Systematic Botany (2014), 39(1): pp. 239–252 © Copyright 2014 by the American Society of Plant Taxonomists DOI 10.1600/036364414X678008 Date of publication 02/05/2014 A Phylogeny of the Violaceae (Malpighiales) Inferred from Plastid DNA Sequences: Implications for Generic Diversity and Intrafamilial Classification Gregory A. Wahlert,1 Thomas Marcussen,2 Juliana de Paula-Souza,3 Min Feng,4 and Harvey E. Ballard, Jr.5,6 1 Department of Biology, University of Utah, Salt Lake City, Utah 84112, U. S. A. 2 Monrads gate 21A, NO-0564 Oslo, Norway. 3 Instituto de Biociências, Universidade de São Paulo, São Paulo, São Paulo, 05508, Brazil. 4 Key Laboratory of Systematic and Evolutionary Botany, Institute of Botany, Chinese Academy of Sciences, 20 Nanxincun, Xiangshan, Beijing 100093, China. 5 Department of Environmental & Plant Biology, Ohio University, Athens, Ohio 45701, U. S. A. 6 Author for correspondence (ballardh@ohio.edu) Communicating Editor: Thomas L.P. Couvreur Abstract—The Violaceae consist of 1,000–1,100 species of herbs, shrubs, lianas, and trees that are placed in 22 recognized genera. In this study we tested the monophyly of genera with a particular focus on the morphologically heterogeneous Rinorea and Hybanthus, the second and third most species-rich genera in the family, respectively. We also investigated intrafamilial relationships in the Violaceae with taxon sampling which included all described genera and several unnamed generic segregates. Phylogenetic inference was based on maximum parsimony, maximum likelihood, and Bayesian analyses of DNA sequences from the trnL/trnL–F and rbcL plastid regions for 102 ingroup accessions. Results from phylogenetic analyses showed Rinorea and Hybanthus to be polyphyletic, with each genus represented by three and nine clades, respectively. Results also showed that most intrafamilial taxa from previous classifications of the Violaceae were not supported. The phylogenetic inferences presented in this study illustrate the need to describe new generic segregates and to reinstate other genera, as well as to revise the traditionally accepted intrafamilial classification, which is artificial and principally based on the continuous and homoplasious character state of floral symmetry. Keywords—Non-monophyletic genus, rbcL, Rinorea, trnL/trnL–F, Viola, zygomorphy. families is now better understood (Wurdack and Davis 2009, Xi et al. 2012, and citations therein). Results from an 83-gene phylogeny (Xi et al. 2012) placed the Violaceae in a highly supported clade of 10 families united by the synapomorphy of parietal placentation, and resolved the Violaceae + Goupiaceae as sister to the large ‘salicoid’ clade containing Salicaceae, Passifloraceae, Turneraceae, and four other families. Morphology—Features defining the Violaceae include five sepals and petals, five stamens, a three-carpellate compound pistil (except in a few dioecious Melicytus, most Anchietea, and one Leonia), a single style, parietal placentation, polygonum type embryo sac formation, fleshy endosperm, tetrasporangiate anthers opening longitudinally, and a secretionary tapetum (Ballard et al. 2013). The family is perhaps best known for the large and diverse genus Viola—predominately small herbs with bilaterally symmetrical, spurred flowers that are distributed worldwide in temperate regions and montane habitats in the tropics. Nevertheless, the herbaceous habit and floral morphology of many Viola species are exceptional in this otherwise primarily tropical woody family: most genera are composed of species that are trees, shrubs, or lianas and more than half bear weakly zygomorphic flowers that lack spurs and nectaries. While most genera produce three-valved, thick-walled capsules well-known in the Violaceae, two genera (Gloeospermum and Melicytus) bear juicy berries, Leonia produces an indehiscent nut, and Anchietea produces a paper-thin, bladder-like capsule. Genera diverge greatly in structure and elaboration of the androecium, including varying degrees of fusion of the filaments, orientation and dehiscence of the paired thecae, presence and morphology of the dorsal connective scale terminating the anthers, and morphology of the glands often associated with the filaments. Cleistogamy has been reported for most lineages within Viola and in a few species of certain hybanthoid groups The Violaceae are a medium-sized cosmopolitan family containing 22 recognized genera and 1,000–1,100 species of herbs, shrubs, lianas, and trees (Melchior 1925a, 1925b; Hekking 1988; see Table 1 for a synopsis of generic diversity). The family is predominantly composed of woody genera distributed in the tropics and encompasses a wide range of growth forms, inflorescence architectures, floral morphologies, and fruit types (Fig. 1). Most genera are monotypic or oligotypic and have limited distributions (e.g. Decorsella with one species in west Africa; Isodendrion with four species in Hawaii; Mayanaea with one species in Guatemala; and Hybanthopsis and Noisettia, each with one species in Brazil; Table 1). In contrast, the three largest genera, Viola, Rinorea, and Hybanthus account for ca. 98% of the species in the family and are widely distributed throughout the world. Several molecular phylogenetic studies have been conducted in Viola, the largest genus in the family with 580–620 species (Ballard 1996; Ballard et al. 1999; Yockteng et al. 2003; Marcussen et al. 2012). Rinorea, the second largest genus, is composed of 225–275 species of small trees and shrubs and has a pantropical distribution. Previous molecular phylogenetic studies have suggested that Rinorea is not monophyletic (Feng 2005; Wahlert and Ballard 2012). Likewise, similar studies have shown Hybanthus to be polyphyletic, consisting of several morphologically well-defined and phylogenetically disparate clades distributed around the world in mostly tropical and subarid regions (Feng 2005; Tokuoka 2008; Paula-Souza 2009). Circumscription and Systematic Position of the Violaceae— The familial composition of the order Malpighiales based on molecular phylogenetic evidence was first circumscribed by Chase et al. (1993). The APG III concept of the Malpighiales (Angiosperm Phylogeny Group 2009; Chase and Reveal 2009) encompasses 42 families, including the Violaceae. The monophyly of the order has been corroborated by various multigene phylogenetic studies, and the resolution among 239 2014] WAHLERT ET AL.: PHYLOGENY OF THE VIOLACEAE 241 Fig. 1. Some representative genera in the Violaceae. A. Amphirrhox longifolia. B. Anchietea pyrifolia. C. Hybanthopsis bahiensis. D. Hybanthus arenarius. E. Hybanthus enneaspermus. F. Leonia glycycarpa. G. Paypayrola confertiflora. H. Paypayrola hulkiana. I. Schweiggeria fruticosa. J. Rinorea pubiflora. K. Rinorea arborea. L. Viola cuicochensis. 242 SYSTEMATIC BOTANY few character states that have been shown to be continuous and homoplasious (Feng 2005). In the most recent classification of the Violaceae, Hekking (1988) isolated the aberrant genera Fusispermum and Leonia in their own subfamilies (Fusispermoideae and Leonioideae, respectively), and divided the remainder (forming the subfamily Violoideae) into the tribes Rinoreeae and Violeae based primarily on floral symmetry and presence/absence of nectaries and spurs. In the Rinoreeae, Hekking (1988) maintained four subtribes, but left genera in the Violeae unclassified as to subtribes. Melchior (1925a) had previously divided Violeae into two subtribes, Violinae and Hybanthinae, depending on whether the corolla was fully spurred (former) or merely saccate (latter). Table 2 shows a comparison of the most recent intrafamilial classifications of Melchior (1925a, 1925b) and Hekking (1988). Phylogenetic Relationships within the Violaceae—Previous investigations in the family utilizing morphology, anatomy, chromosome numbers, and DNA sequences provided evidence that two of the largest genera—Rinorea and Hybanthus—are non-monophyletic (Feng 2005; Tokuoka 2008). Table 1 summarizes the generic diversity in the Violaceae as well as several generic segregates subsumed under broad taxonomic concepts of Rinorea and Hybanthus. It has also been shown that the genera with strongly zygomorphic corollas [Volume 39 and the higher-level groups containing them were likely derived from weakly zygomorphic lineages, with multiple origins of strongly zygomorphic corollas and spurs (Feng 2005). The most recent phylogeny of the Violaceae was inferred from five genes (Tokuoka 2008) and aimed at examining higher-level relationships in the family. The results corroborated those of Feng (2005), but taxon sampling left out some genera, as well as many morphologically divergent species from the unnamed generic segregates in Rinorea and Hybanthus. The phylogeny of Tokuoka (2008) showed that most intrafamilial taxa were not congruent with phylogenetic relationships, but it did not have sufficient taxon sampling needed to reevaluate generic boundaries and diversity in the family. The main goal of this study was to produce a well resolved phylogeny of the Violaceae inferred from plastid DNA sequences, with complete taxon sampling of all genera as well as 10 additional unnamed generic segregates. With a new phylogenetic framework, and drawing on previous morphological evidence, we discuss the resulting taxonomic diversity at the rank of genus, which is substantially greater than previously understood for the family. We also discuss the need to completely revise the intrafamilial classification for the Violaceae in light of new phylogenetic hypotheses. Table 2. A comparison of genera and unnamed generic segregates and their placement in Melchior’s (1925a, 1925b) and Hekking’s (1988) classifications of the Violaceae, as well as the molecular clade in which each genus or segregate was resolved in the present study. A dash (—) indicates a genus, generic segregate, or intrafamilial group which was not recognized or treated in the classifications of Melchior or Hekking. The intrafamilial taxa included in this table are subfamilies Fusispermoideae, Violoideae, and Leonioideae; tribes Rinoreeae and Violeae; and subtribes Hybanthinae, Hymenantherinae, Paypayrolinae, Rinoreinae, and Violinae. Genus or generic segregate Fusispermum Rinorea crenata group Melchior (1925a, 1925b) — Violoideae Rinoreeae, Rinoreinae Hekking (1988) Fusispermoideae Violoideae Rinoreeae, Rinoreinae Molecular clade, this study Sister to all Violaceae Sister to Rinorea s. s. + Violaceae or unresolved trichotomy with Rinorea s. s. and Violaceae Rinorea s. s. Rinoreeae, Rinoreinae Rinoreeae, Rinoreinae Sister to remaining Violaceae or unresolved trichotomy with R. crenata group and Violaceae Decorsella — Rinoreeae, Rinoreinae Unplaced Allexis Rinoreeae, Rinoreinae Rinoreeae, Rinoreinae Clade 1 Schweiggeria Violeae, Violinae Violeae Clade 1 Noisettia Violeae, Violinae Violeae Clade 1 Viola Violeae, Violinae Violeae Clade 1 Hekkingia — — Clade 2 Paypayrola Rinoreeae, Paypayrolinae Rinoreeae, Paypayrolinae Clade 2 Leonia Leonioideae Leonioideae Clade 3 Gloeospermum Rinoreeae, Rinoreinae Rinoreeae, Rinoreinae Clade 3 Amphirrhox Rinoreeae, Paypayrolinae Rinoreeae, Paypayrolinae Clade 3 Hybanthus caledonicus group Violeae, Hybanthinae Violeae Clade 3 Hybanthus havanensis group Violeae, Hybanthinae Violeae Clade 3 Hybanthus concolor Violeae, Hybanthinae Violeae Clade 3 Mayanaea — Violeae Clade 3 Orthion — Violeae Clade 3 Rinoreocarpus Rinoreeae, Rinoreinae Rinoreeae, Rinoreinae Unplaced Hybanthus enneaspermus group Violeae, Hybanthinae Violeae Clade 4 Melicytus Rinoreeae, Hymenantherinae Rinoreeae, Hymenantherinae Clade 4 Hybanthus guanacastensis group — Violeae Clade 4 Anchietea Violeae, Violinae Violeae Clade 4 Hybanthopsis — — Clade 4 Corynostylis Violeae, Violinae Violeae Clade 4 Agatea Violeae, Hybanthinae Violeae Clade 4 Hybanthus fruticulosus group Violeae, Hybanthinae Violeae Clade 4 Hybanthus thiemei group Violeae, Hybanthinae Violeae Clade 4 Rinorea virgata group Rinoreeae, Rinoreinae Rinoreeae, Rinoreinae Clade 4 Hybanthus mexicanus group Rinoreeae, Rinoreinae Rinoreeae, Rinoreinae Clade 4 Isodendrion Rinoreeae, Isodendriinae Rinoreeae, Isodendriinae Clade 4 Hybanthus calceolaria group Violeae, Hybanthinae Violeae Clade 4 2014] WAHLERT ET AL.: PHYLOGENY OF THE VIOLACEAE Materials and Methods + Taxon Sampling—We sampled 99 different species from the Violaceae (102 ingroup accessions; three outgroup taxa) that represent all 22 currently recognized genera, as well as 10 morphologically distinctive and biogeographically disjunct generic segregates which previous research has suggested may be lineages deserving of taxonomic recognition. The genus Hybanthus was heavily sampled to include 38 accessions from nine morphologically and geographically disparate groups. Likewise, the 16 accessions of Rinorea encompassed exemplars from three groupings and included taxa from across its pantropical distribution. The use of Malesherbia, Passiflora, and Turnera (Passifloraceae s.l.) as outgroup taxa was based on Wurdack and Davis (2009). Voucher specimens, geographic origin, and GenBank accession numbers are given in Appendix 1. DNA Extraction, Amplification, and Sequencing—Genomic DNA was isolated from herbarium or fresh, silica-dried leaf tissue using one of three methods: 1) DNEasy Plant Mini Kit (Qiagen, Valencia, CA, USA), 2) SDS “mini-extraction” protocol (Edwards et al. 1991), or 3) a modified 6% CTAB method (Doyle and Doyle 1987; Smith et al. 1991). The trnL (UAA) intron and trnL (UAA)–trnF (GAA) intergenic spacer region was amplified either in one fragment using the primer pair Tab C + Tab F or in two fragments using Tab C + Tab D and Tab E + Tab F (Taberlet et al. 1991). The rbcL region was amplified in two parts, rbcL I and rbcL II. The rbcL I region used rbcL F (Rieseberg et al. 1992) and rbcL 724R (Fay et al. 1997). The rbcL II region was amplified rbcL 536F (500 GGTTATCCGCTAAGAATTATGGG-300 ) and rbcL 3R (500 -TGTGTTG CGGCCGCCTTTTAGTAAAAGGG-300 ). In some cases where freshly silica-dried material was available, the entire rbcL region was amplified in one reaction using rbcL 1A (500 -ATGTCACCACAAACAGARAC TAAA-300 ) and rbcL GR95 (500 -CTTCACAAGCAGCAGCTAGTTC-300 ). Double-stranded DNA was amplified in 25 ml reactions containing 10– 20 ng DNA, 1.0 unit AmpliTaq polymerase (Applied Biosystems, Foster buffer, 1.5 ml 20 mmol MgCl2, 1.0 ml dNTPs, City, CA, USA), 2.5 ml 10 and 0.63 ml 20 mmol amplification primers. PCR amplifications for both chloroplast regions were carried out under the following conditions: an initial denaturation step (94 C, 2 min) followed by 40 cycles of denaturation (94 C, 30 sec), annealing (52 C, 1 min), and elongation (72 C, 1 min), with a final extension step (72 C, 7 min). The PCR amplicons were visualized by electrophoresis on a 1.3% agarose gel stained with ethidium bromide and cleaned with a Promega WizardÒ PCR clean-up system (Promega Corp., Madison, Wisconsin). Cycle-sequencing reactions used the same primers and annealing temperature as for PCR and employed BigDyeÒ fluorescent dye-labeled chemistry (BigDyeÒ terminator cycle sequencing kit, Applied Biosystems, Foster City, California). All cycle-sequenced reactions were cleaned with BigDyeÒ XTerminator purification kit (Applied Biosystems, Foster City, California) and visualized on an Applied Biosystems 3130xl genetic analyzer at the DNA sequencing facility in the Department of Environmental and Plant Biology, Ohio University, Athens, OH. Phylogenetic Analysis—All sequence chromatograms were inspected manually and sequences were aligned with MUSCLE (Edgar 2004), as implemented in Geneious version 5.3.5 (Biomatters; available from http://www.geneious.com/), using standard settings. The trnL/trnL–F alignment was further optimized manually in order to ensure that all putatively homologous indels were consistently aligned using the criteria of identical length and position and > 80% sequence similarity for putative insertions. Indel characters were coded using the Simple IndelCoding algorithm (Simmons and Ochoterena 2000) as implemented in the SeqState software (Müller 2005). The concatenated alignment was 3320 bp long, including 360 coded indels, of which 844 characters were parsimony-informative. Combinability of the chloroplast data sets (trnL/ trnL–F and rbcL) was assessed using the incongruence length difference (ILD) test (Farris et al. 1995), implemented in PAUP* v4.0b10 (Swofford 2002) as the Partition Homogeneity Test. The test was conducted using 1,000 partition replicates, each with 10 random addition replicates, and TBR branch swapping. Data sets were analyzed using maximum parsimony, maximum likelihood, and Bayesian inference approaches. Maximum parsimony (MP) analyses were conducted in PAUP* v4.0b10 (Swofford 2002) using a heuristic search strategy with TBR branch swapping, 1,000 random addition replicates, saving one tree per replicate, steepest descent off, and MULTREES in effect. All characters were equally weighted and unordered. Internal branch support of phylogenetic trees from each MP analysis was estimated with 1,000 bootstrap (BS) replicates (Felsenstein 1985) using a full heuristic search with TBR branch swapping, 10 random stepwise addition replicates, and MULTREES in effect. Maximum Likelihood (ML) analyses were performed with TREEFINDER version of March 2008 (Jobb et al. 2004). Nucleotide sub- 243 stitution models for the trnL/trnL–F and rbcL partitions were estimated by TREEFINDER based on the AICc model selection criterion. The three partitions trnL/trnL–F, rbcL, and coded indels were analyzed under the GTR + G, J3 + G, and JC substitution models, respectively. Confidence in the configuration of branches was estimated by local rearrangement of expected likelihood weights (LR-ELW) edge support (Strimmer and Rambaut 2002). Bayesian inference (BI) was conducted on the partitioned data set using MrBayes 3.2.1 (Ronquist and Huelsenbeck 2003) using the same models of nucleotide substitution as for ML and the binary model for coded indels. Ten million generations were run to estimate probabilities using MCMC (average standard deviation of split frequencies = 0.0029). Parameters and trees were sampled every 500 generations, examined for convergence, and trees from the first 50,000 generations were discarded as “burn in”. Tree files from the two chains were combined and the maximum clade credibility tree was calculated. The ML and BI analyses were run on the Bioportal facility, University of Oslo (Kumar et al. 2009). Maximum parsimony BS values were plotted on a strict consensus cladogram, and ML LR-ELW support values and BI posterior probabilities were plotted on the best ML tree. The data matrix consisting of concatenated trnL/trnL –F and rbcL sequences and coded gaps were deposited in TreeBASE (study number S14488). Results Sequence Characteristics—Of the 102 ingroup accessions, 80 sequences for the trnL/trnL–F region and 70 sequences for the rbcL region were newly generated in this study (Appendix 1). Additional sequences for both ingroup and outgroup were downloaded from GenBank and used in the analyses (23 for the trnL/trnL–F region and 20 for rbcL; Appendix 1). The concatenated matrix had 9.52% missing data for the ingroup. Descriptive statistics for MP analysis of each dataset and the combined matrix are given in Table 3. The incongruence test showed no substantial discordance between the gene regions, and the sequence data were combined for subsequent analysis (as different partitions for ML and BI). The strict consensus of 18,660 most parsimonious trees with bootstrap values is shown in Fig. 2. The best ML tree had a log likelihood score of –23657.3 and is shown in Fig. 3 with both ML LR-ELW support values and BI posterior probabilities. Where all three support values (MP, ML, and BI) are reported below, they are listed sequentially as: MP bootstrap %/ML LR-ELW %/BI posterior probability. Phylogeny of the Violaceae—The monophyly of the Violaceae as currently circumscribed was unambiguously supported (100/100/1.0; Figs. 2, 3). Fusispermum and the Rinorea crenata group were strongly supported as successive sisters to the remaining Violaceae, (100/100/1.0) and (100/ 96/1.0), respectively. Rinorea s. s. was resolved as a strongly supported clade (100/100/1.0), but its sister relationship to the remaining Violaceae was weakly supported (68/70/0.92). The clade containing the remaining Violaceae was highly supported (100/100/1.0). The four major clades recovered Table 3. Descriptive statistics for each chloroplast data partition and the combined matrix used in phylogenetic analyses. MPTs, most parsimonious trees; CI, consistency index; RI, retention index. Aligned length Parsimony informative sites Constant sites Number of MPTs Length of MPTs CI; RI trnL/trnL–F rbcL trnL/trnL–F + rbcL + gaps 1,659 402 (24.2%) 1,002 (60.4%) > 35,000 1,511 0.638; 0.810 1,301 262 (20.1%) 894 (68.7%) > 100,000 1,116 0.467; 0.711 3,320 848 (25.5%) 1,648 (49.%) 18,660 3,242 0.561; 0.756 2014] WAHLERT ET AL.: PHYLOGENY OF THE VIOLACEAE 245 Fig. 3. Maximum likelihood (ML) phylogram for 102 species of Violaceae inferred from analysis of a combined dataset of rbcL and trnL/trnL–F DNA sequences and coded indels. Numbers above the branches are ML LR-ELW support percentages/BI posterior probabilities; nodes marked with a solid black circle represent combined ML LR-ELW support percentages and BI probabilities ³ 95% and 0.99, respectively. Numbered clades (i.e. Clades 1–4) highlight important groupings of genera, but do not correspond to intrafamilial taxa. Clade 1: Viola, Schweiggeria, Noisettia, and Allexis— Viola, Schweiggeria, Noisettia, and Allexis were resolved in Clade 1 with very strong branch support (100/99/1.0) (Figs. 2, 3). Viola and Allexis were unambiguously resolved as monophyletic (both 100/100/1.0), and the two monotypic genera Schweiggeria and Noisettia were strongly supported as a clade (100/100/1.0) which had a highly supported sister relationship to Viola (100/99/1.0). 248 SYSTEMATIC BOTANY Clade 4—THE RINOREA VIRGATA GROUP—The two species in the Rinorea virgata group from Asia are often segregated in the genus Scyphellandra based on leaf morphology, the weakly zygomorphic corolla, few-flowered fasciculate inflorescences borne on short shoots, and small capsules with a single seed, but it was submerged under Rinorea following Kuntze (1891). The distant phylogenetic relationship of the group to Rinorea s. s., as well as substantial and numerous morphological differences (e.g. short-shoot morphology, dioecy in R. virgata), support the reinstatement of Scyphellandra as an earlier generic segregate name. A report of intrastaminal processes has been confirmed in some specimens and may be a synapomorphy of the R. virgata group. THE HYBANTHUS ENNEASPERMUS GROUP—This unnamed Hybanthus segregate lineage contains perhaps as many as 25 species, including several unnamed ones in Africa, Madagascar, IndoChina, Southeast Asia, and northern Australia. They are distinct from the other hybanthoid groups by the combination of a highly reduced axillary inflorescence bearing a single flower, zygomorphic perianth with distinct basal spur, anthers atop a low filament tube, usually elongate staminal nectary glands, and elongate honey-yellow seeds with longitudinally arranged alveolae and/or ribs. THE HYBANTHUS FRUTICULOSUS GROUP—This clade represents another small, unnamed hybanthoid segregate lineage. The Hybanthus fruticulosus group includes two described species from Mexico and Central America, and an additional 1– 2 undescribed species from Baja California (Ballard, unpubl. data). The H. fruticulosus group, and the H. thiemei group to which it is sister, are readily distinguished from other Hybanthus groups by stamens with a very short filamental “collar”, short globose staminal glands, and 6-seeded capsules with pale yellow seeds. The H. fruticulosus group diverges from the H. thiemei group in its terminal densely flowered racemes with short pedicels, apically positioned staminal glands, weakly zygomorphic corollas, and conspicuously rugulose seeds. Ongoing morphological studies are under way to better understand generic limits and species diversity in the H. fruticulosus group, as well as the morphological features that separate it from the H. thiemei group and the other Hybanthus groups. THE HYBANTHUS THIEMEI GROUP—This group includes four to six species, all distinguished by axillary solitary flowers on very long filiform pedicels, strongly zygomorphic corollas with a strongly clawed bottom petal, medially positioned globose staminal glands, and pale yellow minutely alveolate seeds. It shares some features with its sister group, the H. fruticulosus clade but deserves generic segregation from that group based on the differentiation of other character states. Member species include the widespread Neotropical H. thiemei, the Central American H. galeottii, and the South American H. nanus. The group may also include two poorly known taxa endemic to Hispaniola (H. domingensis Urb. & Ekman and H. leucanthus Urb. & Ekman) which, aside from their diminutive habit, closely resemble H. thiemei. THE HYBANTHUS MEXICANUS GROUP—This group contains two species, Hybanthus mexicanus and Rinorea uxpanapana and is endemic to Mexico and Central America. Rinorea uxpanapana was highlighted by Hekking (1988) as an anomalous species in comparison to other members of the Neotropical Rinorea s. s. in its vegetative and floral morphology, but the absence of fruiting material made a reliable generic assignment difficult [Volume 39 (he did speculate a close relationship to Hybanthus mexicanus, however). The two species share a distinctive combination of character states, including leaves arising from short-shoots, fascicled cymes bearing 1–3 flowers, weakly zygomorphic corollas, a short staminal tube with 2–3 free dorsal nectary glands, sessile anthers, and 3-seeded capsules. The combined molecular phylogenetic and morphological evidence strongly supports the segregation of the group as a new genus. ISODENDRION—The endemic Hawaiian genus Isodendrion, with its long tube-shaped corolla composed of closely adherent petals and rudimentary to vestigial connective scale, is placed in a clade with the Hybanthus calceolaria and H. mexicanus groups. Isodendrion shares some traits of the H. mexicanus group, particularly the highly reduced, 1-flowered inflorescence arising from short axillary shoots. The systematic position also indicates that the prolonged coherent nature of the corolla in Isodendrion may have a relatively simple ontogenetic basis from the shorter, open-petaled corolla of related hybanthoid groups. THE HYBANTHUS CALCEOLARIA GROUP—The Hybanthus calceolaria group represents the largest segregate lineage of Hybanthus, and the third largest in the family, with ca. 65 species that are distributed mainly in South America, but also in Central America and one species (H. verticillatus) extending into the southern United States. The group shares the following combination of character states: a strongly zygomorphic corolla, the inflorescence a depauperate to well-developed raceme, the filaments free or fused only at the base, seeds black, prismatic or alveolate, and 6–12 seeds per capsule. Although the taxon sampling in the present study is insufficient to allow for a revised infrageneric classification of species, a recent study (Paula-Souza 2009) indicated that characters such as phyllotaxy, structure of root system, and morphology of nectar gland may be useful in delimiting groupings that correspond with at least four distinct sublineages that are potentially referable to infrageneric taxa. Based on morphological synapomorphies and the highly supported monophyly of the group, there is ample evidence to recognize the H. calceolaria group as a distinct genus. MELICYTUS—Until recently, this genus was split into Melicytus and Hymenanthera on the basis of flower sexuality and carpel number, and recognition of the two genera has been somewhat controversial. Phylogenetic data, however, have shown that the two genera are not monophyletic (Mitchell et al. 2009) and should be merged under the name Melicytus, a move supported by diagnostic characters that have been shown to overlap (Beuzenberg 1961; Green 1970; Connor and Edgar 1987; Molloy and Druce 1994). All species in this lineage are polyploids (2n = 32, 48, 64, 96), and many hybridize in nature, resulting in a complex and reticulating allopolyploid history. The genus is endemic to New Zealand (where it may have originated), easternmost Australia, and adjacent South Pacific Islands (Mitchell et al. 2009). Species are hermaphroditic or dioecious, bear weakly zygomorphic flowers and produce fleshy, blueberry-like fruits instead of capsules. Other recent studies (Hoffmann et al. 2010; Paula-Souza 2009) have shown that the same kind of variation in carpel number and sexuality of flowers also occurs in Anchietea. THE HYBANTHUS GUANACASTENSIS GROUP—Hybanthus denticulatus and H. guanacastensis were recovered in yet another Hybanthus clade. The combination of characters that distinguish the group 2014] WAHLERT ET AL.: PHYLOGENY OF THE VIOLACEAE from other Hybanthus groups include 1–2 flowered axillary inflorescences borne on short shoots, strongly zygomorphic perianth and androecium, lack of staminal glands but the free portions of filament subtending the anthers conspicuously enlarged or swollen, seeds shiny purple-black and minutely alveolate, and 6–9 seeds per capsule. This Central American group, potentially including H. hespericlivus H. E. Ballard, Wetter & N. Zamora, represents another generic segregate lineage that warrants taxonomic recognition upon further morphological characterization. AGATEA, ANCHIETEA, HYBANTHOPSIS, AND CORYNOSTYLIS—The four lianescent genera of the Violaceae, Agatea, Anchietea, Hybanthopsis, and Corynostylis, were resolved in a well supported clade (90/94/1.0; Figs. 2, 3). The genera are united by a vining/lianescent habit, strongly flattened seeds, and certain characters of wood anatomy in Agatea, Anchietea, and Corynostylis (Taylor 1972). A hypothesis of a close evolutionary relationship was rejected by Taylor (1972) because of the large geographical disjunction found in the group (i.e. Agatea in New Guinea-South Pacific islands and the rest in the Neotropics), and he suggested the several shared features of wood anatomy were the consequence of convergent evolution. However, the phylogenetic results presented here have confirmed a well supported relationship demonstrated in Tokuoka (2008) and have also shown that such amphi-oceanic disjunctions are not uncommon between closely related genera of the Violaceae. The finding that the monotypic genus Hybanthopsis, a recently described twining herbaceous plant from Brazil (Paula-Souza and Souza 2003), belongs to this lineage suggests that the twining/lianescent habit is a synapomorphy for this group. Implications for a Revised Intrafamilial Classification for the Violaceae—The molecular phylogeny presented here provides new views on generic diversity and intrafamilial groupings in the Violaceae. Intrafamilial taxa for the family were almost wholly rejected by our phylogenetic inferences, with only a few clades congruent with the original circumscription (Table 2). Previously unsuspected generic affinities were revealed, and while some relationships appear to be discordant with morphology (e.g. Leonia and Gloeospermum), many groupings can be re-interpreted upon reexamination of morphological characters in combination with other characteristics of anatomy, chromosome numbers, embryology, etc. A completely revised intrafamilial classification of the whole family is now necessary in order to re-circumscribe higher-level groups that better reflect relationships in the family. Future circumscriptions of intrafamilial taxa need to be based on different character states instead of the homoplasious and continuous state of floral symmetry, which has resulted in an artificial classification of the family. The phylogenetic hypotheses presented in this study will serve as a springboard for the description of new genera and recircumscription of intrafamilial taxa, as well as a new framework from which trait evolution and biogeographic patterns can be inferred. Much additional research needs to be done, however, across all known genera and unnamed segregate groupings within Rinorea and Hybanthus. Research utilizing floral development (Feng 2005), leaf anatomy (Hoyos-Gómez 2011), pollen morphology (Mark et al. 2012), and calcium oxalate crystal morphology and distribution (Ballard, unpubl. data) has highlighted taxonomically informative character states that generally support higher-level 249 relationships inferred from DNA sequence data, and provide new insights into evolutionary processes and biogeographic affinities. Chromosomal evidence may also serve to add additional support to clades and to delimit intrafamilial taxa. While there are ca. 1,300 chromosome numbers presently available, most have been obtained from Viola. The counts published so far suggest base chromosome numbers of x = 4, 6, 7, or 8 for individual groups within the family, but the relationships among these numbers are not known. A future focus should be placed on acquisition of chromosome numbers for poorly sampled genera to reinterpret cytogenetic evolution in the Violaceae from within a phylogenetic context. Reported chromosome counts from Violaceae demonstrate extensive polyploidy within several genera and among species (e.g. Miyaji 1913; Brizicky 1961; Turner and Escobar 1991). Recent work on Viola has uncovered extensive allopolyploidy (Marcussen et al. 2011; Marcussen et al. 2012), and duplication of several low-copy nuclear genes appear to suggest at least one polyploidization event relatively early in the diversification of the family (Marcussen et al. 2010). The additional sequencing of unlinked nuclear genes will doubtless be necessary for the construction of a robust phylogeny and a phylogeny-based classification of the family. Many of the deeper relationships within the family are still unresolved, and little is known about past reticulations via allopolyploidy, a mechanism not detectable with the markers used so far owing to the maternal inheritance of the chloroplast (Harris and Ingram 1991) and concerted evolution in rDNA loci (Álvarez and Wendel 2003). All of the under-utilized avenues of investigation discussed above deserve further exploitation in the Violaceae. Certainly, there is strong indication from molecular phylogenetic studies that the reliance on floral symmetry (i.e. “actinomorphy” vs. zygomorphy) alone provides misleading inferences of relationships and heterogeneous generic circumscriptions, whereas less evolutionarily labile traits and micromorphological characters may be more broadly conserved and useful at circumscribing intrafamilial taxa. Until a full understanding of taxonomic diversity, relationships, and character distribution is available for all recognizable groups, our ability to produce a meaningful, predictive, and natural classification will remain out of our grasp. Acknowledgments. The authors would like to thank curators at the following institutions for access to herbarium material: BLH, BNRH, ESA, F, G, K, L, LL, MG, MO, NY, P, PRE, SPF, TAN, TEF, TEX, US, WAG, WIS. Assistance in the field was provided by Bil Alverson, Harvey Ballard Sr., Jean-Jacques de Granville, Barry Hammel, AnaLu MacVean, Ross McCauley, David Neill, Antonio Vazquez, Mark Wetter, Nelson Zamora, John and Sandie Burrows, Mervyn Lötter, Franck Rakotonasolo, and Hanta Razafindraibe. We wish to thank the people who provided leaf tissue or extractions for use in this phylogeny: Gaston Achoundong, Tracy Nowell, Robin Van Velzen, Jay Bolin, Martin Callmander, Neil Crouch, and P. Peter Lowry, II. The Editor-in-Chief (Tom Ranker) and an Associate Editor at Systematic Botany and two anonymous reviewers provided insightful reviews and thoughtful commentary that greatly improved the manuscript. Funding was provided by Ohio University Department of Environmental and Plant Biology and the Graduate Student Senate to GAW; the Myndel Botanical Foundation, International Association for Plant Taxonomy, Society of Systematic Biologists, São Paulo Research Foundation, and University of São Paulo to JPS; SigmaXi to MF; and the National Science Foundation (DEB-0211054) to HEB. The use of some DNA sequences from Scott Hodges and Mark Chase is gratefully acknowledged. 250 SYSTEMATIC BOTANY Literature Cited Álvarez, I. and J. F. Wendel. 2003. Ribosomal ITS sequences and plant phylogenetic inference. Molecular Phylogenetics and Evolution 29: 417–434. Angiosperm Phylogeny Group. 2009. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III. Botanical Journal of the Linnean Society 161: 105–121. Baillon, H. E. 1873. Violacées. Pp. 265–356 in Histoire des Plantes vol. 4. Paris: L. Hachette et Cie. Ballard, H. E. Jr. 1996. Phylogenetic relationships and infrageneric groups in Viola (Violaceae) based on morphology, chromosome numbers, natural hybridization and Internal Transcribed Spacer (ITS) sequences. Ph. D. thesis. Madison, Wisconsin: University of Wisconsin. Ballard, H. E. Jr., J. Paula-Souza, and G. A. Wahlert. 2013. Violaceae. In The Families and Genera of Vascular Plants, ed. K. Kubitzki, in review. Berlin: Springer-Verlag. Ballard, H. E. Jr., K. J. Sytsma, and R. R. Kowal. 1999. Shrinking the violets: Phylogenetic relationships of infrageneric groups in Viola (Violaceae) based on internal transcribed DNA sequences. Systematic Botany 23: 439–458. Bentham, G. and J. D. Hooker. 1862. Genera Plantarum, vol. 1, part 1. London: Reeve and Co. Beuzenberg, E. J. 1961. Observations on sex differentiation and cytotaxonomy of the New Zealand species of the Hymenantherinae (Violaceae). New Zealand Journal of Science 4: 337–349. Brizicky, G. K. 1961. The genera of Violaceae in the southeastern United States. Journal of the Arnold Arboretum 42: 321–333. Chase, M. W. and J. L. Reveal. 2009. A phylogenetic classification of the land plants to accompany APG III. Botanical Journal of the Linnean Society 161: 122 – 127. Chase, M. W., D. E. Soltis, R. G. Olmstead, D. Morgan, D. H. Les, B. D. Mishler, M. R. Duvall, R. A. Price, H. G. Hills, Y. L. Qiu, K. A. Kron, J. H. Rettig, E. Conti, J. D. Palmer, J. R. Manhart, K. J. Sytsma, H. J. Michaels, W. J. Kress, K. G. Karol, W. D. Clark, M. Hedren, B. S. Gaut, R. K. Jansen, K. J. Kim, C. F. Wimpee, J. F. Smith, G. R. Furnier, S. H. Strauss, Q. Y. Xiang, G. M. Plunkett, P. S. Soltis, S. M. Swensen, S. E. Williams, P. A. Gadek, C. J. Quinn, L. E. Eguiarte, E. Golenberg, G. H. Learn, S. W. Graham, S. C. H. Barrett, S. Dayanandan, and V. A. Albert. 1993. Phylogenetics of seed plants: An analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528–580. Connor, H. E. and E. Edgar. 1987. Name changes in the indigenous flora, 1960–1986 and Nomina Nova IV, 1983–1986. New Zealand Journal of Botany 25: 115–170. Doyle, J. J. and J. L. Doyle. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15. Edgar, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32: 1792–1797. Edwards, K., C. Johnstone, and C. Thompson. 1991. A simple and rapid method for the preparation of plant genomic DNA for PCR analysis. Nucleic Acids Research 19: 1349. Farris, J. S., M. Källersjö, A. G. Kluge, and C. Bult. 1995. Constructing a significance test for incongruence. Systematic Biology 44: 570–572. Fay, M. F., S. M. Swensen, and M. W. Chase. 1997. Affinities of Medusagyne oppositifolia (Medusagynaceae). Kew Bulletin 52: 111–120. Feng, M. 2005. Floral morphogenesis and molecular systematics of the family Violaceae. Ph. D. thesis. Athens, Ohio: Ohio University. Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39: 783–791. Gingins de la Sarraz, F. C. J. de. 1823. Mémoire sur la Famille des Violacées. Mémoires de la Société de Physique et d´Histoire naturelle de Genéve 2: 1–28. Green, P. S. 1970. Notes relating to the floras of Norfolk and Lord Howe Islands. I. Journal of the Arnold Arboretum 51: 204–220. Harris, S. A. and R. Ingram. 1991. Chloroplast DNA and biosystematics: the effects of intraspecific diversity and plastid transmission. Taxon 40: 393–412. Hekking, W. H. A. 1988. Monograph 46. Violaceae Part I–Rinorea and Rinoreocarpus. Flora Neotropica Monograph 26: 1–207. Hoffmann, M. H., J. Paula-Souza, A. Fläshchendräger, and M. Röser. 2010. The gynoecium of male Anchietea pyrifolia (Violaceae): Preserved structure with a new function. Flora 205: 429–433. Hoyos-Gómez, S. E. 2011. Towards an understanding of the basal evolution of Violaceae from an anatomical and morphological perspective. Master’s thesis. St. Louis, Missouri: University of Missouri. [Volume 39 Jobb, G., A. von Haeseler, and K. Strimmer. 2004. TREEFINDER: a powerful graphical analysis environment for molecular phylogenetics. BMC Evolutionary Biology 4: 18. Kumar, S., Å. Skjæveland, R. S. J. Orr, P. Enger, T. Ruden, B.-H. Mevik, F. Burki, A. Botnen, and K. Shalchian-Tabrizi. 2009. AIR: A batchoriented web program package for construction of supermatrices ready for phylogenomic analyses. BMC Bioinformatics 10: 357. Kuntze, O. 1891. Revisio Generum Plantarum, vol. 1. Leipzig: Arthur Felix. Marcussen, T., K. Blaxland, M. D. Windham, K. E. Haskins, and F. Armstrong. 2011. Establishing the phylogenetic origin, history and age of the narrow endemic Viola guadalupensis (Violaceae). American Journal of Botany 98: 1–11. Marcussen, T., K. S. Jakobsen, J. Danihelka, H. E. Ballard, K. Blaxland, A. K. Brysting, and B. Oxelman. 2012. Inferring species networks from gene trees in high-polyploid North American and Hawaiian violets (Viola, Violaceae). Systematic Biology 61: 107–126. Marcussen, T., B. Oxelman, A. Skog, and K. S. Jakobsen. 2010. Evolution of plant RNA polymerase IV/V genes: evidence of subneofunctionalization of duplicated NRPD2/NRPE2-like paralogs in Viola (Violaceae). BMC Evolutionary Biology 10: 1. Mark, P. J. L., A. H. Wortley, and C. A. Furness. 2012. Not a shrinking violet: pollen morphology of Violaceae (Malpighiales). Grana 51: 181–193. Mayers, A. M. and E. M. Lord. 1983. Comparative flower development in the cleistogamous species of Viola odorata. I. A growth rate study. American Journal of Botany 70: 1548–1555. Melchior, H. 1925a. Violaceae. Pp. 329–377 in Die Natürlichen Pflanzenfamilien 21, eds. H. G. A. Engler and K. A. Prantl. Leipzig: Wilhelm Engelmann. Melchior, H. 1925b. III. Die phylogenetishe Entwicklung der Violaceen und die natürlichen Verwandtschaftsverhältnisse ihrer Gattungen. Feddes Repertorium Specierum Novarum Regni Vegetabilis. 36(Beiheft.): 83–125. Mitchell, A. D., P. B. Heenan, B. G. Murray, B. P. J. Molloy, and P. J. de Lange. 2009. Evolution of the south-western Pacific genus Melicytus (Violaceae): evidence from DNA sequence data, cytology and sex expression. Australian Systematic Botany 22: 143–157. Miyaji, Y. 1913. Untersuchungen über die Chromosomenzahlen bei einigen Viola-Arten. Botanical Magazine Tokyo 27: 443–460. [in Japanese]. Molloy, B. P. J. and A. P. Druce. 1994. A new species name in Melicytus (Violaceae) from New Zealand. New Zealand Journal of Botany 32: 113–118. Müller, K. 2005. SeqState–primer design and sequence statistics for phylogenetic DNA data sets. Applied Bioinformatics 4: 65–69. Nieuwland, J. A. and R. M. Kaczmarek. 1914. Studies in Viola. I. Proposed segregates of Viola. American Midland Naturalist 3: 207–217. Paula-Souza, J. and V. C. Souza. 2003. Hybanthopsis, a new genus of Violaceae from Eastern Brazil. Brittonia 55: 209–213. Paula-Souza, J. 2009. Estudos Filogenéticos em Violaceae com ênfase na Tribo Violeaee Revisão Taxonômica dos Gêneros Lianescentes de Violaceae na Região Neotropical. Ph. D. thesis. São Paulo: Universidade de São Paulo. Reiche, K. F. and P. H. W. Taubert. 1895. Violaceae. Pp. 322–336 in Die Natürlichen Pflanzenfamilien. 1st. ed. Teil 3, Heft 6, eds. H. G. A. Engler and K. A. Prantl. Leipzig: Wilhelm Engelmann. Rieseberg, L. H., M. A. Hanson, and C. T. Philbrick. 1992. Androdioecy is derived from dioecy in Datiscaceae: evidence from restriction site mapping of PCR-amplified chloroplast DNA fragments. Systematic Botany 17: 324–336. Ronquist, F. and J. P. Huelsenbeck. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. Simmons, M. P. and H. Ochoterena. 2000. Gaps as characters in sequencebased phylogenetic analyses. Systematic Biology 49: 369–381. Smith, J. F., K. J. Sytsma, J. S. Shoemaker, and R. L. Smith. 1992. A qualitative comparison of total cellular DNA extraction protocols. Phytochemical Bulletin 23: 2–9. Strimmer, K. and A. Rambaut. 2002. Inferring confidence sets of possibly misspecified gene trees. Proceedings. Biological Sciences 269: 137–142. Souza, J. P. 2002. Levantamento das espécies de Hybanthus Jacq. do Brasil. Master’s thesis: São Paulo: Universidade de São Paulo. Swofford, D. L. 2002. PAUP* Phylogenetic Analysis Using Parsimony, vers. 4.0b10. Sinauer Associates, Sunderland, MA. 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 Reporter 17: 1105–1109. Taylor, F. H. 1972. The secondary xylem of the Violaceae: a comparative study. Botanical Gazette (Chicago, Ill.) 133: 230–242. 2014] WAHLERT ET AL.: PHYLOGENY OF THE VIOLACEAE Thiers, B. 2013. Index Herbariorum: A global directory of public herbaria and associated staff. New York Botanical Garden’s Virtual Herbarium. http://sweetgum.nybg.org/ih/. Tokuoka, T. 2008. Molecular phylogenetic analysis of Violaceae (Malpighiales) based on plastid and nuclear DNA sequences. Journal of Plant Research 121: 253–260. Turner, B. L. and L. K. Escobar. 1991. Documented chromosome numbers 1991: 1. Chromosome numbers in Hybanthus (Violaceae). Sida 14: 501–503. Wahlert, G. A. 2010. Phylogeny, biogeography, and a taxonomic revision of Rinorea (Violaceae) from Madagascar and the Comoro Islands. Ph. D. thesis. Athens, Ohio: Ohio University. Wahlert, G. A. and H. E. Ballard Jr. 2009. A new zygomorphic-flowered Rinorea (Violaceae) from the Neotropics. Novon 19: 416–420. Wahlert, G. A. and H. E. Ballard Jr. 2012. A phylogeny of Rinorea (Violaceae) inferred from plastid DNA sequences with an emphasis on the African and Malagasy species. Systematic Botany 37: 964–973. Wurdack, K. J. and C. C. Davis. 2009. Malpighiales phylogenetics: Gaining ground on one of the most recalcitrant clades in the angiosperm tree of life. American Journal of Botany 96: 1551–1570. Xi, Z., B. R. Ruhfel, H. Schaefer, A. M. Amorim, M. Sugumarane, K. J. Wurdack, P. K. Endress, M. L. Matthews, P. F. Stevens, S. Mathews, and C. C. Davis. 2012. Phylogenomics and a posteriori data partitioning resolve the Cretaceous angiosperm radiation Malpighiales. Proceedings of the National Academy of Sciences USA 109: 17519–17524. Yockteng, R., H. E. Ballard Jr., G. Mansion, I. Dajoz, and S. Nadot. 2003. Relationships among pansies (Viola section Melanium) investigated using ITS and ISSR markers. Plant Systematics and Evolution 241: 153–170. Appendix 1. List of taxa used in phylogenetic analyses showing voucher information (species, geographic origin, collector and number, herbarium,) and GenBank accession number (trnL/trnL–F; rbcL). An asterisk (*) denotes a sequence obtained from GenBank; a dash (—) indicate the absence of a sequence for that locus. Herbarium acronyms follow Thiers (2013). Agatea longipedicellata (Baker f.) Guillaumin & Thorne. NEW CALEDONIA. Munzinger et al. 1760 (MO) KC699633; NEW CALEDONIA. McPherson 6152 (MO), KC699560. Agatea pancheri Brongn. NEW CALEDONIA. J. Munzinger & G. McPherson 831 (MO), KC699630; KC699561. Agatea schlechteri Melch. NEW CALEDONIA. J. Munzinger & G. McPherson 690 (MO), KC699632; AB354401*. Allexis cauliflora Pierre. GABON. Achoundong 12 (BHO), KC699709; KC699562. Allexis obanensis (Baker f.) Melch. GABON Wilks 2628 (MO), KC699634; AB354403*. Allexis zygomorpha Achound. & Onana. CAMEROON. Kribi s. n. (BHO), KC699635; KC699563. Amphirrhox longifolia (A. St.-Hil.) Spreng. FRENCH GUIANA. Ballard & Feng 02-102 (BHO), KC699636; KC699564. Amphirrhox surinamensis Eichler. BRAZIL. Assunção et al. 388 (MO), KC699637; AB354404*. Anchietea frangulaefolia (Kunth) Melch. COLOMBIA. Cuatrecasas 5477 (US), KC699638; —. Anchietea peruviana Melch. ECUADOR. Fierro 1273 (NY), KC699639; —. Anchietea pyrifolia (Mart.) G. Don. BRAZIL. Paula-Souza et al. 3685 (ESA), KC699640; KC699565. Anchietea selloviana Cham. & Schltdl. BRAZIL. Paula-Souza et al. 5608 (ESA), KC699641; KC699566. Corynostylis arborea (L.) S.F. Blake. FRENCH GUIANA. Ballard & Feng s. n. (BHO), KC699642; AB354408*. Corynostylis sp. BRAZIL. Kubitzki 84-336 (MG), KC699643; KC699567. Decorsella paradoxa Chev. GABON. de Wilde et al. 495 (MO), KC699644; KC699568. Fusispermum laxiflorum Hekking. JN714121*; FJ670178*. Gloeospermum diversipetalum L.O. Williams. COSTA RICA. Alverson s. n. (BHO), KC699645; KC699569. Gloeospermum grandifolium Hekking. ECUADOR. Ballard 02-305 (BHO), KC699646; KC699570. Hekkingia bordenavei H.E. Ballard & Munzinger. FRENCH GUIANA. Bordenave 5040 (P), KC699647; KC699571. Hybanthopsis bahiensis PaulaSouza. BRAZIL. Paula-Souza et al. 5419 (ESA), KC699648; KC699572. Hybanthus arenarius Ule. BRAZIL. Paula-Souza et al. 4673 (ESA), KC699649; KC699573. Hybanthus atropurpureus (A. St.-Hil.) Taub. BRAZIL. Paula-Souza 3639 (ESA), KC699650; KC699574. Hybanthus attenuatus (Humb. & Bonpl. ex Schult.) Schulze-Menz. MEXICO. Escobedo 2108 (WIS), KC699651; KC699575. Hybanthus aurantiacus (Benth.) F. Muell. AUSTRALIA Lazarides & Palmer 135 (MO), KC699652; KC699576. Hybanthus bigibbosus (A. St.-Hil.) Hassl. BRAZIL. Paula-Souza 3638 (ESA), KC699653; KC699577. Hybanthus buxifolius (Vent.) Baill. MADAGASCAR. Dorr & Barnett 4451 (MO), KC699654; —. Hybanthus calceolaria (L.) Oken. BRAZIL. Paula-Souza et al. 4924 (ESA), KC699655; 251 KC699578. Hybanthus caledonicus (Turcz.) Cretz. NEW CALEDONIA. Tronche & Lowry 616 (MO), KC699656; KC699579. Hybanthus calycinus (DC.) F. Muell. AUSTRALIA. Strid 20287 (MO), KC699657; KC699580. Hybanthus communis (A. St.-Hil.) Taub. BRAZIL. Paula-Souza et al. 5464 (ESA), KC699658; KC699581. Hybanthus concolor (T. F. Forst.) Spreng. U. S. A. Ballard 92-013 (WIS), KC699659; KC699582. Hybanthus danguyanus H. Perrier. KENYA. Mwangangi & Fosberg 617B (MO), KC699660; —. Hybanthus denticulatus H.E. Ballard, M.A.Wetter & N.Zamora. PANAMA. Wahlert 140 (BHO), KC699661; KC699583. Hybanthus enneaspermus (L.) F. Muell. (1). GHANA. Morello et al. 1264 (MO), KC699662; AB354418*; (2), SOUTH AFRICA. Nowell 155 (BHO), KC699663; KC699584. Hybanthus floribundus (Lindl.) F. Muell. AUSTRALIA. DQ407242*; —. Hybanthus fruticulosus (Benth.) I.M. Johnst. MEXICO. Sanders 9166 (TEX), KC699664; KC699585. Hybanthus galeottii (Turcz.) C.V. Morton. MEXICO. Breedlove 57023 (TEX), KC699665; KC699586. Hybanthus guanacastensis Standl. COSTA RICA. Ballard 00-007 (BHO), KC699666; KC699587. Hybanthus havanensis Jacq. DOMINICAN REPUBLIC Liogier 7546 (F), KC699667; KC699588. Hybanthus leucopogon Sparre. ARGENTINA. Paula-Souza et al. 7000 (ESA), KC699668; KC699589. Hybanthus mexicanus Ging. MEXICO. Ballard s. n. (BHO), KC699669; KC699590. Hybanthus micranthus Guillaumin. NEW CALEDONIA. Munzinger JM476 (MO), KC699670; —. Hybanthus monopetalus Domin. AUSTRALIA. Crisp 1625 (US), KC699671; —. Hybanthus nanus (A. St.-Hil.) Paula-Souza. ARGENTINA. Paula-Souza et al. 7045 (ESA), KC699672; KC699591. Hybanthus oppositifolius (L.) Taub. BRAZIL. Paula-Souza et al. 5465 (ESA), KC699673; KC699592. Hybanthus parviflorus (Mutis ex L f.) Baill. BRAZIL. Paula-Souza et al. 3649 (ESA), KC699674; KC699593. Hybanthus prunifolius (Humb. & Bonpl. ex Schult.) Schulze-Menz. PANAMA. Garwood 1297A (F), KC699675; —. Hybanthus serrulatus Standl. MEXICO. Lott & Butterwick 1534 (F), KC699676; KC699594. Hybanthus setigerus (A. St.-Hil.) Baill. BRAZIL. Paula-Souza et al. 4248 (ESA), KC699677; KC699595. Hybanthus sp. nov. MEXICO. McCauley 440 (BHO), KC699678; KC699596. Hybanthus thiemei (Donn. Sm.) C.V. Morton. GUATEMALA. Ballard 02-314 (BHO), KC699679; KC699597. Hybanthus verbenaceus (Kunth) Loes. MEXICO. Breedlove 10524 (F), KC699680; —. Hybanthus verticillatus (Ortega) Baill. MEXICO. Cowan 5436 (TEX), KC699681; —. Hybanthus yucatanensis Millsp. MEXICO. Martı́nez 35211 (ESA), KC699682; KC699598. Isodendrion hosakae H. St. John. U. S. A. Hawaiian Islands, Wagner et al. 5301 (US), KC699683; —. Isodendrion laurifolium A. Gray. U. S. A. Hawaiian Islands, Perlman 5481 (MO), KC699684; KC699599. Isodendrion longifolium A. Gray. U. S. A. Hawaiian Islands, Perlman et al. 6295 (MO), KC699685; AB354421*. Leonia cymosa Mart. BRAZIL. Paula-Souza et al. 9528 (SPF), KC699686; KC699600. Leonia glycycarpa Ruiz & Pav. AY739763*; SURINAME. Evans & Peckham 2917 (BLH), KC699601. Malesherbia lanceolata Ricardi. trnL intron/trnL-trnF spacer: AY636104*. Malesherbia weberbaueri Gilg. rbcL: AY632722*. Mayanaea caudata (Lundell) Lundell. GUATEMALA. Lundell 11442 (MO), KC699687; KC699602. Melicytus alpinus (Kirk) Garn.-Jones. NEW ZEALAND. Holloway & Knowles CHR369010 (MO), KC699688; KC699603. Melicytus chathamicus (F. Muell.) Garn.-Jones. Menzies CHR359614 (MO), KC699689; KC699604. Melicytus dentatus (R. Br.) Molloy & Mabb. AUSTRALIA. Adams 2302 (US), KC699690; —. Melicytus lanceolatus Hook. f. NEW ZEALAND. Gardner 1954 (MO), KC699691; —. Melicytus latifolius (Lindl.) P. S. Green. AUSTRALIA. Norfolk Island, Gardner 5925 (MO), KC699692; —. Melicytus novae-zelandiae (A. Cunn.) P. S. Green. NEW ZEALAND. Sykes 216/93 (MO), KC699693; AB354427*. Noisettia orchidiflora (Rudge) Ging. JF767152*; FRENCH GUIANA. Mori & Pennington 17950 (MO), KC699620. Orthion montanum Lundell. MEXICO. Breedlove 49920 (MO), KC699694; KC699605. Orthion subsessile (Standl.) Standl. & Steyerm. GUATEMALA. Lundell & Contreras 18894 (LL), KC699695; —. Orthion subsessile (Standl.) Standl. & Steyerm. Reyes-Garcı́a & Challenger 2555 (MO), KC699696; AB233941*. Passiflora auriculata Kunth. DQ284534*; DQ445921*. Paypayrola blanchetiana Tul. BRAZIL. Paula-Souza et al. 5682 (ESA), KC699697; KC699606. Paypayrola confertiflora Tul. FRENCH GUIANA. Feng 497 (BHO), KC699698; KC699607. Paypayrola grandiflora Tul. BRAZIL. Paula-Souza et al. 9527 (SPF), KC699699; KC699608. Rinorea anguifera Kuntze. THAILAND. Larsen et al. 43071 (MO), JN714074*; KC699618. Rinorea apiculata Hekking. ECUADOR. Tipaz & Rubio 336 (MO), JN714120*; AB354430*. Rinorea arborea Baill. COMOROS. Labat et al. 3197 (P), JN714094*; KC699619. Rinorea bullata H. Perrier. MADAGASCAR. Skema et al. 217 (P), JN714069*; KC699621. Rinorea crenata S.F. Blake. COSTA RICA. Ballard 94-006 (BHO), JN714119*; KC699622. Rinorea elliptica Kuntze. MOZAMBIQUE. Burrows & Wahlert 10128 (BNRH), JN714084*; KC699623. Rinorea flavescens Kuntze. FRENCH GUIANA. 252 SYSTEMATIC BOTANY Ballard & Feng 02-100 (BHO), JN714112*; KC699624. Rinorea horneri Kuntze. THAILAND. Larsen et al. 41308 (MO), JN714080*; KC699625. Rinorea ilicifolia Kuntze. JN714037*; AB354432*. Rinorea mutica Baill. MADAGASCAR. Wahlert & Rakotonasolo 12 (MO), JN714067*, KC699626. Rinorea neglecta Sandwith. FRENCH GUIANA. Prevost 3457 (MO), KC699700; AB354433*. Rinorea riana Kuntze. FRENCH GUIANA. Saint Jean Forest, Ballard & Feng 02-104 (BHO), JN714106*; KC699627. Rinorea sp. BRAZIL. Acre, Souza 30019 (ESA), JN714107*; KC699628. Rinorea subintegrifolia Kuntze. NIGERIA. Brown & Opayemi 995 (MO), JN714058*; KC699629. Rinorea uxpanapana T. Wendt. MEXICO. Wendt et al. 3907 (MO), KC699701; KC699609. Rinorea virgata (Thwaites) Kuntze. THAILAND. Larsen et al. 2596 (L), KC699702; KC699610. Rinorea [Volume 39 virgata. THAILAND. Larsen et al., 42122 (MO), KC699703; KC699611. Rinoreocarpus ulei Ducke. AY739759*; AB354435*. Schweiggeria fruticosa Spreng. BRAZIL. Paula-Souza 3612 (ESA), KC699631; KC699612. Turnera ulmifolia L. AY636110*; Z75691*. Viola biflora L. DQ085922*; —. Viola blanda A. Gray. U. S. A. Michigan: Ballard 92-004 (BHO), KC699704; KC699613. Viola canadensis L. U. S. A. Michigan: Ballard 92-012 (BHO), KC699705; KC699614. Viola cerasifolia A. St.-Hil. BRAZIL. Paula-Souza et al. 5800 (ESA), KC699706; KC699615. Viola chaerophylloides Makino. DQ787749*; —. Viola collina Besser. DQ085887*; —. Viola grypoceras A. Gray. DQ085891*; —. Viola subdimidiata A. St.-Hil. BRAZIL. PaulaSouza et al. 5853 (ESA), KC699707; KC699616. Viola tricolor L. U. S. A. Wisconsin: Ballard s. n. (BHO), KC699708; KC699617.