly supported as sister to that genus, but lacks the morphological synapomorphies of Multidentia and should no be transferred until its phylogenetic position can be ascertained with better support. Multidentia is one of the best delimited genera in Vanguerieae (Bridson 1987). Eleven species are known, and the genus has a wide distribution in tropical Africa. The genus is easily recognized by a combination of vague tertiary nerves, a well-developed calyx limb tube, and thickly woody pyrenes. The sample here includes most of the morphological variation in the genus and the monophyly of Multidentia is strongly supported. Pygmaeothamnus is a genus of two geofrutices of which P. chamaedendrum is restricted to South Africa and P. zeyheri has a wide essentially Zambesian distribution. A combination of a geofrutescent habit and bilocular ovaries is the major morphological character used to delimit the genus. The genus is polyphyletic and is represented both in the outgroup and in the large-ﬂowered group. H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Vangueriopsis includes four species and is recognized by long (over 15 mm) and linear corolla lobes, long exserted stamens, and hairs usually present on stamens. The specimen here sampled is a fruiting specimen tentatively identiﬁed to Vangueriopsis longiﬂora (see also Lantz and Bremer 2004) but is without doubt a member of the genus. Even though not really tested here, we ﬁnd it highly unlikely that the Vangueria group is nested within any of these genera. The monophyly of Multidentia is strongly demonstrated here (Fig. 4), Cuviera and Vangueriopsis are strongly delimited in a morphological sense, Robynsia is monotypic and Pygmaeothamnus is known to be represented only by the type species P. zeyheri in the large-ﬂowered group. The phylogenetic aﬃnities of Canthium oligocarpum remain unknown and, as was concluded also in an earlier study (Lantz and Bremer 2004), the species cannot remain in Canthium but neither morphology nor phylogeny enables us to suggest how to deal with the species. Incongruencies. As discussed above, the ITS ( Fig. 2) and chloroplast (Fig. 3) trees are partly incongruent. Phylogenetic incongruence can be caused by a number of phenomena, some reﬂecting biological processes such as hybridization or lineage sorting, but there are also technical causes such as taxon sampling or insuﬃcient data which can lead to incongruence (Wendel and Doyle 1998). A common case of incongruence is when weakly supported clades contradict each other. An example of this is the position of Rytigynia induta. Such weakly supported results are not necessarily reﬂecting the evolutionary history of the taxa studied, but can be the result of homoplasy (e.g. Kellogg et al. 1996) and are therefore usually not taken into account in a discussion on phylogeny. However, there are also some strongly supported incongruencies in this study, and these cannot easily be dismissed as artefacts caused by technical phenomena. In the Vangueria group we are able to identify a single taxon, Ancylanthos rubiginosus, as incongruent in the comparison between the ITS and chloroplast phylogenies. Based on comparisons with morphology (see above) the internal position in the Vangueria group suggested by the ITS phylogeny is the most likely for the organism. The incongruent position of Ancylanthos rubiginosus in the chloroplast phylogeny in comparison to the position supported by both the nuclear phylogeny and morphology can be explained by a chloroplast capture event (e.g. Soltis et al. 1996). In chloroplast capture, hybridization and introgression results in a cytoplasmic transfer with little or no transfer of nuclear DNA (Rieseberg and Soltis 1991). An additional factor that potentially could explain the incongruence is the diﬀerent ITS paralogues known to exist for this species (Lantz and Bremer 2004). As these paralogues do not share the same evolutionary history, their presence can result in erroneous phylogenies (Sanderson and Doyle 1992, Buckler et al. 1997). However, it was concluded that the variation did not pass the species boundary (Lantz and Bremer 2004; see also Razaﬁmandimbison et al. 2004 for a similar situation). It is therefore unlikely that it is the presence of diﬀerent ITS paralogues that are causing the incongruence. The incongruence in the Fadogia-Rytigynia group is more complex, especially if comparisons with morphological characters also are made. Four taxa, i.e. Hutchinsonia barbata, Rytigynia beniensis, R. decussata, and R. eickii have strongly supported incongruent positions in the ITS ( Fig. 2) and chloroplast (Fig. 3) phylogenies. There are also strongly supported nodes present in the individual phylogenies which collapse when the data sets are combined. Morphological support exists for some of the strongly supported clades, but not all. In the Fadogia-Rytigynia group, we cannot ﬁnd a single plausible explanation for the incongruence as we could in the Vangueria group, but we ﬁnd it most likely that the incongruence has a similar origin; hybridization and introgression. Chloroplast capture is also in this group a possibility, and there is also an additional H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera phenomenon known to exist for repetitive DNA such as ITS. The ITS region exists in a multitude of copies, and concerted evolution operates on these copies to homogenize them (e.g. Baldwin 1992). Following a hybridization it is likely that the hybrids have a mix of the paralogues present in the parent species, and if the concerted evolution homogenizes the ITS region to diﬀerent paralogues in diﬀerent hybrid lineages, phylogenies based on these paralogous ITS sequences are likely to give misleading inferences of the organismal phylogenies. This has been shown to occur in Gossypium (Wendel et al. 1995). Concerted evolution can also cause DNA to be exchanged between the paralogues, creating even more complex reticulate relationships between the paralogues (Hillis et al. 1991). A process which does not include hybridization but is known to cause incongruence is lineage sorting (Pamilo and Nei 1988, Wendel and Doyle 1998). In this process an ancestral polymorphism fails to survive a speciation event, so that not all of the descendant species share the polymorphism. As a result of this, alleles (or paralogues in the case of ITS) from diﬀerent species may be more closely related to each other than alleles within the same species. A study including also other parts of the tribe (Lantz and Bremer 2004) using ITS and trnT-F data did not ﬁnd any incongruencies outside of the large-ﬂowered group. An interesting correlation between phylogeny and morphology is that most of the genera in the large-ﬂowered group are known to have a special kind of pollen presenter (type 2; Igersheim 1993). Genera in the large-ﬂowered group known to have this type are Ancylanthos, Fadogia, Fadogiella, Lagynias, Pachystigma, Tapiphyllum, Vangueria, and part of Rytigynia (Igersheim 1989, Bridson 1996). Before the ﬂower bud opens, when the pollen is deposited on the presenter, the stigmatic lobes are exposed but covered by sterile apical appendages present on the anthers (Igersheim 1993). This is in contrast to type 1 (Igersheim 1993), where the stigmatic lobes are tightly pressed together and do not open until the ﬂower has opened and the style is fully elongated. This correlation between exposed stigmatic lobes and the presence of incongruence for these genera raises the question whether the exposed stigmatic lobes somehow might facilitate hybridization. We cannot see why this would be the case unless the type 2 pollen presenter is also coupled with weak barriers to hybridization. No studies on hybridization in Vanguerieae have been performed, so we can only guess that this could be the case. However, it is more likely that the exposed stigmatic lobes could increase the possibility for self-fertilization. The pollen sacs of the anthers and the stigmatic surfaces are spatially separated (Igersheim 1993), but the pollen sacs are nevertheless close and if the sterile apical appendages of the anthers fail to cover the stigmatic surfaces completely, it is possible that pollen could be deposited there in the bud stage. Selﬁng has been suggested to promote establishment of hybrid species (see e.g. discussion in Rieseberg 1997) and this could in part explain why we see possible cases of hybridization in the large-ﬂowered group and not in other parts of Vanguerieae. At present, this is little more than speculation, since it is not known whether any Vanguerieae species are self-fertilizing or whether they are self-compatible or not. We can thus conclude that the Vangueria group, as circumscribed here, should be given generic rank (as Vangueria). In the FadogiaRytigynia group, the presence of several taxa with incongruent positions complicates any conclusions about the phylogeny. Fadogia (including Fadogiella) is morphologically distinct, and should not be combined with Rytigynia before the reasons behind the incongruencies are known. We consider diagnosability to be a major desirable feature of classiﬁcation, second only to the criterion of monophyly and would prefer to describe morphologically well-delimited genera. The phylogenies present some possible answers to why the genera and species of the Fadogia-Rytigynia group have proven to be diﬃcult to delimit using morphological criteria (Verdcourt 1987). H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Hybridization can lead to intermediate morphologies (McDade 1990) and this could explain why these species and genera are hard to delimit morphologically. Surprisingly enough, the possibility for hybridization in Vanguerieae has rarely been discussed. We know of only a single report of a possible hybrid in Vanguerieae (between Vangueria apiculata and V. madagascariensis; Verdcourt 1981). However, the hybridization events may be old and the species might not hybridize today. Nevertheless, studies on self-compatibility and hybridization in Vanguerieae would be a most welcome addition to the complex issue of the evolution of Vanguerieae. Taxonomy of Vangueria Vangueria Juss., Gen. pl.: 206. 1789. Type: V. madagascariensis Gmelin in Syst. Nat., ed. 13, 2: 367. 1791. Ancylanthos Desf., Mém. Mus. Hist. Nat. 4: 5. 1818, syn. nov. Type: A. rubiginosus Desf. Pachystigma Hochst., Flora 25: 234. 1842, syn. nov. Type: P. venosum Hochst. Tapiphyllum Robyns, Bull. Jard. Bot. État. 11: 101. 1928, syn. nov. Type: T. cinerascens (Hiern) Robyns. Lagynias E. Mey. ex Robyns, Bull. Jard. Bot. État. 11: 312. 1928, syn. nov. Type: L. lasiantha (Sond.) Bullock. Lagynias E. Mey. ex Robyns subgen. Bembea Verdc. in Kew Bull. 42: 141. 1987, syn. nov. Type: L. rufescens (E.A. Bruce) Verdc. Vangueria Juss. subgen. Itigi Verdc. in Kew Bull. 42: 186. 1987, syn. nov. Type: V. praecox Verdc. Small trees, shrubs, or geofrutices. Unarmed, or very rarely with paired spines. Leaves usually deciduous, paired or (for geofrutices) in whorls of three, glabrous to thickly pubescent; domatia absent, or very rarely present; stipules triangular or with a broad base and a narrow apex, hairs present within. Flowers 5(6)-merous in axillary pedunculate or rarely fasciculate inﬂorescences, secondary axes well-developed or not, with 1–30 ﬂowers; bracts and bracteoles inconspicuous. Calyx glabrous to densely pubescent, limb tube short, not equalling disk or more developed exceeding the disk; lobes variable but always well developed, triangular, oblong, lanceolate, spatulate, or linear. Corolla usually white to green, less often yellow or orange, externally glabrous to densely pubescent; tube cylindrical to campanulate, with diﬀusely spreading hairs at throat and retrorse straight hairs in a well deﬁned ring inside, usually clearly distanced from hairs present at throat; lobes equalling tube or shorter, rarely longer, often apiculate or shortly appendaged. Stamens inserted at throat; anthers ovate or oblong, apiculate, usually shortly exserted. Ovary 3-5-locular, each locule with a single pendulous ovule; style shortly exserted or sometimes long exserted; pollen presenter cylindrical or sometimes coroniform. Fruit large, above 1 cm in length, only rarely below 1 cm, subglobose or globose, glabrous or pubescent, calyx lobes often persistent; 3–5 pyrenes, or rarely less due to faulty development. A genus of over 50 species distributed in Africa south of the Sahara with one species occurring in Madagascar (V. madagascariensis). The centre of diversity is in East Africa (Kenya, Tanzania) and it is rare in West Africa. Robyns (1928) revised all of the then recognized species and later also described additional species, especially in Tapiphyllum (Robyns 1931, 1962). The East African and Zambesian species were revised in conjunction with the production of the Vanguerieae volumes for Flora of Tropical East Africa and Flora Zambesiaca (Verdcourt 1981, 1987; Verdcourt and Bridson 1991; Bridson 1996, 1998), but the other species have not recently been revised. Several of these taxa are poorly known, especially the ones from Angola and Western/Central Africa, and are possibly not worth maintaining. Unless suggested by other authors (see V. infausta, V. macrocalyx, and V. venosa), we do not address the delimitation of any of the included species here, but it should be noted that further studies are H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera certainly needed. The genus is in this circumscription made more variable in comparison to how it earlier was conceived, especially concerning inﬂorescence and ﬂoral morphology. However, the genus is now easily distinguished from its closest relatives by the following combination of characters (also listed above): domatia rarely present, inﬂorescences usually borne at nodes from which the leaves have fallen, bracteoles small (present on secondary branches, usually not more than 1 mm in length), smooth retrorse hairs in the corolla (Lantz and Bremer 2004), and large fruits (above 1 cm in length) with 3 to 5 locules. Here follows a list of all currently recognized species of Vangueria in the circumscription suggested above. New combinations are made when necessary, but infraspeciﬁc names are given only for taxa here transferred to Vangueria. Basionyms and most recently accepted names are listed, full synonyms are not included. Vangueria apiculata K. Schum. in Pﬂanzenw. Ost-Afrikas, C: 384. 1895. Vangueria agrestis (Schweinf. ex Hiern) Lantz, comb. nov. Fadogia agrestis Schweinf. ex Hiern in Fl. trop. Afr. 3: 154. 1877. Vangueria albosetulosa (Verdc.) Lantz, comb. nov. Pachystigma albosetulosum Verdc. in Kew Bull. 42: 139. 1987. Vangueria bicolor K. Schum. in Bot. Jahrb. Syst. 34: 332. 1904. Vangueria bowkeri (Robyns) Lantz, comb. nov. Pachystigma bowkeri Robyns in Bull. Jard. Bot. État. 11: 128. 1928. Vangueria burnettii (Tennant) Lantz, comb. nov. Tapiphyllum burnettii Tennant in Kew Bull. 19: 280, Fig. 1. 1965. Vangueria burttii (Verdc.) Lantz, comb. nov. Pachystigma burttii Verdc. in Kew Bull. 36: 545, Fig. 11. 1981. subsp. burttii. subsp. hirtiﬂora (Verdc.) Lantz, comb. nov. Pachystigma burttii subsp. hirtiﬂorum Verdc. in in Fl. Trop. E. Afr., Rubiaceae: 769. 1991. Vangueria chariensis A. Chev. ex Robyns in Bull. Jard. Bot. État. 11: 299. Vangueria cinerascens (Welw. ex Hiern) Lantz, comb. nov. Ancylanthos cinerascens Welw. ex Hiern in Fl. trop. Afr. 3: 159. 1877. Tapiphyllum cinerascens (Welw. ex Hiern) Robyns in Bull. Jard. Bot. État. 11: 107. 1928. var. cinerascens. var. inaequalis (Robyns) Lantz, comb. nov. Tapiphyllum inaequale Robyns in Bull. Jard. Bot. Brux. 32: 145. 1962. Tapiphyllum cinerascens var. inaequale (Robyns) Verdc. in Kew Bull. 42: 144. 1987. var. laeta (Robyns) Lantz, comb. nov. Tapiphyllum laetum Robyns in Bull. Jard. Bot. Brux. 32: 143. 1962. Tapiphyllum cinerascens var. laetum (Robyns) Verdc. in Kew Bull. 42: 145. 1987. var. laevior (K. Schum.) Lantz, comb. nov. Vangueria velutina Hiern var. laevior K. Schum. in Bot. Jahrb. Syst. 28: 494. 1900. Tapiphyllum cinerascens var. laevius (K. Schum.) Verdc. in Kew Bull. 42: 144. 1987. var. richardsii (Robyns) Lantz, comb. nov. Tapiphyllum richardsii Robyns in Bull. Jard. Bot. Brux. 32: 142. 1962. Tapiphyllum cinerascens var. richardsii (Robyns) Verdc. in Kew Bull. 42: 145. 1987. Vangueria cistifolia (Welw., ex Hiern) Lantz, comb. nov. Ancylanthus cistifolius Welw. ex Hiern in Fl. trop. Afr. 3: 159. 1877. Tapiphyllum cistifolium (Welw.) Robyns in Bull. Jard. Bot. État. 11: 108. 1928. var. cistifolia. var. latifolia (Verdc.) Lantz. Tapiphyllum cistifolium (Welw. ex Hiern) Robyns var. latifolium Verdc. in Fl. Zamb. 259. 1998. Vangueria coerulea (Robyns) Lantz, comb. nov. Pachystigma coeruleum Robyns in Bull. Jard. Bot. État. 11: 129. 1928. Vangueria cyanescens Robyns in Bull. Jard. Bot. État. 11: 284. 1928. H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Vangueria discolor (De Wild.) Lantz, comb. nov. Fadogia discolor De Wild. in Feddes Repert. 13: 138. 1914. Tapiphyllum discolor (De Wild.) Robyns in Bull. Jard. Bot. État. 11: 105. 1928. Vangueria dryadum S. Moore in J. Linn. Soc., Bot. 40: 93. 1911. Lagynias dryadum (S. Moore) Robyns in Bull. Jard. Bot. État. 11: 315. 1928. Vangueria esculenta S Moore in J. Linn. Soc., Bot. 40: 91. 1911. Vangueria fulva (Robyns) Lantz, comb. nov. Tapiphyllum fulvum Robyns in J. Bot. 69: 186. 1931. Vangueria fuscosetulosa (Verdc.) Lantz, comb. nov. Rytigynia fuscosetulosa Verdc. in Fl. Trop. E. Afr., Rubiaceae: 836. 1991. Vangueria gillettii (Tennant) Lantz, comb. nov. Rytigynia gillettii Tennant in Kew Bull. 19: 279. 1965. Pachystigma gillettii (Tennant) Verdc. in Kew Bull. 42: 140. 1987. Vangueria gossweileri (Robyns) Lantz, comb. nov. Tapiphyllum gossweileri Robyns in J. Bot. 69: 185. 1931. Vangueria induta (Bullock) Lantz, comb. nov. Rytigynia induta (Bullock) Verdc. & Bridson in Kew Bull. 42: 169. 1987. Canthium indutum Bullock in Kew Bull. 1932: 366. 1932. Vangueria infausta Burchell in Trav. S. Africa 2: 258. 1824. Vangueria velutina Hook. in Bot. Mag. 57, t. 3014. 1830, syn. nov. Vangueria velutina Hook. has been suggested earlier (Verdcourt 1981) to be a synonym of V. infausta Burchell, but has not been formally listed as such. From the excellent illustration in Hooker’s description with detail of the calyx lobes there can be no doubt that the specimen depicted indeed is the same taxon as Vangueria infausta Burchell. See also Vangueria zambesiaca Lantz below. Vangueria lasiantha (Sond.) Sond. in Fl. Cap. 3: 14. 1865. Pachystigma lasianthum Sond. in Linnaea 23: 55. 1850. Lagynias lasiantha (Sond.) Bullock in Bull. Misc. Inform. Kew 1931: 274. 1931. Vangueria latifolia (Sond.) Sond. in Fl. Cap. 3: 15. 1864. Pachystigma latifolium Sond. in Linnaea 23: 56. 1850. Vangueria loranthifolia K. Schum. in Pﬂanzenw. Ost-Afrikas, C: 385. 1895. Pachystigma loranthifolium (K. Schum.) Verdc. in Kew Bull. 42: 140. 1987. subsp. loranthifolia. subsp. salaense (Verdc.) Lantz, comb. nov. Pachystigma lorantifolium subsp. salaense Verdc. in Kew Bull. 42: 140. 1987. Vangueria macrocalyx Sond. in Linnaea 23: 59. 1850. Vangueria caﬀra Sim in Forest ﬂ. Cape: 244. 1907, syn. nov. Pachystigma caﬀrum (Sim) Robyns in Bull. Jard. Bot. État. 11: 120. 1928, syn. nov. Pachystigma macrocalyx (Sond.) Robyns in Bull. Jard. Bot. État. 11: 130. 1928. We agree with the view in Retief (2003) that Pachystigma caﬀrum (Sim) Robyns should be considered a synonym of Pachystigma macrocalyx (Sond.). Vangueria madagascariensis J.F. Gmelin in Syst. Nat., ed. 13, 2: 367. 1791. Vangueria micropyren (Verdc.) Lantz, comb. nov. Pachystigma micropyren Verdc. in Kew Bull. 42: 139. 1987. Vangueria mollis (Robyns) Lantz, comb. nov. Tapiphyllum molle Robyns in J. Bot. 69: 186. 1931. Vangueria monteiroi (Oliv.) Lantz, comb. nov. Ancylanthos monteiroi Oliv. in Icon. pl. 8: 7. 1877. Lagynias monteiroi (Oliv.) Bridson in Kew Bull. 51: 351. 1996. Vangueria obtusifolia K. Schum. in Bot. Jahrb. Syst. 28: 493. 1900. H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Tapiphyllum obtusifolium (K. Schum.) Robyns in Bull. Jard. Bot. État. 11: 115. 1928. Vangueria pachyantha (Robyns) Lantz, comb. nov. Tapiphyllum pachyanthum Robyns in Bull. Jard. Bot. Brux. 32: 150. 1962. Vangueria pallidiﬂora (Bullock) Lantz, comb. nov. Lagynias pallidiﬂora Bullock in Kew Bull. 1931: 273. 1931. Vangueria parvifolia Sond. in Linnaea 23: 58. 1850. Vangueria praecox Verdc. in Kew Bull. 36: 554. 1981. Vangueria proschii Briq. in Annuaire Conserv. Jard. Bot. Genève. 6: 7. 1902. Vangueria psammophila (K. Schum.) Lantz, comb. nov. Fadogia psammophila K. Schum. in Bot. Jahrb. Syst. 32: 147. 1902. Tapiphyllum psammophilum (K. Schum.) Robyns in Bull. Jard. Bot. État. 11: 105. 1928. Vangueria pygmaea Schlecht. in J. Bot. 35: 342. 1897. Pachystigma pygmaeum (Schlecht.) Robyns in Bull. Jard. Bot. État. 11: 122. 1928. Vangueria quarrei (Robyns) Lantz, comb. nov. Tapiphyllum quarrei Robyns in Bull. Jard. Bot. Brux. 32: 144. 1962. Vangueria randii S. Moore in J. Bot. 40: 252. 1902. Vangueria rhodesiaca (Tennant) Lantz, comb. nov. Ancylanthos rhodesiacus Tennant in Kew Bull. 19: 283. 1965. Tapiphyllum rhodesiacum (Tennant) Bridson in Kew Bull. 51: 351. 1996. Vangueria rubiginosa (Desf.) Lantz, comb. nov. Ancylanthos rubiginosus Desf. in Mém. Mus. Hist. Nat. 4: 5 1818, (as rubiginosa). Vangueria rufescens (E.A. Bruce) Lantz, comb. nov. Ancylanthos rufescens E.A. Bruce in Kew Bull. 1936: 477. 1936. Lagynias rufescens (E.A. Bruce) Verdc. in Kew Bull. 42: 143. 1987. subsp. rufescens. subsp. angustiloba (Verdc.) Lantz, comb. nov. Lagynias rufescens (E.A. Bruce) Verdc. subsp. angustiloba Verdc. in Kew Bull. 42: 143. 1987. Vangueria schliebenii (Verdc.) Lantz, comb. nov. Tapiphyllum schliebenii Verdc. in Kew Bull. 36: 534. 1981. Vangueria schumanniana (Robyns) Lantz, comb. nov. Tapiphyllum schumannianum Robyns in Bull. Jard. Bot. État. 11: 109. 1928. Pachystigma schumannianum (Robyns) Bridson & Verdc. in Fl. Trop. E. Afr., Rubiaceae: 768. 1991. subsp. schumanniana subsp. mucronulata (Robyns) Lantz, comb. nov. Pachystigma schumannianum (Robyns) Bridson & Verdc. subsp. mucronulatum in Fl. Trop. E. Afr., Rubiaceae: 768. 1991. Vangueria solitariiﬂora (Verdc.) Lantz, comb. nov. Pachystigma solitariiﬂorum Verdc. in Kew Bull. 36: 541, Fig. 9. 1981. Vangueria soutpansbergensis N. Hahn in Bothalia 27: 45. 1997. Vangueria thamnus (Robyns) Lantz, comb. nov. Pachystigma thamnus Robyns in Bull. Jard. Bot. État. 11: 121. 1928. Vangueria triﬂora (Robyns) Lantz, comb. nov. Pachystigma triﬂorum Robyns in Bothalia 3: 184. 1937. Vangueria venosa (Hochst.) Sond. in Fl. Cap. 3: 14. 1864. Pachystigma venosum Hochst. in Flora 25: 235. 1842. Pachystigma cymosum Robyns in Bull. Jard. Bot. État. 11: 127. 1928, syn. nov. Pachystigma cymosum Robyns is listed as a synonym of Pachystigma venosum Hochst. in Retief (2003), and we agree with that view. Vangueria verticillata (Robyns) Lantz, comb. nov. Tapiphyllum verticillatum Robyns in Bull. Jard. Bot. Brux. 32: 141. 1962. H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera Vangueria volkensii K. Schum. in Pﬂanzenw. Ost-Afrikas, C: 384. 1895. Vangueria zambesiaca Lantz, nom. nov. Vangueria velutina Hiern in Fl. trop. Afr. 3: 151. 1877, nom. illegit. Tapiphyllum velutinum (Hiern) Robyns Bull. Jard. Bot. État. 11: 111. 1928. Vangueria velutina Hook. (1830) precedes Vangueria velutina Hiern (1877) and the latter name is thus a later homonym and should not be used. The species epithet used here is derived from the strictly Zambesian distribution of the species (Bridson, 1998). Vangueria velutina Hook. is above combined under V. infausta Burchell. The authors would like to thank K. Bremer, B. Oxelman, M. Popp, M. Thulin, and L. Tibell for suggestions and comments on the manuscript. Plant material was received from the BR, K, LISC, MO, NYBG, S, TEF, and UPS herbaria, their help is greatly appreciated. DNA material was also received from R. Gereau and A. Davis whose help we gratefully acknowledge. Diane Bridson kindly shared with us her view on the classiﬁcation of Vanguerieae, and for this we are very thankful. Thanks also go to A. Bauer and N. Heidari for help with sequencing. Financial support was received from the Swedish Research Council (to B. B.), which is gratefully acknowledged. References Asmussen C. B. (1999) Toward a chloroplast DNA phylogeny of the tribe Genomeae (Palmae). Mem. New York Bot. Gard. 83: 121–129. Baldwin B. G. (1992) Phylogenetic utility of the internal transcribed spacers of nuclear ribosomal DNA in plants: an example from the Compositae. Molec. Phylogenet. Evol. 1:3–16. Bridson D. M. (1987) The recognition and recircumscription of the African genus Multidentia (Rubiaceae-Vanguerieae). Kew Bull. 42: 641–654. Bridson D. M. (1996) The tropical African genus Ancylanthos (Rubiaceae-Vanguerieae) reconsidered. Kew Bull. 51: 343–352. Bridson D. M. (1998) Rubiaceae (Tribe Vanguerieae). In: Pope G. V. (ed.) Flora Zambesiaca. Royal Botanic Gardens, Kew, London, pp. 211–377. Buckler IV E. S., Ippolito A., Holtsford T. P. (1997) The evolution of ribosomal DNA: divergent paralogues and phylogenetic implications. Genetics 145: 821–832. Bullock A. A. (1931) Tropical African Plants: VIII. Bull. Misc. Inform. No. 5: 270–275. Desfontaines M. (1818) Nouveau genre de la famille des Rubiacées: Ancylanthos. Mém. Mus. Hist. Nat. 4: 5–7. Downie S. R., Katz-Downie D. S., Watson M. F. (2000) A phylogeny of the ﬂowering plant family Apicaeae based on chloroplast DNA rpl16 and rpoc1 intron sequences: towards a suprageneric classiﬁcation of subfamily Apioideae. Amer. J. Bot. 87: 273–292. Doyle J. J., Doyle J. L. (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. Bot. Soc. Amer. 19: 11–15. Hahn N. (1997) Rubiaceae: A new species of Vangueria from the Soutpansberg. Bothalia 27: 45–48. Hallé N. (1959) Sur les Cuviera (Rubiacées) d’Afrique intertropicale et description pour ce genre de deux espèces et de deux variétés nouvelles. Bull. Soc. Bot. France 7–8: 342–348. Hillis D. M., Moritz C., Porter C. A., Baker R. J. (1991) Evidence for biased geneconversion in concerted evolution of ribosomal DNA. Science 251: 308–309. Holmgren P. K., Holmgren N. H., Barnett L. C. (1990) Index herbariorum, Part 1: The herbaria of the world, 8th edn. New York Botanical Garden, New York. Hutchinson J. (1931) Robynsia. In: Hutchinson J., Dalziel J. M. (eds.) Flora of West Tropical Africa. The Crown Agents for the Colonies, London, pp. 108–109. Igersheim A. F. (1989) Beiträge zur Klärung der Gattungsabgrenzungsprobleme innerhalb der Rubiaceae-Vanguerieae. Ph. D. Thesis. Universität Wien. Igersheim A. F. (1993) Gynoecium development in Rubiaceae-Vanguerieae, with particular reference to the ‘‘stylar-head’’-complex and secondary pollen presentation. Pl. Syst. Evol. 187: 175– 190. Jussieu A. L. de (1789) Genera Plantarum Secundum Ordines Naturales Disposita. Paris. Kellogg E. A., Appels R., Mason-Gamer R. J. (1996) When genes tell diﬀerent stories: the H. Lantz and B. Bremer: Phylogeny of Vangueria and closely related genera diploid genera of Triticeae (Gramineae). Syst. Bot. 21: 321–347. Lantz H., Andreasen K., Bremer B. (2002) Nuclear rDNA ITS data used to construct the ﬁrst phylogeny of Vanguerieae (Rubiaceae). Pl. Syst. Evol. 230: 173–187. Lantz H., Bremer B. (2004) Phylogeny inferred from morphology and DNA data: characterizing well-supported groups in Vanguerieae (Rubiaceae). Bot. J. Linn. Soc. 146: 257–283. McDade L. (1990) Hybrids and phylogenetic systematics I. Patterns of character expression in hybrids and their implications for cladistic analysis. Evolution 44: 1685–1700. Pamilo P., Nei M. (1988) Relationships between gene trees and species trees. Mol. Biol. Evol. 5: 568–583. Razaﬁmandimbison S., Kellogg E., Bremer B. (2004) Recent origin and phylogenetic utility of divergent ITS putative pseudogenes: a case study from Naucleeae (Rubiaceae). Syst. Biol. 53: 177– 192. Retief E. (2003) Rubiaceae. In: Germishuizen G., Meyer N. L. (eds.) Plants of southern Africa: an annotated checklist. Strelitzia 14: 825–841. National Botanical Institute, Pretoria. Rieseberg L. H., Soltis D. E. (1991) Phylogenetic consequences of cytoplasmic gene ﬂow in plants. Evol. Trends Pl. 5: 65–84. Rieseberg L. H. (1997) Hybrid origins of plant species. Annual Rev. Ecol. Syst. 28: 359–389. Robbrecht E. (1988) Tropical woody Rubiaceae. Opera Bot. Belg. 1: 1–271. Robyns W. (1928) Tentamen Monographiae Vanguerieae Generumque Aﬃnium. Bull. Jard. Bot. État. 11: 1–359. Robyns W. (1931) Vanguerieae Gossweilerianae. J. Bot. 69: 189–190. Robyns W. (1962) Contribution a l’étude du genre Tapiphyllum Robyns. Bull. Jard. Bot. État. 32: 133–153. Sanderson M. J., Doyle J. J. (1992) Reconstruction of organismal and gene phylogenies from data on multigene families: concerted evolution, homoplasy, and conﬁdence. Syst. Biol. 41: 4–17. Simmons M. P., Ochoterena H. (2000) Gaps as characters in sequence-based phylogenetic analyses. Syst. Biol. 49: 369–381. 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. Soltis D. E., Johnson L. A., Looney C. (1996) Discordance between ITS and chloroplast topologies in the Boykinia group (Saxifragaceae). Syst. Bot. 21: 169–185. Swoﬀord D. L. (2002) PAUP*. Phylogenetic Analysis Using Parsimony (*and other methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. Verdcourt B. (1958) Remarks on the classiﬁcation of the Rubiaceae. Bull. Jard. Bot. État. 28: 209– 281. Verdcourt B. (1981) Notes on African Rubiaceae. Kew Bull. 36: 493–557. Verdcourt B. (1987) Notes on African RubiaceaeVanguerieae. Kew Bull. 42: 123–199. Verdcourt B., Bridson D. M. (1991) Rubiaceae (Part 3). In: Polhill R. M. (ed.) Flora of Tropical East Africa. A.A. Balkema, Rotterdam/Brookﬁeld, pp. 749–956. Wendel J. F., Schnabel A., Seelanan T. (1995) Bidirectional interlocus concerted evolution following allopolyploid speciation in cotton (Gossypium). Proc. Natl. Acad. Sci. USA 92: 280– 284. Wendel J. F., Doyle J. J. (1998) Phylogenetic incongruence: window into genome history and molecular evolution. In: Soltis P. S., Soltis D. E., Doyle J. J. (eds.) Molecular systematics of plants, 2nd edn. Chapman and Hall, New York, pp. 265–296. Addresses of the authors: Henrik Lantz (email: henrik.lantz@ebc.uu.se), Department of Systematic Botany, Evolutionary Biology Centre, Uppsala University, Norbyvägen 18D, 75236, Uppsala, Sweden. Birgitta Bremer (e-mail: birgitta.bremer@ bergianska.se) The Bergius Foundation at the Royal Swedish Academy of Sciences, P. O. Box 50017, 10405 Stockholm, Sweden.

RELATED PAPERS

RELATED TOPICS

Log In

Phylogeny of the complex Vanguerieae (Rubiaceae) genera Fadogia, Rytigynia, and Vangueria with close relatives and a new circumscription of Vangueria

Phylogeny of the complex Vanguerieae (Rubiaceae) genera Fadogia, Rytigynia, and Vangueria with close relatives and a new circumscription of Vangueria

Related Papers

RELATED PAPERS

RELATED TOPICS