Research Paper Phylogeny of the Genus Peperomia (Piperaceae) Inferred from the trnK/matK Region (cpDNA) S. Wanke1, M.-S. Samain2, L. Vanderschaeve2, G. Mathieu2, P. Goetghebeur2, and C. Neinhuis1 1 Institut für Botanik, Plant Phylogenetics and Phylogenomics Group, Technische Universität Dresden, 01062 Dresden, Germany 2 Department of Biology, Research Group Spermatophytes, Ghent University, K. L. Ledeganckstraat 35, 9000 Gent, Belgium Received: June 13, 2005; Accepted: October 13, 2005 Abstract: The genus Peperomia is one of the largest genera of basal angiosperms, comprising about 1500 – 1700 pantropically distributed species. The currently accepted infrageneric classification divides Peperomia into nine subgenera and seven sections. This classification is based on some 200 species, primarily using fruit morphology. The monophyly of these infrageneric taxa has never been tested and molecular phylogenetic studies of a representative sampling within Peperomia do not exist. This paper provides the first molecular phylogeny for the genus Peperomia. Monophyletic clades within Peperomia are identified and previously used morphological characters are critically reviewed. We show that the importance of some morphological characters has been overestimated and that some of these characters presumably have evolved several times independently. Only one previously described subgenus has been confirmed to be monophyletic. Key words: Peperomia, phylogeny, trnK intron, matK, Bayesian inference, giant genera. Introduction The genus Peperomia includes 1500 – 1700 accepted species making it one of the largest genera of basal angiosperms. However, 3031 names are published, making the taxonomic study of this huge genus an intricate search. As a result of this huge number of names and taxa and because of the difficult synonymy, the actual taxonomic knowledge of the genus is quite chaotic. Peperomia species are distributed pantropically. The greatest diversity occurs in the Neotropics, followed by Southern Asia (about 100 species), Africa (about 20 species), Madagascar (about 40 species), and Australia and New Zealand (less than 20 species). The genus Peperomia Ruiz and Pavón (1794), together with another huge genus, Piper, constitutes the core of the Piperaceae. This family is part of Piperales, a clade that also includes Aristolochiaceae and Saururaceae, Lactoridaceae and Hydnoraceae Plant Biol. 8 (2006): 93 – 102 © Georg Thieme Verlag KG Stuttgart · New York DOI 10.1055/s-2005-873060 ISSN 1435-8603 (APG1 1998, APG2, 2003) although the exact position of the latter two families within the order remains unclear. Within Piperales, Peperomia has also been assigned to its own family, Peperomiaceae (Novak, 1954; Burger, 1977; Airy-Shaw, 1966, 1987; Brummit, 1992; Heywood, 1978; Walker, 1976 a, b; Smith, 1981) or subfamily Peperomioideae (Thorne, 1992). The current infrageneric classification, which we used as reference classification, includes some 200 species and is primarily based on fruit morphology (Dahlstedt, 1900). The relationships and groupings described by Dahlstedt in the introduction, circumscription of subgenera and in his “phylogenetic tree” are completely incongruent with each other, e.g., the subgenus Tildenia which forms a clade with Panicularia according to the “tree”, but is more or less related with every subgenus according to the text. The earliest classification of Piperaceae which mentioned Peperomia dates back to Miquel (1843). Within the tribe Peperomieae he recognized 5 genera, from which 4 are now included in Peperomia, and four subgenera. The exact position of the fifth genus, Verhuellia, is unknown within Piperaceae. A detailed analysis of Verhuellia will be published elsewhere. Dahlstedt (1900) divided Peperomia into 9 subgenera and seven sections. Trelease (1930) raised two of Dahlstedt’s sections to subgeneric level, to a total of 11 subgenera. To this, 3 more subgenera were added: Hawaiiana (Yuncker, 1933), Treleaseanum (Stehlé, 1940), and Tildeniidium (Skottsberg, 1947). Burger (1971) questioned the development of the pseudocupula that typifies the subgenus Micropiper. Yuncker (1953) mentioned four of Dahlstedt’s subgenera: Tildenia p.p., Micropiper, Sphaerocarpidium, and Rhyncophorum p.p. He merged Acrocarpidium and Pleurocarpidium in an unnamed group, VI. Further he merged Panicularia and some species of Tildenia in his group VIII. In addition he cited other unnamed groups with numerous species: II and VII. However in 1974 Yuncker mentioned five of Dahlstedt’s subgenera: Acrocarpidium (in which he included Pleurocarpidium), Tildenia (in which he included Ogmocarpidium), Micropiper, Sphaerocarpidium, and Rhyncophorum. This confused grouping caused us to use the more detailed study of Dahlstedt as the basis for our discussion and for our sampling. An overview of the taxonomic history of Peperomia is given in Table 1. Most species descriptions after 1900 are published without reference to their subgeneric affiliation. Additionally, the monophyly of the infrageneric groups has never been tested. Phylogenetic studies based on molecular data of a 93 94 Plant Biology 8 (2006) Table 1 S. Wanke et al. Summary of the taxonomic history of the genus Peperomia (GE = genus, SG = subgenus, SE = section) Miquel, 1843 Henschen, 1873 Dahlstedt, 1900 De Candolle, 1923 Trelease, 1930 Yuncker, 1974 Peperomia (GE) Peperomia (GE) Peperomia (GE) Peperomia (GE) Peperomia (GE) Peperomia (GE) Piperanthera (GE) Piperanthera (GE) Acrocarpidium (GE) Acrocarpidium (SE) Acrocarpidium (SG) Acrocarpidium (SG) Eupeperomia (SE) Erasmia (GE) Tildenia (SE) Ogmocarpidium (SG) Ogmocarpidium (SG) Erasmia (SG) Erasmia (SG) Pleurocarpidium (SG) Pleurocarpidium (SG) Tildenia (SG) Tildenia (SG) Eutildenia (SE) Rhyncophorum (SE) Rhyncophorum (SE) Tildenia (SG) (+ Ogmocarpidium) Hemirhyncophorum (SE) Panicularia (SE) Acrocarpidium (SG) (+ Pleurocarpidium) Hemirhyncophorum (SG) Panicularia (SG) Panicularia (SG) Rhyncophorum (SG) Rhyncophorum (SG) Rhyncophorum (SG) Oxyrhynchum (SE) Malacorhynchum (SE) Leptorhynchum (SE) Leptorhynchum (SG) Sphaerocarpidium (SG) Sphaerocarpidium (SG) Sphaerocarpidium (SG) Micropiper (SG) Micropiper (SG) Alternifoliae (SE) Verticillatae (SE) Micropiper (SE) Micropiper (SE) Micropiper (SG) Phyllobryon (GE) Verhuellia (GE) representative sampling within Peperomia do not yet exist, but a molecular and morphological study regarding the Hawaiian and South-Pacific Peperomia species was carried out by Bradley (2002), with the emphasis on biogeography and speciation. The monophyly of the genus in its most narrow circumscription (excluding Verhuellia) has been proven by a molecular approach (Neinhuis et al., 2004; Jaramillo et al., 2004), flower morphology (Jaramillo et al., 2004), and pollen ultrastructure (Mathew and Mathew, 2001). From a morphological point of view, several synapomorphies can be cited: a single carpel, two unilocular stamens, a 16-nucleate female gametophyte, an ovule with a single integument, small pollen grains that lack apertures, herbaceous habit, and succulent leaves (e.g., Bornstein, 1991; Tebbs, 1993; Tucker et al., 1993). The aim of this study is to infer relationships within the genus Peperomia and to test the monophyly of previously described clades. As we deal with one of the largest genera in angiosperms, a well resolved tree for Peperomia is needed to address all further questions within this genus. To achieve this goal, we decided to use the chloroplast gene matK, because it is known for the highest variability among chloroplast genes (Olmstead and Palmer, 1994; Hilu and Liang, 1997; Borsch et al., 2005). The matK gene is situated within the group II intron of the tRNA gene for lysine (UUU) and encodes a maturase (Neuhaus and Link, 1987). The gene has been used successfully in several studies within genera to resolve phylogenies, e.g., Magnolia (Azuma et al., 1999), Paeonia (Sang et al., 1997), and Utricularia (Müller and Borsch, 2005). In addition the “cost-benefit calculation” of the matK gene has been shown to supersede any other plastid gene, as comparative studies indicate a better reso- Verhuellia (GE) Verhuellia (GE) lution at a similar number of nucleotides sequenced and a lower level of homoplasy (Hilu et al., 2003; Borsch et al., in press; Hilu et al., unpublished). The surrounding trnK intron has also been sequenced to increase the resolution, since it has been shown that this intron has a phylogenetic signal similar to the matK gene (e.g., Müller and Borsch, 2005). Non-coding regions, especially group II introns, have been considered to have a high potential in molecular phylogenetics (Kelchner, 2002; Löhne and Borsch, 2005). Materials and Methods Sampling and outgroup taxa We sampled 48 species representing 8 out of 9 subgenera and a majority of sections described by Dahlstedt (1900). As outgroup we choose the genus Piper (Piperaceae), which has been substantiated to be the sister group of Peperomia, as well as two species from Saururaceae, the sister of Piperaceae. Plant material was collected in the field and from species cultivated in botanical gardens (collections of the Botanical Garden of Ghent University, Belgium as well as from the Botanical Gardens of Bonn, Dresden, and Berlin, Germany). A list of the species sampled, along with collection localities, vouchers, and accession numbers is provided in Table 2. The name of clades (sections/subgenera) to which the species are assigned are given as well. Phylogeny of Peperomia Plant Biology 8 (2006) Table 2 List of investigated species in the present study. The classification for Peperomia follows Dahlstedt (1900) except for Tildenia – Geophila (Hill, 1907). An asterisk refers to species which are not mentioned in Dahlstedt (1900); if possible we assign species to a subgenus on the basis of morphological characters. The field and/or garden origin as well as the voucher information and the GenBank accession number are indicated along with the taxon name (usually species). (Samain and Vanderschaeve = S. and V.; Leuenberger and Hagemann = L. and H.) Species Affilation after Dahlstedt (1900) Field/garden origin Voucher GenBank Acc. P. andina var. pseudoperuviana Pino Tildenia – Geophila (Hill, 1907) S. and V. 2005-018 (GENT) DQ212717 P. argyreia (Miq.) Morr. P. bicolor Sodiro P. blanda (Jacq.) Kunth Tildenia Sphaerocarpidium * Sphaerocarpidium – Verticillatae – Platyphyllae Rhyncophorum – Oxyrhynchum Panicularia Peru, Carretera Cajamarca San Juan; priv. collection R. Mayer BG Berlin, 062-56-74-83 BG Berlin, 107-84-74-83 BG Berlin, 054-24-74-73 Wanke 051 (DD) Wanke 052 (DD) Wanke 055 (DD) DQ212734 DQ212761 DQ212763 Schwertfeger GH 13433 (B) S. and V. 2005-002 (GENT) DQ212753 DQ212720 P. clusiifolia (Jacq.) Hook. P. cotyledon Benth. P. cuspidilimba C. DC. P. dahlstedtii C. DC. P. dolabella Rauh and Kimnach P. dolabriformis Kunth P. emarginella (Sw. ex Wikstr.) C. DC. P. fagerlindii Yunck. P. fraseri C. DC. P. galioides Kunth P. glabella (Sw.) A. Dietr. P. gracillima S. Watson P. graveolens Rauh and Barthlott P. hoffmannii C. DC. P. hylophila C. DC. P. inaequalifolia Ruiz and Pav. P. incana (Haw.) Hook. P. lanceolatapeltata C. DC. P. lancifolia Hook. P. longespicata C. DC. P. macrostachya (Vahl) A. Dietr. P. magnoliifolia (Jacq.) A. Dietr. P. marmorata Hook. f. Micropiper Micropiper Tildenia – Geophila (Hill, 1907) Panicularia * Pleurocarpidium Rhyncophorum * Panicularia Sphaerocarpidium – Verticillatae – Leptophyllae Sphaerocarpidium – Alternifoliae – Macrophyllae Tildenia – Geophila (Hill, 1907) Panicularia * Micropiper Sphaerocarpidium * Sphaerocarpidium – Verticillatae – Leptophyllae Rhyncophorum – Leptorhynchum Tildenia Erasmia Rhyncophorum – Leptorhynchum Rhyncophorum – Leptorhynchum Rhyncophorum – Oxyrhynchum Tildenia P. maypurensis Kunth P. metallica Linden and Rodigas P. pellucida (L.) Kunth P. pereskiifolia (Jacq.) Kunth P. pernambucensis Miq. P. pitcairnensis C. DC. Tildenia * P. ppucuppucu Trel. P. prostrata Williams P. reptilis C. DC. P. rhombea Ruiz and Pav. Micropiper Sphaerocarpidium * * Micropiper Ogmocarpidium Micropiper Panicularia Sphaerocarpidium * BG Berlin, 062-58-74-83 Ecuador, along road Loja to Vilcabamba; BG Gent, 2002-2158 BG Berlin, 062-58-74-83 BG Gent, 2002-1192 priv. collection R. Mayer BG Gent, 2003-1627 Guadeloupe; BG Gent, 2002-1923 BG Berlin, 107-73-74-83 BG Berlin, 285-64-89-80 Peru; BG Gent, 2003-1644 Wanke 050 (DD) S. and V. 2005-005 (GENT) S. and V. 2005-019 (GENT) S. and V. 2005-016 (GENT) S. and V. 2005-010 (GENT) DQ212733 DQ212732 DQ212718 DQ212721 DQ212723 Wanke 054 (DD) Wanke 046 (DD) S. and V. 2005-007 (GENT) DQ212742 DQ212719 DQ212748 BG Bonn, 18749 Wanke 061 (DD) DQ212757 Mexico, La Paz; BG Bonn, 06005 BG Bonn, 20297 Wanke 060 (DD) DQ212716 Wanke 062 (DD) DQ212722 BG Gent, 2002-1932 Costa Rica, Guatuso; BG Berlin, 173-23-95-33 BG Berlin, 213-35-00-80 S. and V. 2005-008 (GENT) Wanke 047 (DD) DQ212730 DQ212758 GH 39585 (B) DQ212749 BG Gent, 2003-1628 BG Gent, 1900-4181 Ecuador; BG Gent, 2002-2275 BG Gent, 2003-1608 Costa Rica, Gobaya Soula; BG Berlin, 039-82-89-23 French Guyana, Gobaya Soula; BG Berlin, 039-34-89-20 southern Brasil; BG Bonn, 17527 BG Bonn, 11132 BG Bonn, 16189 S. and V. 2005-017 (GENT) S. and V. 2005-003 (GENT) S. and V. s.n. S. and V. 2005-015 (GENT) Wanke 048 (DD) DQ212745 DQ212750 DQ212755 DQ212741 DQ212744 L. and H. GH 26231 (B) DQ212752 Wanke 064 (DD) DQ212725 Wanke 006 (DD) Wanke 066 (DD) DQ212735 DQ212740 S. and V. 2005-009 (GENT) Schwertfeger GH 13455 (B) Schwertfeger GH 13432 (B) Wanke 007 (DD) DQ212738 DQ212726 DQ212751 DQ212762 Wanke 067 (DD) S. and V. 2005-013 (GENT) S. and V. 2005-012 (GENT) Wanke 053 (DD) DQ212728 DQ212759 DQ212739 DQ212731 BG Gent, 1996-1016 BG Berlin, 140-28-74-83 BG Berlin, 062-62-74-83 Pitcairn Island; BG Bonn, 17744 BG Berlin, s.n. BG Gent, 2003-1991 BG Gent, 1973-0161 BG Berlin, 224-08-95-80 continued → 95 96 Plant Biology 8 (2006) Table 2 S. Wanke et al. continued Species Affilation after Dahlstedt (1900) Field/garden origin Voucher GenBank Acc. P. rotundifolia (L.) Kunth Sphaerocarpidium – Alternifoliae – Microphyllae Rhyncophorum – Malacorhynchum Panicularia Brasil, Bahia, Camaca; BG Berlin, 166-08-83-20 BG Gent, 1977-0525 Ecuador, Loja; BG Gent, 2003-2003 Zaire, Irangi; BG Dresden s. n. Venezuela, Bocono-Mosquey, BG Gent, 1978-1342 BG Berlin, 078-06-97-80 BG Gent, 2003-1674 Costa Rica, Rio Reventazo; BG Berlin, 173-24-95-33 BG Gent, 1987-1247 Wanke 049 (DD) DQ212754 S. and V. 2005-011 (GENT) S. and V. 2005-004 (GENT) DQ212747 DQ212737 Wanke 068 (DD) S. and V. 2005-006 (GENT) DQ212760 DQ212729 Wanke 044 (DD) S. and V. 2005-001 (GENT) Wanke 045 (DD) DQ212727 DQ212736 DQ212756 S. and V. s.n. DQ212746 BG Gent, 1900-3913 Costa Rica, San Ignacio de Agosta; BG Berlin, 173-25-95-33 S. and V. s.n. Horich and San Jose GH 35591 (B) DQ212724 DQ212743 BG Bonn, 08120 BG Bonn, 00312 Borsch 3481 (BONN) Wanke 001 (DD) DQ212712 DQ212713 BG Bonn, 18143 Guatemala; BG Bonn, 18857 Wanke 070 (DD) Wanke 056 (DD) DQ212714 DQ212715 P. serpens (Sw.) Loud. P. sodiroi C. DC. P. spec. P. tetraphylla (Forst. f.) Hook. and Arn. P. trifolia (L.) A. Dietr. P. tristachya Kunth P. tuisana C. DC. P. urocarpa Fisch. and C. A. Mey. P. verschaffeltii Lem. P. vinasiana C. DC. Sphaerocarpidium * Micropiper Micropiper * Sphaerocarpidium * Rhyncophorum – Malacorhynchum Tildenia * Rhyncophorum – Leptorhynchum Outgroup Houttuynia cordata Thunb. Saururus chinensis Hort. ex Loudon Piper crocatum Ruiz and Pav. Piper spec. DNA-Isolation, amplification, and sequencing Sequence alignment and treatment of microstructural changes Total genomic DNA was isolated from fresh material. Details of protocols are given in Borsch et al. (2003) using a modified triple-extraction approach with CTAB following the miniprep procedure of Liang and Hilu (1996). Sequences were manually aligned using PhyDe (Müller et al., 2005) following alignment rules proposed by Borsch et al. (2003) and Löhne and Borsch (2005). Indels have been considered to be an additional source for phylogenetic information with a low level of homoplasy. The origin of indels was usually easy to recognize (simple sequence repeats). Based on the alignment, an additional indel matrix, compiled with SeqState (Müller, 2005) using the simple coding algorithm (Simmons and Ochoterena, 2000), was analyzed. The alignment and the indel matrix can be obtained from the first author upon request. The trnK/matK region was generally amplified in two parts with an overlapping region of 250 to 400 bp, using the primers shown in Table 3. Amplification profiles generally follow an initial denaturation 1 min at 94 8C, annealing 1 min at 48 8C, elongation 1.5 min at 72 8C, denaturation 1 min at 94 8C, 1 min at 50 8C, elongation 1.5 min at 72 8C (34 cycles), and a final extension 7 min at 72 8C. Due to the high sequence variability within the trnK intron as well as the matK gene and due to periodically occurring poly A or T mononucleotid microsatellites repeats, a large number of internal primers had to be specifically designed for the sequencing process. The Polymerase Chain Reaction (PCR) was carried out as a 25-μl reaction, containing 15 μl DNA template (1 : 100 delution of the total genomic DNA), 3.3 μl dNTP mix (1.25 mM each), 0.5 μl of each primer (20 pmol/μl), and 1 uTaq Polymerase (PeqLab). After separation by gel, the PCR products were then extracted using different commercial gel extraction kits and directly sequenced with an ABI PrismTM BigDye Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer) on ABI 310 or 377 automated sequencers or directly sequenced with a CEQ DTCS Quick Start Kit (Beckman Coulter) on CEQ 8000 automated sequencer, following the standard protocol provided with each kit. Tree reconstruction and their evaluation Parsimony analysis (MP) using PAUP 4.0 (Swofford, 1998) and PRAP (Müller, 2004) were conducted either based on substitutions with gaps treated as missing characters or a combined matrix including indels. PRAP generates command files which implements ratchet search strategies (Nixon, 1999) for PAUP. We used 10 random addition cycles of 200 ratchet iterations, using PRAP standard settings. A strict consensus tree was compiled from the shortest trees collected from the different tree islands. The evaluation of the MP tree was performed using the Bootstrap approach (1000 replicates), in PAUP. For further measurement of support, posterior probabilities were calculated using MrBayes V3.1 (Ronquist and Huelsenbeck, 2003). The GTR model of nucleotide substitution was Phylogeny of Peperomia Table 3 Plant Biology 8 (2006) Primers used in the present study Primer Sequence 5′ – 3′ MG1 MG15 Pi-matK-370R Pi-matK-950R Pe-matK-2000F Pi-matK-1820R AS-matK-460F Pi-matK-2030F Pi-matK-470F Pe-matK-2500R Pe-matK-2700F Pi-matK-730R Pi-matK-1060F Pi-matK-1480F ZiSa-trnK-F AAC ATC TTT CCT TTC ACA TAT CCT TTC TTC AAA ATA ACT TCG AAC TAG TCG TGG GTT YCC TAT ATC GCT CTT ACG CTA ATT TTC CCT CTT TGC AAA CCC GCA ATA CAA TCT GAA ATG TRT GGT TAA ACA CGT GCT Design GAT GCT AAT CTT AAT GGA TTT ATT TTC AAT TTT GAY CTC YAA TGC GGA GTA AAC TCA TGG AGC TTG ATT CCA TAG AGG AGA HGA GG TAT TGC GYT RCT GCA AAG CAT TTA TCG TTC AAC YG AAG TAC ATT TTT employed, assuming site-specific rate categories following a gamma distribution. The substitution model was tested using Modeltest (Posada and Crandall, 1998) resulting in a GTR + G + I submodel (Rodriguez et al., 1990). The a priori probabilities supplied were those specified in the default settings of the program MrBayes (version 3.1). Posterior probability (PP) distributions of trees were created using the Metropolis-coupled Markov chain Monte Carlo method and following search strategies suggested by Huelsenbeck et al. (2002) and Huelsenbeck et al. (2001). Only PP’s of 95 and higher were considered significant (alpha = 0.05). Four chains were run simultaneously (1000 000 generations), starting from random trees. Chains were collected every 10 generations and the respective trees were written to a tree file. Calculation of the consensus tree and of the posterior probability of clades was done based upon the trees sampled after the burn-in, and trees were compiled and drawn using TreeGraph (Müller and Müller, 2004). Results and Discussion GAT ATG TTG GAA A ATG G G GG AGG CG AAG CAT TG Liang and Hilu (1996) Liang and Hilu (1996) this study this study this study this study this study this study this study this study this study this study this study this study this study length mutations is beyond the scope of the paper and will be published elsewhere. The trees resulting from the different approaches are highly congruent and therefore only cited in the discussion if incongruent to each other. The MP ratchet search retains the shortest trees of 1649 steps (CI 0.714, RI 0.789, RC 0.563) from which the strict consensus tree was calculated (Fig. 1). Including the indel matrix the shortest trees needed 1988 steps (CI 0.684, RI 0.765, RC 0.523) (strict tree not shown). In the following paragraphs the bootstrap support (BS) is only given if differences were observed between the original matrix (first) and with indels included (second). BS support from both matrices were plotted on the strict consensus tree. The tree obtained by Bayesian inference is shown as phylogram in Fig. 2. As the support for nearly all branches reveals maximum values, PPs are only indicated by an asterisk on the respective tree if support is not significant (lower then 95). The PPs are indicated in the text. Alignment Phylogenetic relationships The alignment resulted in a total of 3204 characters, based on 2419 – 2737 bp depending on the taxa. GC content (32.2 %) of the studied region is comparable to other matK/trnK studies e.g., Müller et al. (2004). The length of the trnK 5′ intron ranges from 682 to 763 bp and the shorter trnK 3′ intron from 185 to 349 bp. The matK coding region varies from 1509 to 1557 bp (503 to 519 amino acids). A total of 182 indels were coded. Six small regions were excluded from the analysis due to uncertain homology. Within the trnK 5′ intron (position 500 to 507) and the matK gene (position 1616 – 1701) only one region was excluded, whereas the remaining four were excluded in the small trnK 3′ intron (position 2654 to 2712, 2855 to 2891, 2915 to 2971, 3117 to 3132). Indels were observed in both coding and non-coding regions. The allocation of indels within the different regions is: 86 within the trnK 5′ intron, 32 within the matK gene and 64 within the trnK 3′ intron. A detailed investigation of the molecular evolution especially focusing on the In all our analyses the monophyly of the genus Peperomia gains maximum support (100% BS/100% PP). This is also supported by several other molecular (Neinhuis et al., 2004; Jaramillo et al., 2004) and morphological (Jaramillo et al., 2004; Mathew and Mathew, 2001) studies, where only few species were represented. Compared to the sister group Piper, Peperomia shows a stronger reduction of floral structures e.g., the single carpel and only two stamens which are both partly “covered” by a bract. Within the genus, the tuberous Peperomia species (Geophila) (100% BS/100% PP), P. gracillima (representing the Central American species), P. andina var. pseudoperuviana, and P. dolabella (the latter two representing the Andean species) were resolved as sister to all other non-tuber forming species (100% BS/100% PP). Species of this group with their unusual life form were described by Dahlstedt (1900) in the subgenus Tildenia 97 98 Plant Biology 8 (2006) S. Wanke et al. Fig. 1 Phylogenetic relationships of the genus Peperomia, based on parsimony method. Bootstrap values from 1000 replicates are plotted along the branches, based on the original matrix (first), and the matrix including indels (second). Differences in branching patterns (indicated with an asterisk), between the two strict trees are marginal and described in the text. The subgeneric affiliation (second column) follows Dahlstedt (1900) except for Geophila (*), this section follows Hill (1907). P. lancifolia (!) has been described in the subgenus Erasmia by Dahlstedt (1900), but is clearly part of the subgenus Sphaerocarpidium. and section Tildenia (“Eutildenia”) on the basis of morphological characters (e.g., tubers, peltate leaves). However, the subgenus Tildenia appears in several clades within our trees. This is not surprising as Dahlstedt mentioned that certain morphological characters of some species of Tildenia are linked to other subgenera e.g., the shape of the fruit apex, links Tildenia either to Rhyncophorum, Sphaerocarpidium, Erasmia, and Panicularia on one hand or to Micropiper on the other hand. Hill (1907) placed the geophytic species in a different subsection Geophila within section Tildenia. However, he also included P. cotyledon based on a “similar tuber” and peltate leaves, as mentioned for Geophila. P. cotyledon is a species belonging to Panicularia (Dahlstedt, 1900) and this is also confirmed by our results. Indeed, according to our tree, peltate leaves evolved several times independently. The next clade (A) is formed by 4 species of the subgenus Panicularia (93 % BS/87 % BS/100% PP). One species that was placed by Dahlstedt in this subgenus as well, P. pernambucensis, comes out in a different clade near section Oxyrhynchum. Another species placed in Panicularia, P. sodiroi, comes out to- gether with P. verschaffeltii and P. marmorata. The morphological character used by Dahlstedt to distinguish the species of this subgenus is the paniculate inflorescence. This inflorescence is sharply delimited from the vegetative part and is deciduous as a whole after fruiting (except in P. pernambucensis). But apparently, the paniculate inflorescence has arisen at least three times: once in the subgenus Panicularia itself, once in the clade of P. pernambucensis (with several paniculate inflorescences, which are not clearly delimited from the vegetative parts of the plants), and once in the clade of P. sodiroi (with several paniculate inflorescences, which are clearly delimited from the vegetative parts of the plants). In his revision, Dahlstedt already recognized the fact that transitional stages to this inflorescence type also occur in other subgenera, namely Sphaerocarpidium and Micropiper (Dahlstedt, 1900). Clade A, representing Panicularia (excluding P. pernambucensis and P. sodiroi) receives high support in all our analyses. P. fraseri and P. cotyledon also form a clade (100% BS/100% PP), which is supported by the thickened scape and which is not comparable with the tubers of the geophilous species. Additionally the basal leaf rosette can be stated as synapomorphy, too. Phylogeny of Peperomia Plant Biology 8 (2006) Fig. 2 Phylogram based on the Bayesian analysis with posterior probabilities only indicated as asterisk if support is lower than 95% PP along the respective branch. Branches with PPs lower than 50% are collapsed. Of the 9 subgenera recognized by Dahlstedt, the analysis supports only the monophyly of the subgenus Micropiper (clade B, 98 % BS/100% PP). A very clear morphological synapomorphy for this subgenus is the so-called pseudocupula at the basis of the drupe (Dahlstedt, 1900). The nature of this structure has been a subject of controversy. Dahlstedt recorded the pseudocupula as a real structure, covered by a sticky secretion, which can be removed easily and is probably necessary for dispersal. Burger (1971) suggested that the structure might be formed as a result of differential drying of the lower part of the fruit surface and possibly of little taxonomic value. Our own observations in the field and in living collections of botanical gardens, suggest that the pseudocupula is not a drying artefact as stated by Burger (1971). The fruits in all the other sub- genera are sticky as well, but the secretion never covers the fruit surface. The only representative of the subgenus Pleurocarpidium sampled, P. emarginella, gains significant support as sister to Micropiper, (98 % BS/99 % BS/100% PP). From morphological point of view the position remains unclear. P. emarginella is a small prostrate epiphyte with thin, membranous, alternating leaves, whereas the subgenus Micropiper generally has succulent or coriaceous, opposite leaves. Henschen (1873) mentioned P. emarginella as a member of the subgenus Micropiper. Dahlstedt (1900) on the contrary believed that the subgenus Pleurocarpidium is a clade on its own because of the presence of a fruit pedicel, but he supposed that the group was related 99 100 Plant Biology 8 (2006) to Tildenia in which some species have pedicellated fruits as well. The relationships, based on the different matrices (parsimony method), between the clades A, B, and the remaining clades reveal uncertainty: I) The analysis without indels shows A branching first (100% BS), B secondly (100% BS), followed by the remaining clades (97 % BS/95 % BS); II). However, when indels are included, clade A and B show a sister group relationship (69 % BS) and are sister to the remaining clades (100% BS) (tree not shown). For the remaining clades the relationships obtained from the different matrices were identical, differing only by support. The Bayesian analysis shows the same branching patterns as the MP tree I, but the support for the sister group relationship of clade A and B is low (79 % PP). Clade C, a clade characterized by peltate leaves and a short irregular stem with short internodes is highly supported (99 % BS/100% PP). Members of this clade (C) were placed into the subgenus Tildenia by Dahlstedt and cause the paraphyly of the subgenus (clade E also contain members of Dahlstedts subgenus Tildenia, as P. lanceolatopeltata in clade G). Since peltate leaves are the main character for the subgenus Tildenia, the investigated species are related to numerous clades. The next clade (D) comprises the annual species, P. pellucida. The support of this clade remains unquestionable (100% BS/ 100% PP), but a clear morphological synapomorphy is missing. Based on the wrinkled fruit wall, P. pellucida has been placed together with the annual species P. exigua in its own subgenus Ogmocarpidium (Dahlstedt, 1900). He mentioned that the subgenus Acrocarpidium, which is not included in our analysis due to lack of material, is also characterized by the annual life form. Clade E is not supported (64 % BS/52 % BS/92 % PP). In addition, the relationships between this clade and the two clades described above, remain questionable as the support for these branches is low. It might be that the two “Tildenia” clades (clades C and E) together form one group. This is also supported by the presence of stems with short internodes. The three sections of the subgenus Rhyncophorum (clades F and G) are each monophyletic. The sections Leptorhynchum (96 % BS/88 % BS/100% PP) and Malacorhynchum (100% BS/ 100% PP) constitute a monophyletic clade (F) together with P. fagerlindii (100% BS/100 PP). Besides the species described by Dahlstedt in the third section Oxyrhynchum (100% BS/100 PP), we can assign two other species in clade (G) to this section as well on the basis of their stout habit and their fleshy, succulent, oblong to lanceolate leaves. Additionally, P. lanceolatopeltata has peltate leaves and therefore was placed in the subgenus Tildenia by Dahlstedt, but we showed previously that the character has a multiple origin. Yuncker (1953, 1974) placed P. lanceolatopeltata in Sphaerocarpidium. The inflorescences in this group are very different from each other, varying from solitary spadices (P. lanceolatopeltata, P. magnoliifolia, and P. clusiifolia) to paniculate inflorescences (P. pernambucensis) which are not as clearly delimited from the vegetative parts as the inflorescence in the subgenus Panicularia. S. Wanke et al. Subgenus Sphaerocarpidium and its sections are paraphyletic because subgenus Erasmia represented by P. lancifolia is placed within it. Sphaerocarpidium including Erasmia (clade H) reveals high support (92 % BS/98 % PP). We propose to include Erasmia into Sphaerocarpidium, which has already been discussed by Dahlstedt (1900). He assigned species to Sphaerocarpidium on the basis of the globose to ovoid fruit shape and the oblique apex. The delimitations of the sections within the subgenus were made on the basis of the leaf position. The sections themselves are once more divided in subsections. The clade P. galioides, P. inaequalifolia (section Verticillatae, subsection Leptophyllae) reveals high support (98 %/99 % BS/100% PP) and is easily recognized by the small, very succulent, verticillate leaves. All other species of the subgenus Sphaerocarpidium have alternate leaves, including P. lancifolia (Erasmia). P. blanda also causes the paraphyly of Alternifoliae which is supported by the opposite leaves as present in Verticillatae. The clade Alternifoliae s.l. (including Erasmia and P. blanda) receives maximum support (100% BS/100% PP). The phylogeny of the Hawaiian and South-Pacific species of the subgenus Sphaerocarpidium has been investigated by Bradley (2002). Yuncker (1933) distinguishes about 20 species from the Hawaiian Islands in his new subgenus Hawaiiana, based on a slightly different fruit shape and fruit apex compared to those of Sphaerocarpidium. This particular problem has been addressed in Bradley (2002), where she showed phylogenetic trees based on nuclear and chloroplast sequence data. In all her analyses the species of the two groups were mixed. We propose to include Hawaiiana again in Sphaerocarpidium. Conclusion – Prospects The importance of some morphological characters to subdivide Peperomia has been overestimated. It has been shown that these characters, e.g., peltate leaves, have evolved several times. In the present molecular study different groupings are proposed. Morphological studies, especially on ultrastructure, are necessary to improve the present groupings and to adjust the actual taxonomic knowledge. In addition to this study the sampling has to be enlarged especially to test the monophyly of the underrepresented groups Ogmocarpidium, Erasmia, Acrocarpidium, and Pleurocarpidium. In addition to this marker it could be useful to add ITS sequence data to resolve unsupported nodes and to improve the phylogenetic relationships found. Acknowledgements Financial support for this study came from the Department of Biology, Ghent University (travel grant to M. S.), the Flemish Fund for Scientific Research (FWO-Vlaanderen) (travel grant to G. M.), BOF-Ghent University (to L. V.), and the Friends of the Botanical Garden, Gent. The authors would like to thank Dietmar Quandt, TU Dresden for assistance in any kind of problems and Khidir W. Hilu, Virginia Tech, Kai Müller, Bonn for comments on the manuscript. We gratefully acknowledge material received from the Botanical Gardens Bonn and Berlin. Many thanks to Robert Maijer (The Netherlands) for sending us material from P. andina var. pseudoperuviana and P. dolabella. Phylogeny of Peperomia References Airy-Shaw, H. (1966) Diagnoses of new families, new names etc. for the seventh edition of Willis’ Dictionary. Kew Bulletin 18, 249 – 273. Airy-Shaw, H. (1987) Willis’ Dictionary: A Dictionary of the Flowering Plants and Ferns (8th edition). Cambridge: Cambridge University Press. APG1 (1998) An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85, 531 – 553. APG2 (2003) An update of the Angiosperm Phylogey Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141, 399 – 436. Azuma, H., Thien, L. B., and Kawano, S. (1999) Molecular phylogeny of Magnolia (Magnoliaceae) inferred from cDNA sequences and evolutionary divergence of floral scents. Journal of Plant Research 112, 291 – 306. Bornstein, A. J. (1991) The Piperaceae in the southeastern United States. Journal Arnold Arboretum (Suppl. 1), 349 – 366. Borsch, T., Hilu, K. W., Quandt, D., Wilde, V., Neinhuis, C., and Barthlott W. (2003) Noncoding plastid trnT-trnF sequences reveal a well resolved phylogeny of Basal Angiosperms. Journal of Evolutionary Biology 16, 558 – 576. Borsch, T., Löhne, C., Müller, K., Hilu, KW., Wanke, S., Worberg, A., Barthlott, W., Neinhuis, C., and Quandt, D. (2005) Towards understanding basal angiosperm diversification: recent insights using rapidly evolving genomic regions. Nora Acta Leopoldina 342, 85 – 110. Bradley, U. (2002) Biogeography and speciation of the genus Peperomia Ruiz and Pavón in Eastern Polynesia. PhD thesis, University of Dublin, Trinity College. Brummit, R. (1992) Vascular Plants, Families and Genera. Oxford: Clarendon Press. Burger, W. (1971) Flora Costaricensis (family 41 Piperaceae). Fieldiana Botany 35, 5 – 79. Burger, W. (1977) The Piperales and the monocots – alternate hypothesis for the origin of monocotyledonous flowers. The Botanical Review 43, 345 – 393. Dahlstedt, H. (1900) Studien über süd- und central-amerikanische Peperomien mit besonderer Berücksichtigung der brasilianischen Sippen. Stockholm: Kongliga Svenska Vetenskaps – Akademiens handlingar 33. de Candolle, C. (1923) Piperacearum clavis analytica. Candollea 1, 286 – 415. Henschen, S. E. (1873) Études sur le Genre Peperomia Comprenant les Espéces de Caldas, Brésil, Acta Universitatis Upsaliensis. Nova acta Regiae Societatis Scientiarum Upsaliensis. Heywood, V. (1978) Flowering Plants of the World. New York: Mayflower Books, Inc. Hill, A. W. (1907) A revision of the geophilous species of Peperomia with some additional notes on their morphology and seedling structure. Annals of Botany 21, 141 – 160. Hilu, K. W., Borsch, T., Müller, K., Soltis, D. E., Soltis, P. S., Savolainen, V., Chase, M. W., Powell, M. P., Alice, L. A., Evans, R., Sauquet, H., Neinhuis, C., Slotta, T. A. B., Rohwer, J. G., Campbell, C. S., and Chatrou, L. W. (2003) Angiosperm phylogeny based on matK sequence information. American Journal of Botany 90, 1758 – 1776. Hilu, K. W. and Liang, H. (1997) The matK gene: sequence variation and application in plant systematics. American Journal of Botany 87, 830 – 839. Huelsenbeck, J. P., Larget, B., Miller, R. E., and Ronquist, F. (2002) Potential applications and pitfalls of Bayesian inference of phylogeny. Systematic Biology 51, 673 – 688. Huelsenbeck, J. P., Ronquist, F., Nielsen, R., and Bollback, J. P. (2001) Bayesian inference of phylogeny and its impact on evolutionary biology. Science 294, 2310 – 2314. Plant Biology 8 (2006) Jaramillo, M. A., Manos, P. S., and Zimmer, E. A. (2004) Phylogenetic relationships of the perianthless Piperales: reconstructing the evolution of floral development. International Journal of Plant Sciences 165, 403 – 416. Kelchner, S. A. (2002) Group II introns as phylogenetic tool: structure, function, and evolutionary constraints. American Journal of Botany 89, 1651 – 1669. Liang, H. and Hilu, K. W. (1996). Application of the matK gene sequences to grass systematics. Canadian Journal of Botany 74, 125 – 134. Löhne, C. and Borsch, T. (2005) Molecular evolution and phylogenetic utility of the petD group II intron: a case study in basal angiosperms. Molecular Biology and Evolution 22, 317 – 332. Mathew, P. J. and Mathew, P. M. (2001) Pollen morphology of some members of Pipereaceae and its bearing on the systematics and phylogeny of the family. Rheedea 11, 65 – 78. Miquel, F. A. W. (1843) Systema Piperacearum (Kramers, H. A., ed.), Rotterdam, pp. 64 – 199. Müller, K. (2004) PRAP – computation of Bremer support for large data sets. Molecular Phylogenetics and Evolution 31, 780 – 782. Müller, K. (2005) SeqState – primer design and sequence statistics for phylogenetic DNA data sets. Applied Bioinformatics 4, 65 – 69. Müller, K. and Borsch, T. (2005) Phylogenetics of Utricularia (Lentibulariceae) and molecular evolution of the trnK intron in a lineage with high substitutional rates. Plant Systematics and Evolution 250, 39 – 67. Müller, K., Borsch, T., Legendre, L., Porembski, S., Theisen, I., and Barthlott, W. (2004) Evolution of Carnivory in Lentibulariaceae and the Lamiales. Plant Biology 6, 477 – 490. Müller, J. and Müller, K. (2004) TreeGraph: automated drawing of complex tree figures using an extensible tree description format. Molecular Ecology Notes 4, 786 – 788. Müller, J., Quandt, D., and Neinhuis, C. (2005). PhyDe – Phylogenetic Data Editor. Programm distributed by the author. Neinhuis, C., Wanke, S., Hilu, K. W., Müller, K., and Borsch, T. (2004) Phylogeny of Aristolochiaceae based on parsimony, likelihood, and Bayesian analyses of trnL-trnF sequences. Plant Systematics and Evolution 250, 7 – 26. Neuhaus, H. and Link, G. (1987) The chloroplast tRNA Lys (UUU) gene from mustard (Sinapis alba) contains a class II intron potentially coding for a maturase related polypeptide. Current Genetics 11, 251 – 257. Nixon, K. C. (1999) The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15, 407 – 414. Novak, F. (1954) System angiosperm. Preslia 26, 337 – 364. Olmstead, R. G. and Palmer, J. (1994) Chloroplast DNA systematics: a review of methods and data analysis. American Journal of Botany 81, 1205 – 1224. Posada, D. and Crandall, A. (1998) Modeltest: testing the model of DNA substitution. Bioinformatics 14, 817 – 818. Rodriguez, F. J., Oliver, J. L., Marín, A., and Medina, J. R. (1990) The generic stochastic model of nucleotide substitution. Journal of Theoretical Biology 142, 485 – 501. Ronquist, F. and Huelsenbeck, J. P. (2003) MRBAYES: Bayesian phylogenetic inference under mixed models. Bioinformatics 19, 1572 – 1574. Ruiz, H. and Pavón, J. (1794) Florae Peruvianae et Chilensis Prodromus. Madrid. Sang, T., Crawford D. J., and Stuessy, T. F. (1997) Chloroplast DNA phylogeny, reticulate evolution, and biogeography of Paeonia (Paeoniaceae). American Journal of Botany 84, 1120 – 1136. Simmons, M. P. and Ochoterena, H. (2000) Gaps as characters in sequence-based phylogenetic analyses. Systematic Biology 49, 369 – 381. Skottsberg, C. (1947) The genus Peperomia in Chile. Acta Horti Gotoburgensis 17, 1 – 47. 101 102 Plant Biology 8 (2006) Smith, A. C. (1981) Flora Vitiensis. Nova 2, 75 – 97. Stehlé, H. (1940) Flore Descriptive des Antilles Françaises 2 – Les Pipérales. Fort-de-France: Imprimerie du Gouvernement. Swofford, D. L. (1998) PAUP*. Phylogenetic Analysis Using Parsimony (*and other Methods). Sunderland: Sinauer Associates. Tebbs, M. C. (1993) Piperaceae. In The Families and Genera of Vascular Plants (Kubitzki, K., ed.), Vol. 2. Flowering Plants – Dicotyledons – Magnoliid, Hamamelid and Caryophyllid Families. Berlin: Springer-Verlag. Thorne, R. F. (1992). Classification and geography of the flowering plants. Botanical Review 58, 225 – 348. Trelease, W. (1930) The geography of American peppers. Proceedings of the American Philosophical Society 69, 309 – 327. Tucker, S. C., Douglas, A. W., and Liang, H. X. (1993) Utility of ontogenetic and conventional characters in determining phylogenetic relationships of Saururaceae and Piperaceae (Piperales). Systematic Botany 18, 614 – 641. Walker, J. (1976 a) Comparative pollen morphology and phylogeny of the Ranalean complex. In Origin and Early Evolution of Angiosperms (Beck, C., ed.), New York: Columbia University Press, pp. 241 – 249. Walker, J. (1976 b) Evolutionary significance of the exine in the pollen of primitive angiosperms. In The Evolutionary Significance of the Exine (Fergusen, I. and Muller, J. eds.), London: Academic Press, pp. 251 – 292. Yuncker, T. G. (1933) Revision of the Hawaiian Species of Peperomia. Bulletin of the Bernice P. Bishop Museum, Honolulu, Hawaii, 112, 3 – 131, illust. Yuncker, T. G. (1953) The Piperaceae of Argentina, Bolivia and Chile. Lilloa 27, 97 – 303. Yuncker, T. G. (1974) The Piperaceae of Brasil III. Peperomia: taxa of uncertain status. Hoehnea 4, 71 – 413. S. Wanke et al. S. Wanke Institut für Botanik Plant Phylogenetics and Phylogenomics Group Technische Universität Dresden 01062 Dresden E-mail: stefan.wanke@tu-dresden.de Editor: F. Salamini