American Journal of Botany 96(5): 989–1010. 2009. A PRELIMINARY PHYLOGENY OF THE ‘DIDYMOCARPOID GESNERIACEAE’ BASED ON THREE MOLECULAR DATA SETS: INCONGRUENCE WITH AVAILABLE TRIBAL CLASSIFICATIONS1 Michael Möller,2,5 Martin Pfosser,3,6 Chang-Gee Jang,3,7 Veronika Mayer,3 Alexandra Clark,2 Michelle L. Hollingsworth,2 Michael H. J. Barfuss,3 Yin-Zheng Wang,4 Michael Kiehn,3 and Anton Weber3 2Royal Botanic Garden Edinburgh, 20A Inverleith Row, Edinburgh EH3 5LR, Scotland, UK; 3Faculty Centre of Biodiversity, University of Vienna, Rennweg 14, A-1030 Vienna, Austria and 4Institute of Botany, Chinese Academy of Sciences, Beijing, China The ‘didymocarpoid Gesneriaceae’ (traditional subfam. Cyrtandroideae excluding Epithemateae) are the largest group of Old World Gesneriaceae, comprising 85 genera and 1800 species. We attempt to resolve their hitherto poorly understood generic relationships using three molecular markers on 145 species, of which 128 belong to didymocarpoid Gesneriaceae. Our analyses demonstrate that consistent topological relationships can be retrieved from data sets with missing data using subsamples and different combinations of gene sequences. We show that all available classifications in Old World Gesneriaceae are artificial and do not reflect natural relationships. At the base of the didymocarpoids are grades of clades comprising isolated genera and small groups from Asia and Europe. These are followed by a clade comprising the African and Madagascan genera. The remaining clades represent the advanced Asiatic and Malesian genera. They include a major group with mostly twisted capsules. The much larger group of remaining genera comprises exclusively genera with straight capsules and the huge genus Cyrtandra with indehiscent fruits. Several genera such as Briggsia, Henckelia, and Chirita are not monophyletic; Chirita is even distributed throughout five clades. This degree of incongruence between molecular phylogenies, traditional classifications, and generic delimitations indicates the problems with classifications based on, sometimes a single, morphological characters. Key words: atpB-rbcL spacer; Bayesian inference analysis; ITS; maximum parsimony; molecular phylogeny; Old World Gesneriaceae; taxonomy; trnL-F intron-spacer. Gesneriaceae is a medium-sized family, comprising between 150 and 160 genera, and over 3200 species (Weber, 2004; Weber and Skog, 2007). The distribution is mainly tropical and subtropical, both in the Old and the New World, with minor outliers to the north (e.g., Pyrenees, Balkan peninsula, central and northern China) and to the south (southeastern Australia, New Caledonia, New Zealand, southern Chile). While the neotropical Gesneriaceae (subfam. Gesnerioideae, including Coronanthereae, raised to subfamily rank by Wiehler, 1 1983) have received much attention by molecular systematists and apparently approach consolidation regarding a phylogenetic and formal classification (Smith, 1996, 2000; Smith et al., 1997a, b; Zimmer et al., 2002; Perret et al., 2003; Roalson et al., 2005a, b), the paleotropical Gesneriaceae (subfam. Cyrtandroideae, now Didymocarpoideae) lag behind. The last attempt at a formal (morphological) classification was by Burtt and Wiehler (1995), who distinguished five tribes: (1) Cyrtandreae (with indehiscent, berry-like fruits); (2) Trichosporeae (with appendaged seeds); (3) Epithemateae (Klugieae and Loxonieae sensu Burtt (1963), a group rather difficult to characterize, partly with peculiar anatomical characters such as secretory canals and meduallary vascular bundles); (4) Didymocarpeae (the many remaining genera arranged alphabetically to express the wide lack of phylogenetic understanding); and Titanotricheae including the sole genus Titanotrichum. Additional tribes (Saintpaulieae, Ramondeae, Rhynchotecheae) have been recognized by Ivanina (1965a, b, 1967) and Wang et al. (1990, 1992). Detailed morphological investigations (Weber, 1975–1988, as cited in Mayer et al., 2003) and, more recently, a molecular analysis based on cpDNA sequences (Mayer et al., 2003) showed that the Epithemateae form a distinct clade, sister to the remaining Old World Gesneriaceae. Both the morphological investigations and the molecular analyses led to a fairly good understanding of the phylogenetic relationships for six of seven genera of the ‘epithematoid Gesneriaceae’, as they are informally referred to in Weber (2004) and Weber and Skog (2007). The present paper deals with the diverse non-epithematoid Old World Gesneriaceae, the ‘didymocarpoid Gesneriaceae’. A number of genera had already been included in the molecular Manuscript received 26 August 2008; revision accepted 12 December 2008. The authors thank the horticulturists at the Royal Botanic Garden Edinburgh, in particular, S. Scott, S. Barber, and A. Ansell; and A. Sieder at the Botanic Garden Vienna, for excellent work in maintaining and accurately curating the living Gesneriaceae collections. They are also grateful to staff at KIB for assistance with fieldwork, especially Dr. D. Z. Li and Dr. L. M. Gao, and for the provision of samples by Profs. Z. J. Gu and C. Q. Zhang (KIB). The work was carried out and is still continued in the frame of a multidisciplinary cooperation project between the University of Vienna (supported by the Austrian Science Fund (FWF)proj. no. P-13107-Bio) and the Royal Botanic Garden Edinburgh (RBGE). RBGE is supported by the Scottish Government Rural and Environment Research and Analysis Directorate (RERAD). Fieldwork of M.M. was supported by the Percy Sladen Memorial Fund, Oleg Polunin Memorial Fund, Davis Expedition Fund of the University of Edinburgh, and FWF. 5 Author for correspondence (e-mail: m.moeller@rbge.ac.uk) 6 Present address: Biology Centre Linz, Oberösterreichisches Landesmuseum, Johann Wilhelm Klein Straße 73, A-4040 Linz, Austria 7 Present address: Department of Biology Education, College of Education, Kongju National University, Gongju, Korea doi:10.3732/ajb.0800291 989 American Journal of Botany 990 analysis of Mayer et al. (2003), and preliminary results of the current study formed the basis of the informal classification presented by Weber (2004). In the meantime, the number of samples included in the molecular analysis has considerably increased, and sufficient results are available to make some headway in our understanding of relationships in this group, particularly with respect to previous subdivisions into tribes. Nonetheless, we abstain from a new formal classification at this point because a significant number of genera are not yet included in our analysis. Effectively analyzing large data sets is challenging and complete analyses computationally intractable (e.g., Savolainen and Chase, 2003). Furthermore, sample-rich analyses including multiple data sets often suffer from missing data or issues of incongruencies between data sets (Sanderson and Shaffer, 2002). Here we address the issue of missing data by analyzing selected subsets of samples with complete data compared to a complete sample set with missing data, and demonstrate that “correct” topologies of phylogenetic trees can be recovered from largely incomplete data matrices. MATERIALS AND METHODS Plant materials—Material for DNA extraction came in various forms and from diverse sources, including fresh leaves from research collections (E, HBV), silica-dried leaves of field collections, and herbarium specimens at various herbaria (Appendix 1). Outgroup choice—To produce a reliable phylogeny of the family Gesneriaceae, we included members of major families of basal Lamiales (Oleaceae, Tetrachondraceae, Scrophulariaceae, Plantaginaceae s.l., Calceolariaceae s.s.) (Oxelman et al., 1999; Bremer et al., 2002; AGP II, 2003) in the analysis and used them as outgroups. The trees were rooted on Oleaceae (Olmstead et al., 2000; Tank et al., 2006). Details of outgroup taxa are given in Appendix 1. Ingroup taxa—The samples of Gesneriaceae included 174 species in 78 genera. Their taxonomic classification followed Weber (2004). Twenty-three species were from gesnerioid Gesneriaceae covering 20 (of 53) genera, and six species belonged to coronantheroid Gesneriaceae (six of nine genera). A total of 17 species were sampled from epithematoid Gesneriaceae (six of seven genera); only the monotypic genus Gyrogyne W.T.Wang is missing. Gyrogyne subaequifolia W.T.Wang is only known from the type collection, and DNA extraction from the isotype failed. Finally, 128 species (from 46 of 78 genera) came from didymocarpoid Gesneriaceae. The sample selection covered all tribes recognized by Burtt and Wiehler (1995, Epithemateae as ‘Klugieae’). Speciose genera of didymocarpoid Gesneriaceae were represented by a higher number of samples; Aeschynanthus Jack: five species of ca. 185, Chirita Buch.-Ham.: 15 of 80–140, Cyrtandra Forst.: six of (perhaps) 450–600, Didymocarpus Wall. s.s.: seven of ca. 70, Henckelia: five of ca. 180, Paraboea: 14 of 87, Streptocarpus: 10 of ca. 140. In addition, Jerdonia Wight (monotypic with J. indica Wight) and Titanotrichum Solereder (monotypic with T. oldhamii Solereder) were included in the analysis, the former described in Weber (2004) with uncertain familial affiliations and the latter as excluded from Gesneriaceae. Jerdonia was originally described in Gesneriaceae (Wight, 1850), but referred to Scrophulariaceae by Burtt (1977). Titanotrichum was originally described in Scrophulariaceae, but recent molecular analyses suggest it to belong to Gesneriaceae, closely allied to New World taxa (Wang et al., 2004). DNA extraction, PCR, and direct sequencing—Genomic DNA was extracted using a modified CTAB procedure (Doyle and Doyle, 1987, 1990) and/ or the Qiagen DNeasy DNA Isolation Kit (Crawley, UK) following the manufacturer’s protocol. PCR amplification of the trnL-F intron-spacer region (trnL-F) and atpBrbcL spacer (atpB-rbcL) followed Mayer et al. (2003). Some trnL-F and ITS sequences were obtained as follows: PCR amplification was performed using primers ‘c’ and ‘f’ (Taberlet et al., 1991) for trnL-F or primers 5P and 8P (Möller and Cronk, 1997) for ITS on an MJ Research PTC-200 DNA Engine [Vol. 96 thermal cycler. The 50 µL reactions contained 5 µL 10× NH4 reaction buffer (Bioline, UK), 5 µL dNTPs (2 mM), 2.5 µL MgCl2 (50 mM), 1.5 µL of each primer (10 µM), 32.25 µL dH2O, 1.25 U Biotaq polymerase (Bioline, UK) and 1.0 µL DNA template DNA. The PCR thermocycle profile for trnL-F was: initial denaturation for 4 min at 94°C; followed by 30 cycles of 45 s at 94°C, 45 s at 55°C and 3 min at 72°C; with a final extension step for 10 min at 72°C. ITS amplifications were carried out using the PCR profile described in White et al. (1990) or Möller and Cronk (1997). PCR products were run on 1% agarose gels to check for amplification success and quality. Amplified fragments were purified using QIAquick PCR purification kits (Qiagen, Crawley, UK) following the manufacturer’s protocol, and sequenced using the dideoxy chain-termination method. Cycle sequencing was performed as in Mayer et al. (2003), or in 10 µL reactions containing 4 µL DTCS Quickstart mix (Beckman Coulter, High Wycombe, UK), 1 µL primer (10 µM), 3 µL dH2O and 2 µL purified PCR product, under the following PCR conditions: 35 cycles of 96°C for 20 s, 50°C for 20 s and 60°C for 4 min. Sequencing primers were identical to those used for PCR with the addition that primer ‘d’ (Taberlet et al., 1991) was also used to sequence the trnL intron region. Sequencing atpB-rbcL followed Mayer et al. (2003). Sequencing PCR products were purified following the manufacturer’s instructions, then run and analyzed on a CEQ 8000 Genetic Analysis System (Beckman Coulter, UK). Editing and assemblage of contigs was performed using the programs CEQuence Investigator (Beckman Coulter, High Wycombe, UK) and Sequencher 4.5 (Gene Codes Corp, Ann Arbor, USA), and for atpB-rbcL as in Mayer et al. (2003). The cpDNA alignment matrices were assembled manually, based on those of Mayer et al. (2003). ITS sequences were initially aligned using the program CLUSTAL W (Larkin et al. 2007) and the alignment then adjusted manually, as described in Möller and Cronk (1997). Phylogenetic analysis—For this study, 123 trnL-F, 59 atpB-rbcL, and 65 ITS sequences were newly acquired and submitted to GenBank; the remaining sequences were retrieved from GenBank (Appendix 1). Because a complete overlap of sequences across the three gene regions used could not be achieved, the data were analyzed in three sets: (1) A combined trnL-F and atpB-rbcL matrix (cpDNA) that included 129 samples (59 trnL-F and 59 atpB-rbcL sequences were new, the rest from GenBank) (TreeBase: SN4215–20651). This complete plastid matrix included nine outgroup taxa; 22 gesnerioid, five coronantheroid, and 18 epithematiod samples; and a backbone set of 73 samples for didymocarpoid taxa and Jerdonia and Titanotrichum. The main purpose for this analysis was to resolve out-/ingroup relationships and the backbone topology for the Gesneriaceae family, specifically for the basal Asiatic and European taxa. The addition of the two samples of Tetraphyllum Griff ex C.B.Clarke, for which no atpB-rbcL data were available, did not alter the tree topology in any analysis, and were thus included in a final matrix of 131 samples. (2) A complete, combined trnL-F and ITS data set with 88 didymocarpoid samples (71 trnL-F and 65 ITS sequences were new, the rest from GenBank) (TreeBase: SN4215–20652). This analysis focused on resolving relationships among the advanced Asiatic and Malesian genera (Weber, 2004), with specific focus on testing the monophyly of members of the tribe Trichosporeae. This analysis was, based on results of analysis 1, rooted on African and Madagascan samples. The ITS alignment at higher taxonomic level above these samples was found to be too ambiguous and the sampling too incomplete to be analyzed. (3) A combined trnL-F, atpB-rbcL, and ITS data matrix with 201 samples across the family, including missing sequences (201 trnL-F, 129 atpB-rbcL, 88 ITS) (TreeBase: SN4215–20650). 123 trnL-F, 59 atpB-rbcL and 65 ITS sequences were new, the rest were from GenBank. The purpose for this analysis was to obtain a topology for all samples included in our study. The topology of major groups in this analysis was basically identical to those of the complete analyses 1 and 2. The matrices were analyzed by maximum parsimony (MP) implemented in the program PAUP* version 4.0b10 (Swofford, 2002) and by Bayesian Markov chain Monte Carlo (MCMC) inference (BI; Yang and Rannala, 1997) using the program MrBayes version 3.1.2 (Huelsenbeck and Ronquist, 2001, 2007) on unordered and equally weighted characters. Combinability of data sets was determined using the incongruence length difference (ILD) test of Farris et al. (1994, 1995), implemented in PAUP* as the partition-homogeneity test, on 1000 replicates of repartitioning with treebisection-reconnection (TBR) on (Yuan et al., 2005). In view of the high number of sequences included, MP starting trees were found using the parsimony ratchet (Nixon, 1999) implemented in the programs PAUPRat (Sikes and Lewis, 2001) and PAUP*. Only the shortest of the 201 May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae trees saved with PAUPRat were further optimized, with TBR and MulTrees activated, in PAUP*. For analysis 3, the MaxTree option was set to 100 000 trees. Descriptive tree statistics (ensemble consistency index [CI; Kluge and Farris, 1969], ensemble retention index [RI; Farris, 1989a], and ensemble rescaled consistency index [RC; Farris, 1989b]) were obtained using PAUP*. Statistical branch support analyses were performed twofold, as 10 000 replicates of heuristic bootstrap replicates (BS; Felsenstein, 1985) with TBR swapping on and MulTrees off (Spangler and Olmstead, 1999) in PAUP* and as decay indices (DI; Bremer, 1988), derived from the program AutoDecay version 4.0.2 (Eriksson, 1999) and PAUP* from 100 replicates of random addition. Models and parameters priors for the BI analyses were obtained independently for each data set and gene sequence in each analysis using Modeltest (Posada and Crandall, 1998). In the higher level analysis 1 (with 131 samples), the model TVM+I+G was suggested by the Akaike information criterion (AIC; Akaike, 1974) for the trnL-F data, and TVM+G for the atpB-rbcL data. In analysis 2, the GTR+G model was selected for trnL-F, and GTR+I+G for ITS under the AIC. For analysis 3, the models GTR+I+G, TVM+G and GTR+I+G were selected for the trnL-F, atpB-rbcL, and ITS sequences, respectively. For analyses 1 and 2, two million generations were run with four MCMC chains in two independent parallel analyses, with one tree sampled every 100 generations (10 000 trees). The first 130 000 generations or the first 1300 trees for the cpDNA analysis 1 and the first 60 000 generations or 600 trees for analysis 2 were discarded as burn-in (generations prior to stationarity of likelihood values). For analysis 3, the number of generations was four million, and the burn-in 400 000 generations. A majority rule consensus tree was constructed in PAUP* from the remaining trees, combined from the two parallel analyses. Although not strictly comparable to bootstrap values (see also Fig. 4) the retrieved posterior probabilities (PP) indicate robustness of clades (Lewis, 2001; Alfaro et al., 2003; but see Cummings et al., 2003; Erixon et al., 2003). Majority rule consensus trees of individual BI runs were identical (except for two branches in analysis 3 with PP values of 0.5 and 0.53). A high correlation of the PP support values was found between the two parallel runs of the Bayesian analysis for all three data sets (Appendix S1, see Supplemental Data with online version of this article). RESULTS Matrix characteristics—The combined cpDNA matrix (analysis 1) had 2525 aligned characters (trnL-F: 1278 characters, atpB-rbcL: 1247 characters), with 366 ambiguous characters excluded, resulting in 2159 characters used for analysis, of which 585 were constant, 176 variable but parsimony uninformative and 386 (33.6%) parsimony informative in the trnL-F, and 484, 211, and 317 (31.3%) respectively in the atpB-rbcL data. The combined matrix for analysis 2 consisted of an alignment matrix of 2141 characters (trnL-F: 1278 characters; ITS: 863), with 276 ambiguous sites excluded, leaving 1865 characters in the analysis. Of these, 752 were constant, 209 variable but parsimony uninformative, and 186 (16.2%) parsimony informative for trnL-F and 248, 127, and 343 (47.8%), respectively, for ITS. The combination of all three genes (analysis 3) resulted in a matrix of 3652 characters, of which 2877 unambiguously aligned characters were included in the analysis. Of these, 1251 were constant, 528 variable but parsimony uninformative, and 1098 (38.1%) parsimony informative. Analysis 1: Tree topology trnL-F and atpB-rbcL— The cpDNA strict consensus tree, based on 6912 most parsimonious trees of 2636 steps, showed a highly resolved topology with strong backbone support in most areas (Fig. 1). Bootstrap values ranged from 50% to 100% with 64 of 98 internal branches possessing support values of 90% or higher, and decay indices from 1 to 69. The Gesneriaceae were highly supported (BS = 100%; DI = 15) with Calceolariaceae its closest allied family (BS = 90%; DI = 3). A monophyletic coronantherioid clade (BS = 99%; DI =7) nested within the gesnerioid clade. This 991 clade also included the Old World (OW) genus Titanotrichum (BS = 82%; DI = 2). Sister to this group were the other OW samples, which formed a well-supported clade (BS = 96%; DI = 5). Within this clade, the epithematoid (BS = 100%; DI = 13) clade was sister to the remaining OW genera (BS = 98%; DI = 3). Among these, the hitherto unassigned genus Jerdonia split off first, followed by several grades and one polytomy for the basal Asiatic and European genera. The African and Madagascan genera formed a monophyletic sister clade (BS = 100%; DI = 11) to the advanced Asiatic and Malesian genera (BS = 65%; DI = 2). The resolution and support among those was not very high, but clades for individual genera were mostly well supported (BS = 98–100%; DI = 2–11), with the exception of samples of Chirita, which clustered in five different clades that were closely associated with other genera. The two genera belonging to tribe Cyrtandreae sensu Burtt (1963) and Burtt and Wiehler (1995), Cyrtandra and Rhynchotechum Blume, were resolved separately in distant positions, among the basal Asiatic and European genera (Rhynchotechum) and among the advanced Asiatic and Malesian genera (Cyrtandra). Tribe Ramondeae sensu Wang et al. (1990, 1992) was also not monophyletic, with Ramonda among the basal Asiatic and European genera, and Conandron Siebold & Zucc. among the advanced Asiatic and Malesian genera. The topology of the majority rule consensus tree of the BI analysis (Appendix S2, see online Supplemental Data) was very similar to the MP strict consensus tree, with only minor differences in terminal clade relationships, which were not highly supported in any analysis. The most obvious discrepancy was a sister rather than grade relationship of the European and part of basal Asiatic genera (including Leptoboea Benth., Platystemma Wall., Rhynchotechum and Boeica C.B.Clarke) (Fig. 1, arrow). Analysis 2: Tree topology trnL-F and ITS— The MP analysis yielded 28 most parsimonious trees of 3230 steps. The resulting strict consensus tree was highly resolved with 44 of 78 internal branches with 90% or higher bootstrap branch support and decay values between 1 and 36 (Fig. 2). The trees showed a strongly supported monophyletic ingroup (BS = 100%; DI = 29) of African and Madagascan genera, followed by a clade of mainly twisted-fruited advanced Asiatic and Malesian genera (Boea group) (BS = 99%; DI = 8). Among the remaining advanced Asiatic and Malesian genera, the tree topology was resolved but with low backbone support. Several subclades harboring several genera received high branch support (discussed later). Of the four genera of tribe Trichosporeae sensu Burtt and Wiehler (1995) included in our analysis (Aeschynanthus, Agalmyla Blume, Loxostigma C.B.Clarke, and Lysionotus D. Don), the first three fell together in a polytomy (with Loxostigma and Agalmyla as sister genera), while Lysionotus appeared distant to these genera and more closely related to Hemiboea C.B.Clarke (Fig. 2). The BI analysis majority rule consensus tree showed more resolution among the twisted-fruited genera (online Appendix S2), but less resolution for the backbone structure of the straightfruited genera, though the branches involved generally received low branch support in the MP analysis for the latter. Analysis 3: Three-gene tree topology— The MP analysis on all three-sequence matrices resulted in 18 405 most parsimonious trees of 5655 steps (Fig. 3). The topology of the strict con- 992 American Journal of Botany [Vol. 96 May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae sensus tree was identical to that in the respective areas of the two two-gene MP analyses 1 and 2. The southern hemisphere Coronantheroideae nested within the New World (NW) Gesnerioid genera and also included the OW Titanotrichum (BS = 82%; DI = 2). This clade was sister to the rest of the NW didymocarpoid genera (BS = 100%; DI = 15), with the Epithematoid (BS = 100%; DI = 13) sister to the rest (BS = 96%; DI = 5), followed by grades of Jerdonia (BS = 98%; DI = 3), Corallodiscus Batalin (BS = 100%; DI = 13), Tetraphyllum (BS = 99%; DI = 4), and a polytomy of the remaining basal Asiatic genera (BS = 92%; DI = 4) and the European genera (BS = 63%; DI = 1). Next followed the African and Madagascan genera, forming a strongly supported monophyletic clade (BS = 100%; DI = 14), with five genera, the African Acanthonema Hook f., Saintpaulia Wendl, and Schizoboea (K.Fritsch) B.L.Burtt, and the Madagascan Colpogyne B.L.Burtt and Hovanella A.Weber & B.L.Burtt nesting within the genus Streptocarpus Lindl.. Among the remaining advanced Asiatic and Malesian genera, the straight-fruited genus Didissandra C.B.Clarke appeared to split the clades of the African (BS = 100%; DI = 14) and Asian genera of the Boea group (BS = 85%; DI = 2). The latter including the genera Boea Comm. Ex Lam., Emarhendia Kiew, A.Weber & B.L.Burtt, Kaisupeea B.L.Burtt, Ornithoboea Parish ex C.B.Clarke, Paraboea Ridl., Rhabdothamnopsis Hemsl., and Spelaeanthus Kiew, A.Weber & B.L.Burtt. Trisepalum C.B.Clarke was found nested within a Paraboea clade (BS = 97%; DI = 5). One of the four Asiatic members of Streptocarpus, S. orientalis Craib included here, was also found in the Boea group, as well as the straight-fruited Henckelia ericii A.Weber (= Loxocarpus holttumii M.R.Hen.) and Chirita lacunosa (Hook f.) B.L.Burtt. The remaining advanced Asiatic and Malesian genera with straight fruits formed a weakly supported clade (BS = 62%; DI = 2) with fairly well-resolved relationships, but with low or no internal branch support. A highly supported clade (BS = 100%; DI = 16) of Chirita species of section Microchirita was sister to the remaining samples, followed by a clade of two SE Asian Henckelia Spreng. species (BS = 99%; DI = 6) and a mixed clade of Chirita and Henckelia from China and Sri Lanka (BS = 70%; DI = 2). The remainder of the samples fell into two large clades, but with little backbone support, one including the genera Anna Pellegr., Briggsia Craib, Calcareoboea C.Y.Wu, Chirita, Chiritopsis W.T.Wang, Hemiboea, Lysionotus, Petrocodon Hance, Petrocosmea Oliv., Primulina Hance, Raphiocarpus Chun and Ridleyandra A.Weber & B.L.Burtt. In this clade, Primulina was sister to Chirita longgangensis W.T.Wang (BS = 96%; DI = 5), nested deep inside a Chirita section Gibbosaccus clade (BS = 73%; DI = 1). Chiritopsis also nested within this clade (BS = 97%; DI = 2). The second large clade included Aeschynanthus, Agalmyla, Ancylostemon Craib, Briggsia, Chirita asperifolia (Blume) B.L.Burtt, Conandron, Cyrtandra, Didymocarpus, Loxostigma, Opithandra B.L.Burtt, and Oreocharis Benth. Chirita asperifolia (the type species of Chirita section Liebigia, reestablished 993 by Hilliard, 2004), was sister to Didymocarpus (BS = 92%; DI = 10). The four samples of Briggsia included here did not form a monophyletic group and the two sister pairs fell into different clades, one closely allied to Ancylostemon, Opithandra and Oreocharis (BS = 96%; DI = 3), the other sister to a group of genera including Raphiocarpus sp., Anna, Lysionotus, and Hemiboea, but with little branch support. The genus Chirita was highly polyphyletic, falling in four places among the straight-fruited, advanced Asiatic and Malesian genera and once in the Boea group. The BI analysis was largely congruent with the MP analysis for strongly supported branches but differed in areas that received low or no BS support in the MP analysis (Fig. 3; online Appendix S2). DISCUSSION Phylogenetic analysis— Our MP analyses on 190 Gesneriaceae samples resulted in well-resolved and stable topologies across the complete two-gene analyses (analyses 1 and 2), and gave a highly resolved strict consensus tree in the three-gene analysis (analysis 3), irrespective of missing sequences for some samples in the latter. This finding is in line with previous studies that suggested that even highly incomplete matrices can yield accurate phylogenetic topologies (Qiu et al., 1999; Murphy et al., 2001; Kearney, 2002; Wiens, 2003, 2006), and the effects of missing sequences can be minor (Wiens and Reeder, 1995). In fact, in our three-gene analysis, the more conserved cpDNA data supported the backbone of the trees, particularly at the higher taxonomic level, while the faster evolving ITS sequences resolved and supported relationships at lower taxonomic levels. This complementarity has previously been observed in other data sets (Qiu et al., 1999; Long et al., 2000; Sinclair et al., 2002). Furthermore, undesired effects such as long-branch attraction (Felsenstein, 1978), suspected to be problematic in combining incomplete data sets (Wiens, 2006) were not found in our analyses, as seen by the high similarity of topologies between analyses using different gene combinations. The ILD test for combinability of our matrices showed some incongruence between the data sets, but the probability values were above the P = 0.05 threshold. The applicability of the ILD test as a determinator of combinability of different data sets has been discussed controversially, with Cunningham (1997) being a proponent, while others opposed its use (Barker and Lutzoni, 2002; Hipp et al., 2004). However, the consistencies of our tree topologies across the three analyses suggest that the hierarchy retrieved from the three combined genes reflects reasonably well the relationships of the genera included here. Controversy surrounds the interpretation of posterior probabilities (PP) and their comparison to other branch support values (Wilcox et al., 2002; Suzuki et al., 2002; Cummings et al., 2003; Erixon et al., 2003). We investigated this issue by plotting the bootstrap against PPs (Fig. 4), and it becomes clear that ← Fig. 1. Strict MP consensus tree of 6912 most parsimonious trees of 2636 steps based on combined trnL-F and atpB-rbcL sequences of 131 samples (CI = 0.62, RI = 0.84, RC = 0.52). Bootstrap values and decay indices (bold and italics) are given above branches. Light gray boxes = Trichosporeae, dark gray boxes = Cyrtandreae; systematics (right side) follows Weber 2004; black bars = outgroups, hatched bars = gesnerioid, lined bars = coronantheroid, open bars = didymocarpoid, crosshatched = unassigned. Branches in bold are supported by posterior probabilities of 1.00 (black) or 0.99 (gray) in the Bayesian analysis. Asterisks indicate generic types. Arrow indicates a relationship found in the Bayesian analysis. 994 American Journal of Botany [Vol. 96 May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae the PP values were over most parts significantly higher than bootstrap values, with branches receiving BS values as little as 51% having PPs of 0.99 or 1.00. Thus, while Bayesian inference analysis here often results in more resolved trees with high PP values, these high PP values do not necessarily reflect strong robustness of the BI tree topologies, irrespective of missing data (see Fig. 4E for data set 2 without missing data, and Fig. 4F for data set 3 with missing data). Our interpretations consequently rely predominantly on the more conservative MP topology and its branch support values. We observed only few differences between the MP and BI analyses with some bearing on our main objectives relating to the monophyly of current tribes; one concerns the relationships between the European and part of the basal Asiatic genera, the other involves a sister relationship (BI) of Aeschynanthus and Agalmyla as opposed to such a relationship of the latter to Loxostigma (MP). In neither case is monophyly of the tribe Trichosporeae supported (discussed later), and thus these incongruencies have no effects on the interpretation of our data and the conclusions drawn. Highly supported relationships among the straight-fruited, advanced Asiatic and Malesian genera remained elusive, but it must be noted that over 30 didymocarpoid genera (most are likely to belong to this group) have not been included in our analysis yet. Their addition may well stabilize the tree topology in this area. It is not lack of sequence variation but lack of topology congruencies and synapomorphies (shared evolved characters) that causes the problem as inferable from individual phylograms (online Appendix S3). It was interesting to see that the distribution of branch support was greatly partitioned across the tree. While intergeneric branches received low or medium support, branches leading to genera were highly supported (Table 1). The latter indicates that the genera included (except nonmonophyletic genera) represented strongly defined genetic entities. Nonetheless, further data are required to better resolve the relationships of the didymocarpoid genera. Tribal classification of Old World Gesneriaceae–– The traditional family classifications of Gesneriaceae are those of Bentham (1876) and Fritsch (1893–1894). The latter represents the most detailed classification available (2 subfamilies, 18 tribes, 9 subtribes). Fritsch’s (1893–1894) classification was largely based on formal morphological characters, and many insufficiencies became apparent on closer inspection. Burtt (1954) even qualified Fritsch’s classification as “negative in quality” and a retrogression. In the Old World Gesneriaceae subfam. Cyrtandroideae, Burtt reduced the number of tribes to five: Cyrtandreae, Trichosporeae, Klugieae, Loxonieae, and Didymocarpeae. Later, based on morphological data of Weber (1975– 1988, cited in Mayer et al., 2003), he united Klugieae and Loxonieae into a single tribe (informally in Burtt, 1977, formally in Burtt and Wiehler, 1995), which is presently known as Epithemateae (Burtt, 1997). Apart from Didymocarpeae, each of the tribes contained only a small number of genera. The strong asymmetry in species number was paralleled by the dif- 995 ficulties in recognizing generic relationships. Burtt and Wiehler (1995) thus listed the genera of Didymocarpeae simply in alphabetical order. The five-tribe classification of OW Gesneriaceae of Burtt and Wiehler (1995) is reflected in our analyses only to a very low degree. Though morphologically not easy to define, the Epithemateae are a well-supported monophyletic group, which is sister to the remaining Old World genera (Mayer et al., 2003; Wang et al., 2004). The tribes Cyrtandreae and Trichosporeae were supported in a morphological cladistic analysis of Smith (1996), but not by molecular investigations (Smith et al., 1997a, b). In principal agreement with the latter data, we found both tribes nested within tribe Didymocarpeae sensu Burtt and Wiehler (1995). Furthermore, genera from neither tribe Cyrtandreae or Trichosporeae formed monophyletic groups, suggesting that the morphological features considered characteristics at tribal level (Cyrtandreae with indehiscent fruits and Trichosporeae with seed appendages) have repeatedly and independently evolved. Tribe Cyrtandreae included three genera in Burtt and Wiehler (1995): Cyrtandra, Sepikea Schltr., and Rhynchotechum. Because the monospecific Sepikea is probably just an abnormally tetrandrous Cyrtandra (perhaps even based on a misobservation, Burtt, 1998), only Cyrtandra and Rhynchotechum remain. According to our data, Rhynchotechum has its place among other basal Asiatic didymocarpoids, while Cyrtandra occupies a position within the advanced Asiatic and Malesian didymocarpoids. Therefore, there is no basis for retaining the tribe. Tribe Trichosporeae was thought to contain six genera in Burtt and Wiehler (1995). From these, two genera (Oxychlamys Schltr., Van Royen, 1983; Micraeschynanthus Ridl., Middleton, 2007) have been sunk into synonymy, thus leaving the four remaining genera Aeschynanthus, Agalmyla, Loxostigma, and Lysionotus. All these were included in our analysis. Some (but not all) of our trees suggest a closer, though unsupported, affinity between the genera Agalmyla, Loxostigma, and to some degree Aeschynanthus, but Lysionotus was always more closely associated with Hemiboea than with any other genus of Trichosporeae. Both the position within the advanced Asiatic and Malesian genera and the partial nonaffinity of the genera require disbandment of the Trichosporeae. Other tribes used in recent classifications— Early definitions of most tribes are problematic, but some have been revived, redefined, or newly established in recent work. These are discussed here in some detail. Ramondeae (Fritsch, 1893–1894)— Originally, this tribe included the European genera (Ramonda Rich., Jancaea Boiss., Haberlea Pohl ex Barker), Corallodiscus, and Petrocosmea from Asia, and Saintpaulia from Africa. The key character for defining this group is straight fruits with septicidal dehiscence. Wang et al. (1990) redefined this tribe to include all OW genera with actinomorphic flowers (Ramonda, Thamnocharis W.T.Wang, Tengia Chun, Bournea Oliv., and Conandron). Neither definition ← Fig. 2. Strict MP consensus tree of 29 most parsimonious trees of 3275 steps based on combined ITS and trnL-F sequences of 89 sampled of advanced Asiatic genera rooted on African genera (CI = 0.43, RI = 0.63, RC = 0.27). Bootstrap values and decay indices (bold and italics) are given above branches. Light gray boxes = Trichosporeae, dark gray boxes = Cyrtandreae; systematics (right side) follows Weber 2004; open bars = didymocarpoid. Branches in bold are supported by posterior probabilities of 1.00 (black) or 0.99 (gray) in the Bayesian analysis. Asterisks indicate generic types. 996 American Journal of Botany [Vol. 96 May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae 997 998 American Journal of Botany of the tribe is supported in our analyses (see also Möller et al., 1999). Saintpaulieae (Ivanina, 1965b)— This tribe included the straight-fruited African genera Saintpaulia, Carolofritschia Engl. (now Acanthonema: Burtt, 1982), Acanthonema, and Linnaeopsis Engl. (now Streptocarpus: Darbyshire, 2006). Our data show that the genera included in our analysis do not form a coherent group (Figs. 2, 3), but are scattered among Streptocarpus (see also Möller and Cronk, 1997, 2001a, b). Streptocarpeae (Fritsch, 1893–1894)— This tribe (with Phylloboea Benth., Boea, Ornithoboea, and Streptocarpus) was reduced to subtribe Streptocarpinae of tribe Didymocarpeae by Ivanina (1965b) and expanded to include also Paraboea, Rhabdothamnopsis, and Trisepalum. All, except most Streptocarpus species, are distributed in Asia and indeed form a coherent alliance in our analyses: the Boea group. The majority of Streptocarpus species, as currently circumscribed, occur in Africa, Madagascar, and the Comoro Islands (~140 species Hilliard and Burtt, 1971), but four species have been described from Asia: S. burmanicus Craib (Burma), S. orientalis (Thailand), S. sumatranus B.L.Burtt (Sumatra), and S. clarkeanus (Hemsl.) Hilliard & B.L.Burtt (China) (cf. Hilliard and Burtt, 1971). At first sight, the occurrence of Streptocarpus species in Asia seems to support the idea of a transcontinental distribution of a group comprising Streptocarpus and the Asiatic genera with twisted fruits. However, Hilliard and Burtt (1971) have doubted that the Asiatic species of Streptocarpus are placed in the correct genus. They placed these in the genus with considerable reservations because of the lack of distinguishing morphological features. Indeed, one of them, S. clarkeanus, has been recently transferred to Boea by Wang et al. (1990). Another, S. orientalis, was examined molecularly in the current study and cytologically by M. Kiehn (University of Vienna, unpublished data). Both the molecular and the cytological data (2n = 18) indicated that there is no relationship between the Asian S. orientalis and the African and Madagascan species. We believe that it is simply a question of time before the remaining Asiatic species follow suit. The African and Madagascan species of Streptocarpus plus the straight-fruited compatriots formed a highly supported monophyletic clade in our analyses. If retained in a future classification, tribe or subtribe Streptocarpeae (-inae) will most probably form a group comprising exclusively African and Madagascan genera and species. Rhynchotecheae (Ivanina, 1965a)— This tribe included Rhynchotechum and Isanthera Nees (included in Rhynchotechum by Burtt 1962a). Ivanina indeed was right to separate Rhynchotechum from the Cyrtandreae, but she still thought there would be a close relationship between the two tribes (Ivanina, 1965b: fig. 14). As shown here, this is not the case. Therefore, the indehiscent fruit characterizing Cyrtandra and Rhynchotechum has no taxonomic bearing. It is simply a convergence. [Vol. 96 Indication of para/polyphyly of some genera— Our data indicate that a number of genera are artificial and polyphyletic: Streptocarpus (see discussion above), Chirita, Henckelia, Briggsia, and possibly Paraboea, Boea, and Raphiocarpus. Chirita—Species of this large genus (see revisions of Wood, 1974, and Wang, 1985a, b) appear distributed over at least five clades in our analysis. Some conform to current sections (sects. Chirita, Microchirita, Gibbosaccus, Liebigia), but others do not (C. lacunosa). Because of its straight fruits, C. lacunosa (S Thailand, NW Peninsular Malaysia) seems misplaced among the predominantly twisted-fruited Boea group. But it must be remembered that some species of Paraboea also possess straight fruits (similar to non-Streptocarpus taxa among the African and Madagascan taxa), suggesting that loss of fruit twist can occur frequently. The polyphyly of Chirita species over different parts of the didymocarpoid Gesneriaceae, however, is not erratic, but strongly indicates the artificial nature of the genus and requires a thorough (re) examination and (re)definition of the entire genus and its future segregates. Henckelia—This genus was reestablished when splitting the large genus Didymocarpus into smaller and more natural entities (Weber and Burtt, 1998b). As to the present data, the split of Didymocarpus proves more than justified. Neither ‘African’ nor ‘South Indian’ nor ‘Malesian Didymocarpus’ (as discussed in Weber and Burtt, 1998b) are parts of, or closely allied to ‘Didymocarpus s.s.’, which has its center of distribution in the Himalaya region, but a possible Malay Peninsula origin (Palee et al., 2006). Because ‘South Indian’ and ‘Malesian’ Didymocarpus as well as Loxocarpus R.Br. appeared somewhat linked, the genus Henckelia was revived and used in an inclusive sense for their accommodation (Weber and Burtt, 1998b). The present data give a first indication, though, that Henckelia is not monophyletic. Briggsia—This genus has over 20 species, ranging from the Himalayas to Vietnam. Species number, the wide distribution, and the wide range of morphologies (including rosette and caulescent habits, differences in indumentum and anther morphology) allow the prediction that this rather variable taxon is not monophyletic. Therefore, it is not too surprising that our molecular data confirm this prediction. Paraboea—This genus appears to be paraphyletic, with Trisepalum sharing the polytomy with Paraboea clades and species in our analysis 3 (Fig. 3). The close relationship of Paraboea and Trisepalum is undisputed because both have the characteristic arachnoid indumentum. Boea—The species analyzed formed two clades in analysis 3, with Spelaeanthus seemingly linked to the Australian / Papua New Guinean Boea species in one strongly supported clade, the other in a basal polytomy of the Boea group. However, more ← Fig. 3. Strict MP consensus tree of 10977 most parsimonious trees of 5694 steps based on combined ITS, trnL-F and atpB-rbcL sequences of 202 samples (CI = 0.47, RI = 0.75, RC = 0.35). Bootstrap values and decay indices (bold and italics) are given above branches. Light gray boxes = Trichosporeae, dark gray boxes = Cyrtandreae; black bars = outgroups, hatched bars = gesnerioid, lined bars = coronantheroid, open bars = didymocarpoid; numbers in bars = number of anthers; p, posterior pair, rest with two anterior anthers. Branches in bold are supported by posterior probabilities of 1.00 (black) or 0.99 (gray) in the Bayesian analysis. Asterisks indicate generic types. Arrows indicate relationship suggested in the Bayesian analysis. May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae 999 Fig. 4. Plots of branch support values, (A–C) boostrap vs. decay index and (D–F) posterior probabilities of (A, D) the combined trnL-F and ITS; (B, E) the trnL-F and atpB-rbcL, and (C, F) the trnL-F, atpB-rbcL, and ITS data sets (bootstrap values >50% included only). Only decay values >5 receive constantly >90% bootstrap values (vertical dotted red lines, top). In the bottom graphs, the diagonal line indicates the hypothetical 1 : 1 relationship between bootstrap and posterior probabilities. However, in the trnL-F and ITS analysis, branches with boostrap values as low as 51% receive PP values of 1.0 (arrow), for the trnL-F and atpB-rbcL dataset boostrap values as low as 65%, and for the three-gene analysis branches with boostrap values as low as 51% receive PP values of 0.99 (arrows), illustrating the overoptimism of PP values. species need to be included and missing data completed in a future analysis to elucidate the exact relationships in the Boea group as a whole. Raphiocarpus— This genus, revived by Weber and Burtt (1998a) for the accommodation of a number of Sino-Vietnamese species of Didissandra, was noted as potentially artificial even at its inception. Moreover, the identity of the Raphiocarpus sp. samples included in our analysis cannot be ascertained at present; these may well be misidentified. Thus, a more detailed analysis is needed for clarification here. require a detailed morphological discussion of many groups which cannot be presented here. Informally, the didymocarpoid Gesneriaceae can be divided into three groups: (1) the ‘basal Asiatic and European group’; (2) the ‘African and Madagascan group’; (3) the ‘advanced Asiatic and Malesian group’. Our phylogenetic results support that the phytogeography and distribution patterns form an essential component in the understanding of the evolutionary diversification of the didymocarpoid Gesneriaceae. Now we have to discuss whether these groupings make sense in morphological respects. Toward a new classification— As discussed, none of the available classifications and definition of tribes agree with the data presented here. Obviously, a revised classification needs to be established. In view of the many isolated genera and cascades of clades (= grades), a formal classification will prove difficult at this point. In the present paper, we do not attempt to establish such a formal classification, particularly because the position of a considerable number of genera (over 30) is still unknown or uncertain. Such a revised classification would also ‘Basal Asiatic and European group’—This group comprises a number of mono- or oligogeneric clades that form grades or polytomies at the base of the didymocarpoid Gesneriaceae clade. It is not a single taxonomic entity. Their morphology is varied, some are flat-rosette plants (Jerdonia, Corallodiscus, Ramonda, Haberlea, Jancaea), some are tall, shrubby (lignescent) plants with decussate (rarely alternate) leaves (Tetraphyllum, Leptoboea, Boeica, Rhynchotechum), and one (Platystemma) is a unifoliate herb. Floral form includes American Journal of Botany 1000 Table 1. Comparison of different branch support values for branches between genera (intergeneric), branches supporting genera (generic, except nonmonophyletic genera) and within genera (intrageneric) for the three-gene analysis of combined trnL-F, atpb-rbcL, and ITS data. Branch support values Index Intergeneric Generic Intrageneric Decay index SE N 3.57 0.40 82 10.12 2.16 33 3.84 0.45 38 Bootstrap SE Branches <50% support N 82.78 2.04 18 64 94.85 1.91 0 33 88.28 2.19 2 36 Posterior probability SE Branches <0.5 support N 0.96 0.01 13 69 0.99 0.01 1 32 0.97 0.01 1 37 Notes: SE = standard error, N = number of branches. a full range from strongly zygomorphic with a distinct tube (Corallodiscus) through slightly zygomorphic with a short, broad tube (‘campanulate’ = e.g., Jancaea, Rhynchotechum) to (sub)actinomorphic and flat-faced with very short tube (‘saintpaulioid’ = Ramonda, Platystemma). Thus, morphologically, there is little evidence that the basal Asiatic and European Gesneriaceae belong together or that they are particularly primitive. However, there are morphological features that are characteristic for the group and (at least partly) can be qualified as plesiomorphic: (1) Presence of four fertile stamens (tetrandry). Tetrandy is characteristic of the coronantheroid, gesnerioid, and epithematoid (except for Epithema Blume and Asian Rhynchoglossum Blume, which possess two fertile anthers in posterior positions, unlike all other diandrous didymocarpoid taxa where the anthers are in anterior position. This character state difference for taxa with two stamens was also overlooked by Smith, 1996) and is clearly ancestral within the family (Fig. 3). The presence of four stamens is also characteristic for the European group (except Ramonda with five). The African and Madagascan group (except for Acanthonema) and the twisted-fruited advanced Asiatic and Malesian group stamen number is always two, the derived state, while in the straight-fruited advanced Asiatic and Malesian group both tetrandrous and diandrous genera/groups occur. (2) Straight capsules with septicidal dehiscence (sometimes combined with loculicidal dehiscence). This type of fruit opening has already been regarded as primitive by Fritsch (1893–1894). This assumption was merely for morphological reasons (dehiscence along the lines of carpel fusion), but can be supported by the fact that it also occurs in presumed basal members of the New World coronantheroid Gesneriaceae (Coronanthera Vieill. ex C.B.Clarke, Depanthus S.Moore; combined with loculicidal dehiscence in Negria F.Muell. and Rhabdothamnus A. Cunn.) and in the basal gesnerioid tribe Beslerieae (Anetanthus Hiern ex Benth. & Hook f. and probably others). (3) Seed cell ornamentation. As can be concluded from the work of Beaufort-Murphy (1983) and personal observations of the last author (A.W.), the genera of the basal Asiatic and European group have seeds without testa cell ornamentation. Reticulate and striate seeds, without ornamentation on the cell surfaces, are also characteristic of the coronantheroid, gesnerioid, and epithematoid Gesneriaceae and are thus considered [Vol. 96 plesiomorphic. These groups also share a twisted testa cell arrangement. Ornamented seeds are found frequently in the African and Madagascan group as well as in the advanced Asiatic and Malesian group and are therefore derived. Thus, the basal Asiatic and European genera with plesiomorphic unornamented seeds and straight testa cell arrangement form a perfect transition to the higher groups of didymocarpoid Gesneriaceae. ‘African and Madagascan group’—While in earlier classifications the African and Madagascan genera appeared scattered over several tribes, Burtt suggested on many occasions (e.g., in Hilliard and Burtt, 1971) that they are closely related. In all our phylogenetic trees, they indeed formed a well-supported clade and can be considered a coherent taxonomic group. Despite the huge variation in vegetative and floral characters (even in the single genus Streptocarpus in the restricted sense), they are tied together by at least three synapomorphies: (1) Diandry. (2) Twisted fruits, with loculicidal dehiscence; the most parsimonious interpretation of the straight capsules found in Acanthonema, Colpogyne, Hovanella, Saintpaulia, and Schizoboea suggests that they are independent reversals to the ancestral state in the family. (3) Ornamented seeds with verruculose surface pattern (but with a reversal to the ancestral, reticulate state in most derived taxa in Streptocarpus subg. Streptocarpus). Before taxonomic changes can be made concerning this group, the status of the Asian Streptocarpus species needs to be addressed (i.e., their unrelatedness to African Streptocarpus species demonstrated). Advanced Asiatic and Malesian group—This group is the largest of the three groups. It needs detailed discussion, which is postponed to a future, more complete analysis. This group comprises, inter alias, the well-known and species-rich genera Didymocarpus, Aeschynanthus, and Cyrtandra. These genera, the leading genera of different tribes in former classifications, thus prove surprisingly closely related. Their characteristic features, including the appendaged seeds of Aeschynanthus and the indehiscent fruits of Cyrtandra, are obviously convergences, characterizing small groups or single genera, but have no major classificatory significance. Like the basal Asiatic and European group, it consists of a number of larger clades, starting with Didissandra (sensu Weber and Burtt, 1998a). This genus has tetrandrous flowers, capsules of a very special type (tardily loculicidally dehiscent, with the valves finally disintegrating along the sclerified vascular bundles (see Weber and Burtt, 1998a), and seeds with mostly knobby ornaments along the testa cell margins (see Sontag and Weber, 1998). A well-supported major clade includes the genera with predominantly twisted capsules, the Boea group (Boea, Emarhendia, Kaisupeea, Ornithoboea, Paraboea p.p., Rhabdothamnopsis, Trisepalum; only Senyumia Kiew, A.Weber & B.L.Burtt has not been included in the analysis, but probably has also its place here), but scattered within this group are also species with straight fruits (Paraboea p.p., Chirita lacunosa, Henckelia ericii). This group will be discussed in more detail in a forthcoming paper. The remainder of clades in the advanced Asiatic and Malesian group are not strongly supported, though many monophyletic genera are (Fig. 3), and the intergeneric relationships of this group will be addressed comprehensively in another publication. Phytogeography— It is still enigmatic in which part of the world Gesneriaceae originated. Burtt (1998) proposed the hypothesis that the family is of southern hemisphere (Gondwana) origin, with the coronantheroid Gesneriaceae representing a sur- May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae vivors of the oldest group. This gave rise both to the gesnerioid Gesneriaceae (invading South America via the Antarctic and southern South America) and the didymocarpoid Gesneriaceae (by migrating northwards, ‘dropping’ representatives in Africa and Madagascar and finally reaching the Eurasiatic continent and spreading from there to the Sunda Islands and the Pacific). This hypothesis faces difficulties both from the geological time scale and the molecular data. The age of the family has been variously estimated, from 65 million years (Raven and Axelrod (1974) to 71 million years (Bremer et al., 2004, though Peltanthera was used erroneously as a member of Gesneriaceae here, thus the family maybe younger), but nowhere near a Gondwana origin (Gondwana break-up began 150 Ma, Storey et al. 1995). As far as the didymocarpoid Gesneriaceae are concerned, the most basal members are found on the Eurasian continent, especially on the Indian subcontinent: Jerdonia (mountains of SW India), Corallodiscus (Himalayas and China), Tetraphyllum, Leptoboea, Boeica (Himalayas and adjacent areas). Only Rhynchotechum, with around 15 species being the largest genus of the group, has a wider distribution. It spreads from the Himalayas to the Malay archipelago and one species even reaches New Guinea. This island spreading would have been aided by the possession of fleshy fruits of the genus, as in Cyrtandra, that are putatively bird dispersed (Cronk et al., 2005). The basal Asiatic genera thus can be marked as an essentially ‘Indian group’ with most genera represented in the Sino-Himalayan area. This point is where the molecular and phytogeographical data and parts of Burtt’s hypothesis meet. These genera may well be relicts of a group that had its origin on the Indian plate. Transgression to the west (Europe) was apparently very early, followed by a transgression to the south (perhaps first Madagascar and then Africa; the highest morphological diversity occurs in Madagascar, and a subset may have entered the African mainland 25–35 Ma; Möller and Cronk, 2001b) and—under explosive radiation and diversification—to the east and southeast (Indochina, Malesia, Pacific). The case of the widespread Rhynchotechum of the basal Asiatic and European group suggests that a migration from the Himalayas to the east may have involved several lineages in parallel. Conclusions— The present work represents a major step forward in our understanding of the largest group of Old World Gesneriaceae, the ‘didymocarpoid Gesneriaceae’. The molecular data strongly suggest that none of the available current classifications properly reflect phylogeny. Neither the classification into tribes nor the delimitation of the tribes established so far is reflected in the molecular phylogenies. All characters that have been considered to be of major taxonomic value, such as actinomorphic flowers, diandry, indehiscent fruits, appendaged seeds, were found to be homoplastic, i.e., having evolved several times independently. Thus, apart from perhaps fruit twist, most other morphological characters and states are not helpful in shaping a new classification for the family. The informal classification proposed by Weber (2004), which was partly based on unpublished molecular data, is essentially confirmed here on a much larger data set. No formal taxonomy is presented at this point, but a new classification and an understanding of the evolutionary pathways is emerging, and it is the biogeography that is best reflected in the resulting phylogenies so far. This study will provoke a reappraisal of the approach for a classification of the OW Gesneriaceae if it is to reflect our results obtained here. Even though the epithematoid, basal Asiatic and European, and African and Madagascan groups are 1001 well defined and supported, relationships among the advanced Asiatic and Malesian taxa are far from resolved. It is now paramount to obtain a stable phylogeny for this group of Gesneriaceae by adding more taxa and more data. At present, molecular data for a considerable number of genera of didymocarpoids are still lacking. The bulk will be relevant for the advanced Asiatic and Malesian Gesneriaceae. We are confident that their inclusion will result in a better resolution of this group and stabilize relationships. Their addition will also contribute to a better definition of particular genera and solve problems of generic delimitation. Even now it is clear that some genera are not monophyletic, that some genera have to be reduced to synonymy and that new genera have to be established. Much work is waiting. LITERATURE CITED Akaike, H. 1974. A new look at the statistical model identification. IEEE Transactions on Automatic Control 19: 716–723. Alfaro, M. E., S. Zoller, and F. Lutzoni. 2003. Bayes or bootstrap? A simulation study comparing the performance of Bayesian Markov chain Monte Carlo sampling and bootstrapping in assessing phylogenetic confidence. Molecular Biology and Evolution 20: 255–266. APG [Angiosperm Phylogeny Group] II. 2003. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141: 399–436. Barker, F. K., and F. M. Lutzoni. 2002. The utility of the incongruence length difference test. Systematic Biology 51: 625–637. Beaufort-Murphy, H. T. 1983. The seed surface morphology of the Gesneriaceae utilizing the scanning electron microscope and a new system for diagnosing seed morphology. Selbyana 6: 220–422. Bentham, G. 1876. Gesneriaceae. In G. Bentham and J. D. Hooker [eds.], Genera plantarum, vol. 2, part 2, 990–1025. Lovell Reeve & Co., London, UK. Bremer, K. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795–803. Bremer, B., K. Bremer, N. Heidari, P. Erixon, R. G. Olmstead, A. A. Anderberg, M. Källersjö, and E. Barkhordariana. 2002. Phylogenetics of asterids based on 3 coding and 3 non-coding chloroplast DNA markers and the utility of non-coding DNA at higher taxonomic levels. Molecular Phylogenetics and Evolution 24: 274–301. Bremer, K., E. M. Friis, and B. Bremer. 2004. Molecular phylogenetic dating of asterid flowering plants shows early Cretaceous diversification. Systematic Biology 53: 496–505. Burtt, B. L. 1954. Studies in the Gesneriaceae of the Old World I. General introduction. Notes from the Royal Botanic Garden Edinburgh 21: 185–192. Burtt, B. L. 1962. Studies in the Gesneriaceae of the Old World. XXI. Rhynchotechum and Isanthera. Notes from the Royal Botanic Garden Edinburgh 24: 35–39. Burtt, B. L. 1963 (‘1962’). Studies in the Gesneriaceae of the Old World. XXIV. Tentative keys to the tribes and genera. Notes from the Royal Botanic Garden Edinburgh 24: 205–220. Burtt, B. L. 1977. Classification above the genus, as exemplified by Gesneriaceae, with parallels from other groups. Plant Systematics and Evolution. Supplementum 1: 97–109. Burtt, B. L. 1982. New Gesneriaceae from tropical West Africa. Bulletin Museum National d’Histoire Naturelle Paris 4, série 3, 1981, sect. B, Adansonia 4: 415–417. Burtt, B. L. 1997. Old World Gesneriaceae V. Suprageneric names. Edinburgh Journal of Botany 54: 85–90. Burtt, B. L. 1998. Climatic accommodation and phytogeography of the Gesneriaceae of the Old World. In P. Mathew and M. Sivadasan [eds.], Diversity and taxonomy of tropical flowering plants, 1–27. Mentor Books, Kerala, India. 1002 American Journal of Botany Burtt, B. L., and H. Wiehler. 1995. Classification of the family Gesneriaceae. Gesneriana 1: 1–4. Cronk, Q. C. B., M. Kiehn, W. Wagner, and J. F. Smith. 2005. Evolution of Cyrtandra (Gesneriaceae) in the Pacific Ocean: The origin of a supertramp clade. American Journal of Botany 92: 1017–1024. Cummings, M. P., S. A. Handley, D. S. Myers, D. L. Reed, A. Rokas, and K. Winka. 2003. Comparing bootstrap and posterior probability values in the four-taxon case. Systematic Biology 52: 477–487. Cunningham, C. W. 1997. Can three incongruence tests predict when data should be combined? Molecular Biology and Evolution 14: 733–740. Darbyshire, I. 2006. Gesneriaceae. In H. J. Beentje and S. A. Ghazanfar [eds.], Flora of tropical East Africa, 1–76. Royal Botanic Gardens Kew, London, UK. Doyle, J. J., and J. L. Doyle. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemistry Bulletin 19: 11–15. Doyle, J. J., and J. L. Doyle. 1990. Isolation of plant DNA from fresh tissue. Focus12: 13–15. Eriksson, T. 1999. AutoDecay, version 4.0. Computer program distributed by the author. Bergius Foundation, Royal Swedish Academy of Sciences, Stockholm, Sweden. Erixon, P., B. Svennblad, T. Britton, and B. Oxelman. 2003. Reliability of Bayesian posterior probabilities and bootstrap frequencies in phylogenetics. Systematic Biology 52: 665–673. Farris, J. S. 1989a. The retention index and homoplasy excess. Systematic Zoology 38: 406–407. Farris, J. S. 1989b. The retention index and the rescaled consistency index. Cladistics 5: 417–419. Farris, J., S. M. Källersjö, A. G. Kluge, and C. Bult. 1994. Testing significance of incongruence. Cladistics 10: 315–319. Farris, J., S. M. Källersjö, A. G. Kluge, and C. Bult. 1995. Constructing a significance test for incongruence. Systematic Biology 44: 570–572. Felsenstein, J. 1978. Cases in which parsimony and compatibility methods will be positively misleading. Systematic Zoology 27: 401–410. Felsenstein, J. 1985. Confidence-limits on phylogenies—An approach using the bootstrap. Evolution 39: 783–791. Fritsch, K. 1893–1894. Gesneriaceae. In A. Engler and K. Prantl [eds.], Die Natürlichen Pflanzenfamilien, vol. 4, part 3b, 133–185. Engelmann, Leipzig, Germany. Hilliard, O. M. 2004. A revision of Chirita sect. Liebigia (Gesneriaceae). Edinburgh Journal of Botany 60: 361–387. Hilliard, O. M., and B. L. Burtt. 1971. Streptocarpus. An African plant study. University of Natal Press, Pietermaritzburg, South Africa. Hipp, A. L., J. C. Hall, and K. J. Sytsma. 2004. Phylogenetic accuracy, congruence between data partitions, and performance of the ILD. Systematic Biology 53: 81–89. Huelsenbeck, J. P., and F. Ronquist. 2001. MrBayes: Bayesian inference of phylogeny. Bioinformatics 17: 754–755. Huelsenbeck, J. P., and F. Ronquist. 2007. MrBayes, version 3.1.2. Bayesian analysis of phylogeny. Application program distributed by the authors under the GNU General Public License, website http:// mrbayes.csit.fsu.edu. Ivanina, L. I. 1965a. Primencenie karpologisheskogo metoda sistematke semeistva Gesneriaceae. Application of carpological method to taxonomy of the family Gesneriaceae. Botanicheski Zhurnal 50: 29–43 (in Russian). Ivanina, L. I. 1965b. Application of the carpological method to the taxonomy of Gesneriaceae. Notes from the Royal Botanic Garden Edinburgh 26: 383–402. Ivanina, L. I. 1967. The family Gesneriaceae (The carpological review). Komarov Institute, Leningrad, Russia (in Russian). Kearney, M. 2002. Fragmentary taxa, missing data, and ambiguity: Mistaken assumptions and conclusions. Systematic Biology 51: 369–381. Kluge, A. G., and J. S. Farris. 1969. Quantitative phyletics and the evolution of anurans. Systematic Zoology 18: 1–32. Larkin, M. A., G. Blackshields, N. P. Brown, R. Chenna, P. A. McGettigan, H. McWilliam, F. Valentin, et al. 2007. Clustal [Vol. 96 W and Clustal X, version 2.0. Bioinformatics (Oxford, England) 23: 2947–2948. Lewis, P. O. 2001. Phylogenetic systematics turns over a new leaf. Trends in Ecology & Evolution 16: 30–37. Long, D. G., M. Möller, and J. Preston. 2000. Phylogenetic relationships of Asterella (Aytoniaceae, Marchantiopsida) inferred from chloroplast DNA sequences. The Bryologist 103: 625–644. Mayer, V., M. Möller, M. Perret, and A. Weber. 2003. Phylogenetic position and generic differentiation of Epithemateae (Gesneriaceae) inferred from cpDNA sequence data. American Journal of Botany 90: 321–329. Middleton, D. J. 2007. A revision of Aeschynanthus (Gesneriaceae) in Thailand. Edinburgh Journal of Botany 64: 363–429. Möller, M., M. Clokie, P. Cubas, and Q. C. B. Cronk. 1999. Integrating molecular phylogenies and developmental genetics: A Gesneriaceae case study. In P. M. Hollingsworth, R. M. Bateman, and R. J. Gornall [eds.], Molecular systematics and plant evolution, 375–402. Taylor & Francis, London, UK. Möller, M., and Q. C. B. Cronk. 1997. Origin and relationships of Saintpaulia (Gesneriaceae) based on ribosomal DNA internal transcribed spacer (ITS) sequences. American Journal of Botany 84: 956–965. Möller, M., and Q. C. B. Cronk. 2001a. Evolution of morphological novelty: A phylogenetic analysis of growth patterns in Streptocarpus (Gesneriaceae). Evolution 55: 918–929. Möller, M., and Q. C. B. Cronk. 2001b. Phylogenetic studies in Streptocarpus (Gesneriaceae): Reconstruction of biogeographic history and distribution patterns. Systematics and Geography of Plants 71: 545–555. Murphy, W. J., E. Eizirik, W. E. Johnson, Y. P. Zhang, O. A. Ryder, and S. J. O’Brien. 2001. Molecular phylogenetics and the origins of placental mammals. Nature 409: 614–618. Nixon, K. C. 1999. The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15: 407–414. Olmstead, R. G., K. J. Kim, R. K. Jansen, and S. J. Wagstaff. 2000. The phylogeny of the Asteridae sensu lato based on chloroplast ndhF gene sequences. Molecular Phylogenetics and Evolution 16: 96–112. Oxelman, B., M. Backlund, and B. Bremer. 1999. Relationships of the Buddlejaceae s. 1. investigated using parsimony jackknife and branch support analysis of chloroplast ndhF and rbcL sequence data. Systematic Botany 24: 164–182. Palee, P., J. Denduangboripant, V. Anusarnsunthorn, and M. Möller. 2006. Molecular phylogeny and character evolution of Didymocarpus (Gesneriaceae) in Thailand. Edinburgh Journal of Botany 63: 231–251. Perret, M., A. Chautems, R. Spichiger, G. Kite, and V. Savolainen. 2003. Systematics and evolution of tribe Sinningieae (Gesneriaceae): Evidence from phylogenetic analyses of six plastid DNA regions and nuclear ncpGS. American Journal of Botany 90: 445–460. Posada, D., and K. A. Crandall. 1998. ModelTest: Testing the model of DNA substitution. Bioinformatics (Oxford, England) 14: 817–818. Qiu, Y.-L., J. Lee, F. Bernasconi-Quadroni, D. E. Soltis, P. S. Soltis, M. Zanis, E. A. Zimmer, et al. 1999. The earliest angiosperms: Evidence from mitochondrial, plastid and nuclear genomes. Nature 402: 404–407. Raven, P. H., and D. I. Axelrod. 1974. Angiosperm biogeography and past continental movements. Annals of the Missouri Botanical Garden 61: 539–673. Roalson, E. H., J. K. Boggan, and L. E. Skog. 2005a. Reorganization of tribal and generic boundaries in the Gloxinieae (Gesneriaceae: Gesnerioideae) and the description of a new tribe in the Gesnerioideae, Sphaerorrhizeae. Selbyana 25: 225–238. Roalson, E. H., J. K. Boggan, L. E. Skog, and E. A. Zimmer. 2005b. Untangling Gloxinieae (Gesneriaceae). I. Phylogenetic patterns and generic boundaries inferred from nuclear, chloroplast, and morphological cladistic data sets. Taxon 54: 389–410. Sanderson, M. J., and H. B. Shaffer. 2002. Troubleshooting molecular phylogenetic analyses. Annual Review of Ecology and Systematics 33: 49–72. May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae Savolainen, V., and M. W. Chase. 2003. A decade of progress in plant molecular phylogenetics. Trends in Genetics 19: 717–724. Sikes, D. S., and P. O. Lewis. 2001. PAUPRat beta software, version 1: PAUP* implementation of the parsimony ratchet. Computer program distributed by the authors. Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, USA. Sinclair, W. T., R. R. Mill, M. F. Gardner, P. Woltz, T. Jaffré, J. Preston, M. L. Hollingsworth, et al. 2002. Evolutionary relationships of the New Caledonian heterotrophic conifer, Parasitaxus usta (Podocarpaceae), inferred from chloroplast trnL-F intron/spacer and nuclear rDNA ITS2 sequences. Plant Systematics and Evolution 233: 79–104. Smith, J. F. 1996. Tribal relationships within Gesneriaceae: A cladistic analysis of morphological data. Systematic Botany 21: 497–513. Smith, J. F. 2000. Phylogenetic resolution within the tribe Episcieae (Gesneriaceae): Congruence of ITS and ndhF sequences from parsimony and maximum-likelihood analyses. American Journal of Botany 87: 883–897. Smith, J. F., K. D. Brown, C. L. Carroll, and D. S. Denton. 1997a. Familial placement of Cyrtandromoea, Titanotrichum and Sanango, three problematic genera of the Lamiales. Taxon 46: 65–74. Smith, J. F., J. C. Wolfram, K. D. Brown, C. L. Carroll, and D. S. Denton. 1997b. Tribal relationships in the Gesneriaceae: Evidence from DNA sequences of the chloroplast gene ndhF. Annals of the Missouri Botanical Garden 84: 50–66. Sontag, S., and A. Weber. 1998. Seed coat structure in Didissandra, Ridleyandra and Raphiocarpus (Gesneriaceae). Beiträge zur Biologie der Pflanzen 70: 179–190. Spangler, R. E., and R. G. Olmstead. 1999. Phylogenetic analysis of Bignoniaceae based on the cpDNA gene sequences rbcL and ndhF. Annals of the Missouri Botanical Garden 86: 33–46. Storey, M., J. J. Mahoney, A. D. Saunders, R. A. Duncan, S. P. Kelley, and M. F. Coffin. 1995. Timing of hot spot-related volcanism and the break-up of Madagascar and India. Science 267: 852–855. Swofford, D. L. 2002. PAUP*: Phylogenetic analysis using parsimony (*and other methods), version 4. Sinauer, Sunderland, Massachusetts, USA. Suzuki, Y., G. V. Glazko, and M. Nei. 2002. Overcredibility of molecular phylogenies obtained by Bayesian phylogenetics. Proceedings of the National Academy of Sciences, USA 99: 16138–16143. Taberlet, P., L. Gielly, G. Pautou, and J. Bouvet. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105–1109. Tank, D. C., P. M. Beardsley, S. A. Kelchner, and R. G. Olmstead. 2006. Review of the systematics of Scrophulariaceae s.l. and their current disposition. Australian Systematic Botany 19: 289–307. Van Royen, P. 1983. Gesneriaceae. In J. Cramer [ed.], The alpine flora of New Guinea, vol. 4, 2939–2964. A. R. Gantner Verlag KG, Vaduz, Liechtenstein. Wang, W. T. 1985a. A revision of the genus Chirita in China (I). Bulletin of Botanical Research 5: 71–97. Wang, W. T. 1985b. A revision of the genus Chirita in China (II). Bulletin of Botanical Research 5: 37–86. 1003 Wang, C.-N., M. Möller, and Q. C. B. Cronk. 2004. Phylogenetic position of Titanotrichum oldhamii (Gesneriaceae) inferred from four different gene regions. Systematic Botany 29: 407–418. Wang, W. T., K. Y. Pan, and Z. Y. Li. 1990. Gesneriaceae. In W.T. Wang [ed.], Flora Reipublicae Popularis Sinicae, vol. 69, 125–581. Science Press, Beijing, China. Wang, W. T., K. Y. Pan, and Z. Y. Li. 1992. Keys to the Gesneriaceae of China. Edinburgh Journal of Botany 49: 5–74. Weber, A. 2004. Gesneriaceae. In K. Kubitzki [ed.], The families and genera of vascular plants, vol. 7. J. W. Kadereit [vol. ed.], Dicotyledons. Lamiales (except Acanthaceae incl. Avicenniaceae), 63–158. Springer, Berlin & Heidelberg, Germany. Weber, A. and B. L. Burtt. 1998a: Didissandra: Redefinition and partition of an artificial genus of Gesneriaceae. Beiträge zur Biologie der Pflanzen 70: 153–177 (‘1997’, publ. 26. XI. 1998). Weber, A., and B. L. Burtt. 1998b. Remodelling of Didymocarpus and associated genera (Gesneriaceae). Beiträge zur Biologie der Pflanzen 70: 293–363 (‘1997’, publ. 26. XI. 1998). Weber, A., and L. E. Skog. 2007. The genera of Gesneriaceae. Basic information with illustration of selected species. http//:www.generagesneriaceae.at White, T. J., T. Bruns, S. Lee, and J. W. Taylor. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White [eds.], PCR protocols, 315–322. Academic Press, London, UK. Wiehler, H. 1983. A synopsis of the neotropical Gesneriaceae. Selbyana 6: 1–219. Wiens, J. J. 2003. Missing data, incomplete taxa, and phylogenetic accuracy. Systematic Biology 52: 528–538. Wiens, J. J. 2006. Missing data and the design of phylogenetic analyses. Journal of Biomedical Informatics 39: 34–42. Wiens, J. J., and T. W. Reeder. 1995. Combining data sets with different numbers of taxa for phylogenetic analysis. Systematic Biology 44: 548–558. Wight, R. 1850. Gesneraceae. In Illustrations of Indian botany, vol. 2, 180–182, plates 159–159b. P.R. Hunt, Madra, India. Wilcox, T. P., D. J. Zwickl, T. Heath, and D. M. Hillis. 2002. Phylogenetic relationship of the dwarf boas and a comparison of Bayesian and bootstrap measures of phylogenetic support. Molecular Phylogenetics and Evolution 25: 361–371. Wood, D. 1974. A revision of Chirita (Gesneriaceae). Notes from the Royal Botanic Garden Edinburgh 33: 123–205. Yang, Z., and B. Rannala. 1997. Bayesian phylogenetic inference using DNA sequences: A Markov chain Monte Carlo method. Molecular Biology and Evolution 14: 717–724. Yuan, Y.-M., S. Wohlhauser, M. Möller, J. Klackenberg, M. W. Callmander, and P. Küpfer. 2005. Phylogeny and biogeography of Exacum (Gentianaceae): A disjunctive distribution in the Indian Ocean Basin resulted from long distance dispersal and extensive radiation. Systematic Biology 54: 21–34. Zimmer, E. A., E. H. Roalson, L. E. Skog, J. K. Boggan, and A. Idnurm. 2002. Phylogenetic relationships in the Gesnerioideae (Gesneriaceae) based on nrDNA and cpDNA trnL-F and trnE-T spacer region sequences. American Journal of Botany 89: 296–311. 1004 American Journal of Botany [Vol. 96 Appendix 1. List of taxa included in the phylogenetic analysis of Gesneriaceae, including voucher number, deposition of voucher, origin and GenBank sequence information.(An asterisk [*] denotes generic types; [C] = Chirita section Chirita; [G] = Chirita section Gibbosaccus; [M] = Chirita section Microchirita; [L] = Chirita section Liebigia.) Taxon Voucher number Deposited in A.Dahl 703 (trnL-F); cult RBGE 19832265 (atpB-rbcL) A.Dahl702 (trnL-F); cult RBGE 19687553 (atpB-rbcL) E E Tetrachondraceae Polypremum procumbens L. Tetrachondra patagonica Skottsb. Struwe 1000 Martinsson and Swenson 314 UPS UPS — — Scrophulariaceae Scrophularia canina L. Verbascum speciosum Opiz Perret S1.119 cult. HBV; Kiehn s.n. G WU Oleaceae Olea europaea L.* Forsythia ×intermedia Zabel Plantaginaceae Antirrhinum majus L.* Tetranema roseum (Martens & Galeotti) Standl. & Steyerm. [= T. mexicanum Benth. ex Lindl.*] Veronica incana L. Veronica incana L. Calceolariaceae Calceolaria arachnoidea Graham Jovellana punctata Ruiz & Pavon* cult. HBV cult. HB München, Dec. 1997; Kiehn s.n. (27.4.98) cult. HB Bonn ex HB Mainz no. XX-0-MJG-19–47720; Albach 155 unknown cult. RBGE 19912379 cult. RBGE 19980599 Gesnerioid Achimenes admirabilis Wiehler cult. CJBG, Nov. 1997, Chautems & Perret 01–033 cult. RBGE 19822666; Besleria labiosa Hanst. Wiehler & Steyermark 72453 Besleria melancholica (Vell.) C.V.Morton Chautems, Leitman & Matinelli 240, 21.5.1987 cult. RBGE 19802568; Chrysothemis pulchella (Donn ex Sims) Wiehler Decne.* cult. RGBE 19822676; Cobananthus calochlamys (Donn. Sm.) H.Wiehler 7553 Wiehler* cult. HBV, ex HB St. Gallen Columnea sanguinea (Pers.) Hanst. 1987; Kiehn s.n. cult. RBGE 19570361 Episcia cupreata (Hook.) Hanst. cult. RBGE 19782220 Eucodonia verticillata (Martens & Galeotti) Wiehler* Skog 7714 Chautems & Perret 97–020 Gesneria humilis L.* Mendoza-T. et al. 506 Glossoloma bolivianum (Britt.) Wiehler Chautems & Perret 01–034 Gloxinia erinoides (DC.) Roalson & Boggan cult. RBGE 19821486; Kohleria hirsuta (Kunth) Regel var. M.Koehnen hirsuta [K.eriantha (Benth.) Hanst.] Chautems & Perret 97–018 Kohleria spicata (Kunth) Oerst. Napeanthus reitzii (L.B.Smith) B.L.Burtt ex Leeuwenb. A. Cervi & al. AC479 Nautilocalyx melittifolius (L.) Wiehler Chautems & Perret 01–025 Nematanthus villosus (Hanst.) Wiehler Paliavana prasinata (Ker Gawl.) Benth. Rhytidophyllum tomentosum (L.) Mart.* Seemannia aff. purpurascens Rusby Sinningia cardinalis (Lehm.) H.E.Moore Sinningia schiffneri Fritsch Smithiantha lauii Wiehler Perret 99–041 Chautems & Perret 00–013 cult. HBV; Kiehn s.n. Chautems & Perret 97–019 Chautems & Perret 97–015 Chautems & Perret 97–010 cult. GRF (Gesneriad Research Foundation, Sarasota, Fl., USA), cult. ID = G-3588 Sequence by M. Perret Origin atpB-rbcL trnL-F ITS1/ITS2 Cyprus; Troodos FJ501367 AF231866 — cultivar FJ501368 AF231824 — — — AJ430938 AJ430939 — — — Austria, exact origin unknown AY423105 AJ490885 AY423123 AJ492271 — — no voucher unknown origin WU Mexico; exact origin unknown AJ490883 AJ490884 AJ492270 AJ492272 — — WU unknown origin — AY486449 — — unknown origin AY818908 — — E E Chile, Los Lagos Chile, Biobío, Prov. de Arauco: AY423108 AY423109 AY423126 AY423127 — — G unknown origin AJ439982 AJ439827 — E AY423110 AY423128 — AJ490923 AJ492310 — E Venezuela; Distrito Federal, Cerro Narguata. Brazil, Rio de Janeiro, Macaé de Cima unknown origin AJ490925 AJ492312 — E Guatemala, Coban AJ490926 AJ492313 — WU unknown origin AJ490927 AJ492314 — E E unknown origin unknown origin AJ490928 FJ501369 AJ492315 — — — USBRG 86–097 cult. CJBG, Oct. 1997 unknown origin cult. CJBG, Nov. 1997 — AJ439976 — AJ439983 AY047120 AJ439821 AY047156 AJ439828 — — — — E Ecuador, unknown locality AY423114 AY423132* — G cult. CJBG AJ439975, part AJ493036 AJ439820 — AJ492321 — AJ439984, part AJ439980 AJ490932 AJ490930 AJ439977 AJ490931 AJ439900 AJ439978 AJ439829 — AJ439825 AJ492319 AJ492317 AJ439822 AJ492318 AJ439745 AJ439823 — — — — — — — G US G US G UPCB G G G WU G G G G Brazil, PR, Morretes, Route de la Graciosa, Volta Grande unknown origin cult. CJBG, Sept. 1998 cult. CJBG unknown origin cult. CJBG, Oct. 1998 cult. CJBG, cult. CJBG Oct. 1997 unknown origin May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae 1005 Appendix 1. Continued. Taxon Vanhouttea calcarata Lem.* Coronantheroid Asteranthera ovata (Cav.) Hanst.* Fieldia australis Cunn.* Lenbrasssia australiana (C.T.White) G.W.Gillett var. australiana* Mitraria coccinea Cav.* Rhabdothamnus solandri Cunn.* Sarmienta scandens (J.D.Brandis) Pers.* Epithematoid Epithema benthamii C.B.Clarke Epithema membranaceum (King) Kiew Voucher number Carvalho et al. 526 Deposited in Origin atpB-rbcL trnL-F ITS1/ITS2 CEPEC Brazil, State of Rio de Janeiro, Nova Friburgo AJ490933 AJ492320 — Chile, Los Lagos, Prov. de FJ501371 Palena, Chaitén Australia, unknown AY423112 locality Australia, Queensland AJ490921 FJ501427 — AY423130 — AJ492308 — — AY423131 — FJ501370 FJ501426 — AJ490922 AJ492309 — Philippines, Luzon, Isabela AY423118 AY423135 — Peninsula Malaysia, Pahang, Jerantut distr. Peninsula Malaysia, Perak, Kinta distr.; Sg. Siput Selatan Taiwan, Kaohsiung Hsien: Tengshih Cameroon, Kupe village Peninsula Malaysia, Pahang; Pulau Tioman Peninsula Malaysia, near Sungai Siput Selatan Borneo, Sarawak, Bkt. Mentawa Peninsula Malaysia, Selangor, Gua Batu AJ490887 AJ492274 — AJ490888 AJ492275 — AJ490889, part AJ490890 AJ490891 AJ492276 — AJ492277 AJ492278 — — AJ490892 AJ492279 — AJ490893 AJ492280 — AJ490894 AJ492281 — U91315 AJ492269 — AJ490895 AJ492282 — AJ490896 AJ492283 — AJ490897 AJ492284 — AJ490898 AJ492285 — AJ490899 AJ492286 — AJ490900 AJ492287 — — FJ501428 — AJ490901 AJ492288 — AJ490902 AJ492289 — — FJ501396 FJ501454 FJ501500 FJ501306 — — — AF349218 / AF349299 — — FJ501501 — — AY047099 — AF349203 / AF349284 AY047040 cult. RBGE 19980608; UCEXC 362 RBGE 19696862 E cult. RBGE19970901; P.D.Hint 6654 cult. RBGE 19792696; M.Mason cult. RBGE 19660192 — cult. RBGE 19882757; M.Gardner & S.Knees 4033 E cult RBGE 19972563; Philippine Expedition 1997 SM9 Weber 860908–2/1 E E E E WU Epithema saxatile Blume Weber & Anthonysamy 870521–3/2 (WU); cult. HBV WU Epithema taiwanense S.S.Yin C.-N. Wang & al. TNU Epithema tenue C.B.Clarke Loxonia hirsuta Jack* cult. RBGE ex DTH 5815 Weber 870602–1/5 E WU Monophyllaea elongata B.L.Burtt Weber & Antonysamy 870518–1/1 Vogel & Weber 790106–1/1 WU Chin & Weber, Chin 2107 (KLU) = Vogel & Weber 790801 cult. HBV, seeds rec. from RBGE Huber & Weissenhofer 722 WU Monophyllaea glauca C.B.Clarke Monophyllaea hirticalyx Franch. Monophyllaea horsfieldii R.Br.* Rhynchoglossum azureum (Schltdl.) B.L.Burtt Rhynchochlossum notonianum (Wall.) B.L.Burtt Rhynchoglossum obliquum Blume* [Malay Peninsula] Rhynchoglossum obliquum Blume* [Philippines] cult. HB München Weber 870510–1/3 Mendum & al. 25349 C.-N. Wang & al. Rhynchoglossum obliquum Blume var. hologlossum (Hayata) W.T.Wang [Taiwan] Weber 870602–1/1 Stauranthera grandiflora Benth.* Whytockia purpurascens Y.Z.Wang MMO 01–87 Whytockia sasakii (Hayata) B.L.Burtt C.-N. Wang & al. Whytockia tsiangiana (Hand.-Mazz.) A.Weber 1986 Sino-Amer. Exedition Nr. 200 Didymocarpoid: Acanthonema strigosum Hook.f.* Aeschynanthus austroyunnanensis W.T.Wang Aeschynanthus austroyunnanensis W.T.Wang Aeschynanthus bracteatus Wall. ex DC. Aeschynanthus bracteatus Wall. ex DC. Aeschynanthus hildebrandtii Hemsl. B.Macinder 49 MMO 01–79 WU Argentina, Bariloche Region New Zealand, North Island Chile, Región X Los Lagos, Prov. de Osorno WU Peninsula Malaysia, Selangor, Batu Caves WU Costa Rica, Prov. Alajuela; Valle Virgen no voucher unknown origin WU E TNU WU E, WU TNU & E WU K E, WU Peninsula Malaysia, Pahang, Lipis district Philippines, Palawan, betw. San Rafael and Cleopatra Needle Taiwan, Kaohsiung Hsien: Tengshih Peninsula Malaysia, Pahang; Pulau Tioman China, Yunnan, Maguan county Taiwan, Hualian Hsien, Hsiulin Hsiang China, Guizhou Prov., Jiangkou county cult RBGE 19951561; A.Reid & J.Fernie 004 E Wang 991113 cult RBGE 19970165; R.Cherry 123 Skog 7777 PE E Cameroon, Kupe village China, Yunnan, Hekou county, China; Yunnan, Xishuangbanna Dai Aut. Pref. China, Yunnan, Xichou Viet Nam; Lao Cai US unknown origin American Journal of Botany 1006 [Vol. 96 Appendix 1. Continued. Taxon Voucher number Deposited in Aeschynanthus lancilimbus W.T.Wang Aeschynanthus longiflorus (Blume) DC. Wang S-10868 Weber 950905 (photo record) PE WU Aeschynanthus longiflorus (Blume) DC. Agalmyla biflora (Elmer) O.M.Hilliard & B.L.Burtt cult RBGE 19680624 cult. RBGE 19980287, DNA no. AG04 E E Agalmyla biflora (Elmer) O.M.Hilliard & B.L.Burtt Agalmyla clarkei (Elmer) B.L.Burtt E Agalmyla clarkei (Elmer) B.L.Burtt cult RBGE 19980292, RBGEPNHE1998–25517 cult. RBGE 19991911, Royal Botanic Garden EdinburghPhilippine National Herbarium Expedition 1999(P99) 13 cult RBGE 19972530A, Agalmyla parasitica (Lam.) Kuntze* 5.9.95 Weber WU Ancylostemon aureus (Franch.) B.L.Burtt Ancylostemon convexus Craib MMO 01–153 E, WU MMO 01–176 E, WU E E Anna mollifolia (W.T.Wang) W.T.Wang & K.Y.Pan Anna submontana Pellegr.* MMO 01–146 E, WU MMO 01–85 E, WU Boea hygrometrica (Bunge) R.Br. Boea hygroscopica F.Muell. Gu 01–6184 Weber 810808–1/1 KUN WU Boea hygroscopica F.Muell. B. Tan, R.G.Coveny & E.A.Brown 443, cult 19970386 Lambinon 87/830 Boea magellanica Lam.* MMO 01–182B ex Zhang Chang Qin 200012 Gu 99–705 Boeica porosa C.B.Clarke Briggsia longipes (Hemsl. ex Oliv.) Craib MMO 01–122 Boeica ferruginea Drake E L E, WU KUN E, WU Briggsia mihieri Craib Wang 11315B PE Briggsia muscicola (Diels.) Craib Briggsia rosthornii (Diels) B.L.Burtt Kew (1995–2229) Sino-Amer. Guizhou Botanical Expedition 398 (US 229325) MMO 01–141 K US Calcareoboea coccinea C.Y.Wu ex H.W.Li* Chirita asperifolia (Blume) B.L.Burtt [L] P.Woods 1071, 30.4.1968 (C6570) cult. HBV GS-96–02 ex HB Chirita caliginosa C.B.Clarke [M] München-Nymphenburg; Kiehn & Pfosser 2000–1 cult. Smithsonian 94–085, Chirita flavimaculata W.T.Wang [G] Skog 7735 (US 590933) cult. RBGE 19941913, Chirita gemella D.Wood [G] Averyanov, L. 1987 Panigrahi 12231, 1969 Chirita hamosa R.Br. [M] (C8032H) cult RBGE 19972897 Chirita lacunosa (Hook f.) B.L.Burtt [C] Chirita lavandulacea Stapf. [M] Chirita longgangensis W.T.Wang [G] Chirita pinnata W.T.Wang [G] Chirita pinnatifida (Hand.-Mazz.) B.L.Burtt [G] Chirita pumila D.Don [C] cult. RBGE 20000897 cult. RBGE 19941915. Takhtajan, A. & Aruzytov, N. 1975 Expedition Beijing 896526 (US 294374) Xie Qingjian J-037 (US 422838) cult. RBGE 19962271, Gaoligong Shan Expedition 1996 7938 E, WU E WU US Origin China, unknown locality — Peninsula Malaysia, AJ490920 Perak, Larut distr. Peninsula Malaysia — Philippines, Palawan, FJ501421 Near summit of Cleopatra Needle Philippines; Palawan, near — Thumb Peak Philippines, Leyte, Leyte — Island, Mt. Lobi, Philippines, Luzon, Barangay Penicuason Peninsula Malaysia; Maxwell’s Hill China, Yunnan, Binchuan county, China, Yunnan, Dali Co., Yu Dai Lu, Cang Shan, China, Guangxi, Napo county China, Yunnan, Maguan county China, unknown locality Australia, N Queensland, Palmerston N.P. Australia, N Queensland, Tchupala Falls US E — trnL-F ITS1/ITS2 FJ501499 AJ492307 — — — FJ501541 FJ501333 - — FJ501361 FJ501540 — — FJ501360 FJ501420 FJ501539, part — FJ501398 FJ501505 FJ501336 — FJ501506 FJ501337 — FJ501543 FJ501422 FJ501542 AF055050 / AF055051 FJ501362 — — FJ501476 FJ501577 FJ501319 — — — FJ501320 — FJ501478 FJ501321 FJ501379 FJ501440 — FJ501380 FJ501423 FJ501441 FJ501545 — FJ501544 — AF055052 / AF055053 FJ501363 — FJ501425 FJ501548 FJ501547 FJ501366 FJ501365 FJ501406 FJ501516 FJ501365 FJ501419, part FJ501391 FJ501538 FJ501359 FJ501488 FJ501325 — FJ501525 — FJ501408 FJ501523 FJ501345 FJ501392 FJ501489 — FJ501384, part FJ501458 FJ501308 FJ501390 AJ490903 FJ501487 AJ492290 FJ501324 FJ501347 China, Guangxi, Rongshui — Xian China, Guangdong Prov., — Lianxian county China, Yunnan, Nujiang FJ501393, Lisu Aut. Pref., Fugong part county FJ501526 FJ501349 FJ501527 FJ501350 FJ501491 FJ501327 Papua New Guinea, Morobe Province China, SE Yunnan China, unknown locality China, Yunnan, Xichou county China, Chongqing, Nanchuan unknown origin China, Guizhou Prov., Jiangkou Xian China, Guangxi, Napo county Indonesia, Java, forest above Tjibodas Garden Peninsula Malaysia China, Guangxi, leg. in US 11.03.1996 E Viet Nam, Hong Quang Special Region, Cat Hai E India, Allahabad, Meizapus no voucher Peninsula Malaysia, Pahang, Lipis distr., Gua Rusa E China E Viet Nam US atpB-rbcL May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae 1007 Appendix 1. Continued. Taxon Voucher number cult. RBGE 19791050. Godfrey, T.C. 369 ex Smithsonian Institute 94– Chirita spadiciformis W.T.Wang [G] 087, cult. RBGE 19951205 Chirita urticifolia Buch.-Ham. ex D.Don* EMAK 109 H 20.9.1991 (Edinburgh-Makalu [C] Expedition 1991) Skog 7736 (US 590934) Chirita walkeri Gardner [C] Chirita sinensis Lindl. [G] Chiritopsis repanda W.T.Wang var. guilinensis W.T.Wang* Colpogyne betsiliensis B.L.Burtt* Conandron ramondioides Sieb. & Zucc.* Deposited in Origin atpB-rbcL trnL-F ITS1/ITS2 E China, Hong Kong FJ501409, part FJ501524 FJ501348 E China AJ490904 AJ492291 FJ501346 E Nepal, Sankhuwasabha distr., Arun valley — FJ501492 FJ501328 FJ501490 FJ501326 AJ492292 FJ501351 FJ501445 FJ501515 FJ501302 FJ501340 FJ501432 — FJ501431 — FJ501430 — US cult. RBGE 19951206 E MM 9894C cult. RBGE 19691267; Takeda Herbal Garden Kyoto MMO 01–138 E E Sino-America Bot. Expedition 1429 cult. RBGE 19943415A, AGS Expedition 1994–1622 US Corallodiscus conchifolia Batalin Cyrtandra cupulata Ridl. Wang, Hong et al 105 Weber 840806–2/4 E WU Cyrtandra glabra Banks ex Gaertn. Cronk & Percy T91 Cyrtandra longifolia (Wawra) Hillebr. ex C.B.Clarke Cyrtandra pendula Blume cult. HBV; Kiehn 920825–2/1 WU China, Yunnan, Xichou— FJ501374 Napo China, Yunnan Prov., FJ501373 Kunming Municipality China, Yunnan, Dêqên — Zang Aut. Reg., Zhongdian county China, unknown locality FJ501375 Pensinsula Malaysia, FJ501414 Perak, Maxwell’s Hill French Polynesia: Society AY423119 Is.: Tahiti: Mt. Tearoa Col USA, Hawaii, Kauai FJ501413 cult. HBV; Weber & Anthonysamy 860730–1/2 cult. HBV; Smith 3905/GES WU Peninsula Malaysia Corallodiscus lanuginosus (Wall. ex R. Br.) B.L.Burtt (G79)* Corallodiscus lanuginosus (Wall. ex R. Br.) B.L.Burtt (G8)* Corallodiscus sp. E, WU cult. Smithsonian 94–250, — origin: Sri Lanka; leg. in US 11.03.1996 China, Guangxi, Zhuang AJ490905 Aut. Reg. Madagascar, Fianarantsoa — Japan FJ501405 E E USA, Hawaii, Maui, East Maui Kapua & al. s.n. SRP (photo USA, Hawaii, Oahu Cyrtandra sessilis H.St.John voucher) WU Peninsula Malaysia, Didissandra frutescens (Jack) C.B.Clarke Weber 840805–1/2 (DI01) Perak, Maxwell´s Hill WU Peninsula Malaysia, Didissandra frutescens (Jack) C.B.Clarke Weber 840805–1/2 (MB) Perak, Maxwell´s Hill cult. RBGE 19650167, Jong E Peninsula Malaysia, Didymocarpus antirrhinoides A.Weber 9009 Perak, Bujong Melakah, Ipoh. GZU Nepal, Langtang Area (N Didymocarpus aromaticus Wall. ex D. Don Poelt s.n. sub GZU Inv.-Nr. 109–86 Kathmandu) cult. RBGE 19830510; P.Davis E Peninsula Malaysia, Didymocarpus citrinus Ridl. 69437 Perlis, Kedat Peak Weber 860816–2/1 WU Peninsula Malaysia, Didymocarpus cordatus Wall. ex DC. Perak, Maxwell’s Hill Noltie, Pradhan, Sherub & E Bhutan, Deothang District Didymocarpus podocarpus C.B.Clarke Wangdi 193, NPSW 193 Wang 991106 PE China, Yunnan, Pingbian Didymocarpus purpureobracteatus W.W.Sm. MMO 01–70 CM China: Yunnan, Pingbian Didymocarpus purpureobracteatus W.W.Sm. MMO 01–156 E, WU China, Yunnan, Binchuan Didymocarpus stenanthos C.B.Clarke county Weber & Anthonysamy WU Peninsula Malaysia, Emarhendia bettiana (M.R.Hend.) 860825–1/1; cult. HBV. Pahang, Kuantan distr. Kiew, A. Weber & B.L.Burtt* cult. RBGE 19754106 E (Greece) Haberlea rhodopensis Friv.* cult. RBGE 19951207 E unknown origin Hemiboea bicornuta (Hayata) Ohwi Gu G3 KUN China, unknown locality Hemiboea cavaleriei H.Lev. Wang 11317 PE China, Chongqing, Hemiboea gracilis Franch. Nanchuan Wang 11306 PE China, Chongqing, Hemiboea subcapitata C.B.Clarke Chengkou Weber 840805–1/12 WU Peninsula Malaysia, Henckelia albomarginata (Hemsl.) Perak, Maxwell’s Hill; A.Weber base Cyrtandra platyphylla A.Gray SRP FJ501433 FJ501532 — AY818826 / AY818861 AY423136* FJ501353 FJ501531 FJ501412 FJ501530 AY818846 / AY818881 FJ501354 FJ501410 FJ501528 — FJ501411 FJ501529 — U91313 FJ501521 — — FJ501522 — — FJ501513 DQ912671 FJ501402 FJ501511 — AJ490906 AJ492293 DQ912669 — AJ492294 DQ912673 FJ501404 FJ501514 DQ912688 FJ501401 FJ501510 — — — DQ912676 FJ501403 FJ501512 DQ912687 AJ490908 AJ492295 — AJ490909 FJ501416 FJ501415 — AJ492296 FJ501534 FJ501533 FJ501536 — FJ501356 FJ501355 — FJ501417 FJ501535 FJ501357 AJ490910 AJ492297 — American Journal of Botany 1008 [Vol. 96 Appendix 1. Continued. Taxon Henckelia corrugata Mendum Henckelia ericii A. Weber [= Loxocarpus holttumii M.R.Hend.] Henckelia floccosa (Thwaites) A.Weber & B.L.Burtt Henckelia humboldtiana (Gardner) A.Weber & B.L.Burtt Hovanella madagascarica (C.B.Clarke) A.Weber & B.L.Burtt* Hovanella sp. nov. Jancaea heldreichii Boiss.* Kaisupeea herbacea (C.B.Clarke) B.L.Burtt* Leptoboea multiflora (C.B.Clarke) Gamble* subsp. grandifolia B.L.Burtt Loxostigma cavaleriei (H.Lev. & Van.) B.L.Burtt Loxostigma fimbrisepalum K.Y.Pan Loxostigma griffithii (Wight) C.B.Clarke* Voucher number Deposited in Origin atpB-rbcL trnL-F ITS1/ITS2 cult RBGE 19981788, RBGEPNH Expedition 1998, DNA no. D12 Weber 840723–1/2 E Philippines, Palawan, Cleopatra Needle — FJ501484 — WU Malaysia, Malaya — FJ501479 — G 157 Jang WU Sri Lanka — FJ501486 — Sri Lanka, Gombiya Ridge Madagascar, Antsiranana Prov. Madagascar, Toamasina FJ501389 FJ501485 — — FJ501451 — — FJ501452 — Greece, Mt Olymp Thailand, Prov. Chachoengsao, Khao Tak Groep Thailand, SE, Khaso Phra Bat, N of Chanthaburi China, Yunnan, Xichou Co., Far Dou China, Yunnan, Jinping Nepal, Yamphudin FJ501378 FJ501385 FJ501439 FJ501459 — FJ501309 FJ501381 FJ501442 — — FJ501509 FJ501355 FJ501399 FJ501400 FJ501507 FJ501508 — FJ501338 E China, Yunnan, Nujiang AY423137 Lisu Aut. Pref., Gongshan AY423137 — PE E China, unknown locality China, Yunnan, Nujiang Lisu Aut. Pref., — FJ501394 FJ501498 FJ501495 FJ501332 AF349152 / AF349233 China, Yunnan, road to Xichou, Cheng Jia Po China, Yunnan, road to Xichou, Cheng Jia Po Japan, unknown locality China, unknown locality China, Hunan Prov., Xining county China, Guizhou Prov. Yinjiang county China; Guizhou, Jiangkou FJ501395 FJ501497 FJ501331 — FJ501496 — FJ501424 — — FJ501546 FJ501483 FJ501482 FJ501364 — — — FJ501481 — — — FJ501323 Thailand, Chiang Mai, Doi Chiang Dao China, Yunnan, Xichou Thailand; Chiang Mai, Grasshopper cave Peninsula Malaysia, Kedah, Pulau Langkawi, Bukit Terbak Peninsula Malaysia, Pahang, Lipis distr., Gua Bama Peninsula Malaysia, Perak, Kinta distr. China, Yunnan, Maguan county China, Guangdong Prov, Lianxian county Peninsula Malaysia, Kedah, Pulau Langkawi, Selat Panchar Peninsula Malaysia, Kedah, Pulau Langkawi, Pulau Dayang Bunting Peninsula Malaysia, Kedah, Pulau Langkawi FJ501387 FJ501461 FJ501312 — FJ501386 FJ501462 FJ501460 FJ501313 FJ501311 — FJ501464 FJ501314 — FJ501465 — AJ490911 AJ492298 FJ501315 — FJ501472 FJ501318 — FJ501463 — — FJ501471, part — — FJ501467 — — FJ501466 — Kostermans 28519 L MM 9880A E T.Sieder & M.Pfosser 101; 9.2.2000 cult. RBGE 19771605 cult. RBGE 19972918; K.Larsen 44272, 6 Nov 1993. Larsen et al. 32065, 26.8.1972 MMO 01–131 WU photo E E E E, WU Lysionotus pauciflorus Maxim. Wang 991005 cult. RBGE 19892473A; Kew/ Edinburgh Kanchenjunga Expedition (1989) 940. cult. RBGE 19962309, Gaoligong Shan Expedition 1996 GSE96–7668 Wang S-10669 cult. RBGE 19962269A, Gaoligong Shan Expedition 1996 7925 MMO 01–101 E, WU Lysionotus petelotii Pellegr. MMO 01–100 E, WU Opithandra primuloides (Miq.) B.L.Burtt* Oreocharis aurea Dunn. Oreocharis auricula (S. Moore) C.B.Clarke G03 Oreocharis auricula (S. Moore) C.B.Clarke G04 Oreocharis auricula (S. Moore) C.B.Clarke G04 Ornithoboea arachnoidea (Diels) Craib Ornithoboea wildeana Craib Ornithoboea sp. nov. cult. RBGE 19842178A Wang S-10725 Luo Lin-bo 0125 E PE WU Sino-America Expedition 1832 WU Loxostigma sp. Lysionotus chingii Chun ex W.T.Wang Lysionotus forrestii W.W.Sm. PE E MMO 03–304 E cult RBGE 19972903 E Wang 00401 MMO-04–439 PE E Paraboea acutifolia (Ridl.) B.L.Burtt Weber 86805–2/1 WU Paraboea brachycarpa (Ridl.) B.L.Burtt Weber 870508–2/6 WU Paraboea capitata Ridl. var. capitata Weber 870522–5/2; cult. HBV. WU Paraboea crassifolia (Hemsl.) B.L.Burtt MMO 01–83 Paraboea dictyoneura (Hance) B.L.Burtt Paraboea ferruginea (Ridl.) Ridl. Xie Qingjian J-040 (US 422817) Weber 860806–1/2 WU Paraboea lanata (Ridl.) B.L.Burtt Weber 860807–1/2 WU Paraboea laxa Ridl. C 4197 E, WU US E May 2009] Möller et al.—Phylogeny of didymocarpoid Gesneriaceae 1009 Appendix 1. Continued. Taxon Paraboea rufescens (Franch.) B.L.Burtt (G23b) Paraboea rufescens (Franch.) B.L.Burtt (G72) Paraboea rufescens (Franch.) B.L.Burtt var. umbellata (Drake) K.Y.Pan Paraboea sinensis (Oliv.) B.L.Burtt (G20) Paraboea sinensis (Oliv.) B.L.Burtt (G21b) Paraboea swinhoei (Hance) B.L.Burtt Petrocodon dealbatus Hance* Petrocosmea kerrii Craib Petrocosmea minor Hemsl. Petrocosmea nervosa Craib Petrocosmea sericea C.Y.Wu ex H.W.Li Platystemma violoides Wall. Voucher number Deposited in Sino-Amer. Bot. Exped 1566 (US 64646) MMO 01–99 E, WU MMO 01–147 E, WU Wen He Qun W049 (US 329798) Xie Qingjian J-003 (US 422825) Wagner 6640 (US 427725) Xie Qingjian J-042 (US 422841) cult. RBGE 19715592 Sino-Amer. Bot. Exped. 1574 (US 56119) cult. RBGE 19933232, SI.78–057 Gu 99–1104 Projektteam 197–241 US US US US US E US E, US KUN WU cult. RBGE 19951540, Xie, Q.J. & Ye, C.X. cult. RBGE 19711477 cult. RBGE 19784020 E Raphiocarpus begoniifolius (H.Lev.) B.L.Burtt Raphiocarpus sp. (G61) Wang 991108 PE MMO 01–55 E, WU Raphiocarpus sp. (G64) MMO 01–69 E, WU Raphiocarpus sp. Beijing Youth team 572 Primulina tabacum Hance* Ramonda myconi (L.) Rchb.* Ramonda nathaliae Panc. & Petr. Raphiocarpus petelotii (Pellegr.)B.L.Burtt cult RBGE 19982405; S.Goodwin & R.Cherry 92/208 ex Kew 1988 4866 Rhabdothamnopsis sinensis Hemsl.* RBGE 1997 2562; RBGERhynchotechum discolor (Maxim.) PNH Expedition 1997/SM8 B.L.Burtt M.Mendum, G.Argent, Rhynchotechum parviflorum Blume Hendrian 00148; coll 25.2.2000 Weber 870420–2/4 Ridleyandra porphyrantha (Kiew & A. Weber) A. Weber Saintpaulia tongwensis B.L.Burtt Saintpaulia velutina B.L.Burtt Schizoboea kamerunensis K.Fritsch (B.L.Burtt)* Spelaeanthus chinii Kiew, A.Weber & B.L.Burtt* E E PE E K E E WU cult. RBGE 19850668, I.C.Mather 2 cult. RBGE 19872179 E J.Lewalle 6693, 9.4.1972 E Weber 860709–2/2 E WU Origin atpB-rbcL trnL-F ITS1/ITS2 — FJ501468 — FJ501388 FJ501469 FJ501316 — FJ501470 FJ501317 — FJ501473 — — FJ501474 — — FJ501418 FJ501475 FJ501537 — FJ501358 FJ501397, part — FJ501502 FJ501334 FJ501504 — AJ490912 AJ492299 FJ501335 — FJ501382 FJ501503 FJ501443 — — AJ490913 AJ492300 FJ501352 AJ490914 — AJ492301 FJ501438 — — — FJ501517 FJ501342 — FJ501493 FJ501329 — FJ501494 FJ501330 FJ501407 FJ501519 FJ501344 — FJ501518 FJ501343 China, unknown locality Philippines; Luzon, Isabela Central Sulawesi, Mt Sojol. AJ490915 FJ501376 AJ492302 FJ501436 FJ501310 — FJ501377 FJ501437 — Malaysia, Pahang, side ridge of Gunung Bunga Buah Tanzania, Tanga Region — FJ501520 — — FJ501446 FJ501303 AJ490916 AJ492303 FJ501304 — FJ501453 FJ501305 — FJ501457 FJ501307 — FJ501449 AF316903 — FJ501448 AF316905 — FJ501456 AF316951 — FJ501450 AF316907 AJ492304 AF316917 FJ501455 AF316926 China, Yunnan Prov., Lunan Xian China, Yunnan, Xichou, Cheng Jia Po China, Guangxi, Napo, Nong Bu China, Guangxi Prov., Napo county China, Guangxi Prov. Jingxi county Taiwan, Taoyuan Hsien China, Guangdong Prov., Lianxian county unknown origin China, Yunnan Prov., Lunan Xian China, N Yunnan China, unknown locality Nepal, SE Kathmandu Pulchoki China, Guangdong, Lian River Spain, Pyrenees Macedonia, unknown locality China, Yunnan, Yuanyang China, Yunnan, Pingbian Co., Dar Wei Shang China, Yunnan, Pingbian Co., Dar Wei Shang China, Kwangsi (Guangxi) Viet Nam, Lao Cai Prov. Tanzania; unknown locality Burundi, Muramvya, Mount Teza Peninsula Malaysia, Pahang, Jerantut distr., Taman Negara Madagascar, Tuléar Prov., Ranomafana Madagascar, Tuléar Prov., Ranomafana Swaziland, Mbabane MM 9717 E MM 9715 E E Streptocarpus hilsenbergii R.Br. cult. RBGE 19941745, Isobel La Croix cult. RBGE 19631505 E Madagascar, Mandrake Valley Streptocarpus holstii Engl. cult. RBGE 19592272 E Streptocarpus ibityensis Humbert cult. RBGE 19932867, E.Fischer 250/93 E Tanzania, unknown AJ490917 locality Madagascar, Antananarivo — Prov. Streptocarpus andohahelensis Humbert Streptocarpus beampingaratrensis Humbert var. beampingaratrensis Streptocarpus dunnii Hook.f. American Journal of Botany 1010 Appendix 1. Continued. Taxon Voucher number Deposited in atpB-rbcL trnL-F ITS1/ITS2 — FJ501480 FJ501322 — FJ501444 AF316929 AJ490918 AJ492305 AF316979 FJ501383 — FJ501447 — — AF316914 — FJ501434 — — FJ501435 — AJ490919 AJ492306 — no voucher India FJ501372 E Taiwan, unknown locality AY423111 FJ501429 AY423129 — — Streptocarpus orientalis Craib A.Weber & M.Kiehn 29.9.98 Streptocarpus papangae Humbert MM 9718 E Streptocarpus rexii Lindl.* cult. RBGE 19870333, K.Jong E Streptocarpus saxorum Engl. Streptocarpus saxorum Engl. Tetraphyllum roseum Stapf (G113) Chautems & Perret 01–023 cult. RBGE 19721499; I.C.Mather 4 Kurzweil HK 798 Tetraphyllum roseum Stapf (G124) Larsen & al. 31190 E Trisepalum speciosum (Ridl.) B.L.Burtt Weber 860805–1/1 WU Unassigned: Jerdonia indica Wight* Titanotrichum oldhamii (Hemsl.) Soler.* G 155 Jang cult. RBGE 19973433 E, WU MP WU Origin Thailand, Prov. Chiang Mai Madagascar, Tuléar Prov., Ranomafana South Africa, NE Cape Prov., Grahamstown cult. CJBG Tanzania, Tanga region Thailand, Krabi Prov., c. 20km N of Krabi Thailand, Lam Tok Lam Pae Peninsula Malaysia, Kedah, Pulau Langkawi, Bukit Puteh