The JapaneseSocietyJapanese Society for Plant Systematics
ISSN1346-7S65 Acta Phytotax, Geobot. 64 (3):ll3-126 (2013)
PhylogeneticAnalysis of Nepenthaceae, Based on Internal Transcribed
Spacer Nuclear Ribosomal DNA Sequences
FiRMAN ALAMsyAHi'2" AND MoToMi IToi
JDepartment oj'General Systems Studies. Uhiverisity 7blyo 153-8902, Jopan; 2Department of7btve, ofBiotogicai Science, FZiculty ofAdLithetnatics and Sciences, Ahmad Dahlan *[email protected] Uitivensi(B lanturan, ibgyakarta 55164, lhdonesia. (tiuthorfor correspondeneoj
Nepenthaceae, a monotypic family of carnivorous pitcher p]ants comprising Nbpenthes, is widely dis- tributed in Southeast Asia. Tb determine the phytogeography of lVlepenthes in Southeast Asia, and te trace the evolutionary trends of taxonomically important characteristics (i.e., peristomes) of the genus, we analyzed S7 internal transcribed spacer (ITS) nuclear ribosomal DNA (nrDNA) sequences of56 spe- cies of ?Vbpenthes and 1 ITS sequence each ofDionaea muscipula ancl Ancistroeladus robertsoniorerm. To clarify the phylogenetic relationships oflVlepenthes, we examined fbur different methods ofphyloge- netic tree reconstruction. The resulting tree topologies were mostly consistent with one another except for the basal polytomies, Seven monophyletic subclades could be recognized, Similarities and differ- ences in terms of the positions of taxa between the present study and previous studies were ebserved. Judgingfromthephylogenetictreesanddistributionareaofeachspecies,Borneoappearstobeasecond-
ary center of diversification fbr Nbpenthes and species ofIVkpenthes may have then radiated within the Sunda Shelfof Southeast Asia. The three character states ofthe peristomes from the upper pitchers were relatively well correlated with the grouping of the species of Nepenthes within seven subclades and showed the limitations ofthe Danser (1928) system for ?Vepenthes,
Key words: ITS,IVepenthes, phylegenetic relationship, phytogeography, peristome
Nepenthaceae, a monotypic family of carniv- Philippines (21 species, 20 endemic). Although orous pitcher plants comprising ?Vbpenthes, is most extant species oflVepenthes are distributed widely distributed in the Asian tropics, mainly in to the west of the Wallace line, some species, Southeast Asia and the Sunda Shelfregion. The such as IVI ampullaria, IVI gracilis, AL mirabilis, vast majority of species grow in moist regions and AC tentaculata, eccur in both Asia and Wal- throughout the Old World tropics, as far west as lacea (McPherson 2009). The distribution ofthe macingascariensis Madagascar (A4 and A[ maso- genus has been attributed to biogeographic fac- alensis), the Seychelles (NlperviUei), and Sri tors occurring both recently and in the past, in- Lanka (Nl distillatoria) to India (IV] khasiana) in cluding the connectien ofthe Sunda Shelfislands the north, Australia (N tenax and N Towanae) in caused by sea level drop and the isolation of the the south; and New Caledonia (?Vl vieiliardii) in islands caused by increased sea levels, changes in the east (Clarke 2007, Krutzsch 1988, McPherson global climate, and the ability of the species of 2009), IVbpenthes to disperse and colonize new habitats The genus comprises 120 species in Southeast (Clarke 2006, Danser 1928, McPherson 2009, Asia, especially Sumatra (37 species, 29 endem- Meimberg et al. 2001). ic), Borneo (36 species, 29 endemic), and the Meimberg et aL (2001) performeda molecular
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
114 Acta Phytetax. Geobot. Vbl.64
to evaluate phylogenetic analysis of ?Vepenthes based on cosa and N alata havebeen studied DNA sequences of the plastid trnK intron. The the potential value of ITS for phylogenetic recon- trnK intron is more variable than rbcL (Johnsonstruction, This study implies that the ITS regions & Soltis 1995, van den Berg et al. 2005) and is of these species have many variable characteris- widely used for reconstruction of lower taxonom- tics that are potentially infbrmative fbr resolving ic levels and has therefore been the locus of choice the phylogeny of Arapenthes as1ejandro et ai, in molecular phylogenetic analyses of Nepentha- 2008). Ribosomal DNA genes, however, are ceae (Meimberg & Heubl 2006). The results of present in high copy numbers and may therefore their study assumed that colonization of South- be subjected to directional concerted evolution east Asia by AJepenthes was initiated from India, (Wendel et aL 1995a) or intergenomic introgres- since the Indian endemic, IV khasiana, was a sis- sion (Wendel et al. 1995b), ter taxon to all Asian taxa of Nepenthes. A com- The objectives ofthis study were to clarify the parative analysis between the Nepenthes trnK in- phylogenetic relationships ofNepenthes based on
tron and its translocated copy, however, has dem- ITS nucleotide sequences fbr 1) determining the onstratedatopological incengruence. Meimberg phytogeography of Nepenthes in Southeast Asia et al. (2006) and Meimberg & Heubl (2006) and to re-evaluate the scenario suggested by 2) trace the evolu- pointed out the possibility ofintrogression or lin- Meimberg et al. (2001), and eage sorting as the reason fbr the topological in- tionary trends ofperistomes in the genus. Peri- cQngruence ofthe phylegenetic trees, Meimberg stomes are stiff structures comprising an inward- & Heubl (2006) then introduced PTRI (peptidely curved rim surrounding the pitcher opening transferase 1), a nuclear low copy gene as a phy- (McPherson 2009). Peristomes have been used logenetic marker, but the resolution of the phylo- in taxonomic studies (Danser 1928) to distin- & Jebb genetic tree was low and some taxa appeared in guish between related species (Cheek different positions from the previous findings in 2009, Lee et aL 2006, McPherson 2009, Robinson the tree topology. Meimberg et al. (2001) provid- et al, 2009) and to determine new species (Cheek ed two possible interpretations concerning the & Jebb 2009, Lee at al, 2006, Robinson et aL
origin ofNepenthes; evolution in the northern [Ib- 2009).
thys, or a Gondwanan origin at a time when the
Indian plate was separated from Madagascar, Materialsand Methods From the molecular analysis based on matK, they suggested that colonization of Southeast Asia Plantmaterials was from an ancient Indian stock and subse- We analyzed 57 samples from 56 species of quently a new secondary center of diversity de- Nepentkes representing all geographical areas veloped in the Malay Archipelago. Re-eyalua- (Table 1), Ofthe included species, seven (AC cha- tion ofthe origin and diversification of Nbpenthes niana, N, lingulata, Any mindnnaoensis, N naga, using another gene is needed because oftopologi- N platychila, N thai, and N vogelii) were found cal incongruence and low resolution ofthe phylo- within the last 10 years (Akhriadi et al. 2009, genetic trees obtained from previous studies. Cheek & Jebb 2009, Lee et aL 2006, McPherson In our study, we used nucleotide sequences of 2009) and have never been used in phylogenetic the internal transcribed spacer (ITS) to resolve studies. TWo samples from distantly distributed phylogenetic relationships within 7Vepenthes. Nl mirabilis were also examined. The first UVI The characteristics of the ITS region, with its mirabitisl) was from Bengkulu, Sumatra, and the small size, highly conserved fianking regions, second (?Vl mirabilis2) was from West Kaliman- and fast evolutionary rate, have made this nuclear tan, Borneo. The seven species and the two dis- ribosornal DNA (nrDNA) sequence a valuable tantly distributed samples of IVI mirahilis from marker fbr phylogenetic analysis (Baldwin et al, the Malay Archipelago were expected to contrib- 1995). In addition, ITS sequences frem N ventri- ute to a better understanding of the phytogeogra-
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
October2013 Al.AMsyAH&ITo-PhylogeneticAnalysisefNepenthaceae 115
phy ofNbpenthes in Southeast Asia. Ancistrocta- sarnple was used for phylogenetic analysis; all se- dus robertsoniorum (Ancistrocladaceae, Gen- quences were deposited in DDBJIEMBLIGen- bank: GQ443551) and Dionaea muscipula (Dro- bank(Table 1), seraceae, Genbank: AB675913) were used as out-
groups, because these two fam{lies have been Phytogenetic analysis recognized as sister groups to Nepenthaceae on DNA sequences obtained from the ITS region the of macromolecular characteristics using basis were aligned with ClustalX (Larkin et al. 2007). nuclear 18S rDNA and pTastid rbcL, acpB, and Phylogenetic analysis involving the maximum matK sequences (Albert et al. 1992, Cuenoud et parsimeny (MP) method was performed using 2002, al, Hilu et aL 2003), the PAUP (Phylogenetic Analysis using Parsimo- ny) program, version 4.0blO (Swofford 2002). Amplijication and sequencing Data were analyzed by the heuristic search meth- Total DNA was extracted from silica gel- .od with the tree bisection-reconnection (TBR) dried leaf samples with a Qiagen DNeasy Mini branch swapping and MulTrees options on and Plant Kit (Qiagen, Valencia, CA, USA) fo11owing stepwise addition with simple addition sequences the manufacturer's protocol. Amplification ofthe using one reference taxon (AC thai), All of the ITS region firom Ancistrocladaceae was per- most parsimonious trees (MPTs) were saved. All formed using a set of primers, AITSI characters were equally weighted and unordered (5'-AGAAGTCCACTGAACCTTATC-3;) and (Fitch 1971). Gaps were treated as missing data. AITS4(5'-CGCTTCTCCAGACTACAATTC-3'), Tb evaluate the internal support of clades, boot-
which are angioseperm specific and do not co- strap analysis (Felsenstein 1985) was conducted amplify fungal DNA (Meimberg et aL 2010). The using 1000 replicates in a heuristic search with amplification reaction for the ITS region included the TBR branch swapping and MulTrees options Ex:fttq buffer and Ex-Taq DNA polymerase (Ta- off1 The number of steps, consistency indices, kara Bio, Shiga, Japan). The polymerase chain and retention indices (Farris 1989) were calculat- reaction (PCR) protocol consisted of an initial 90 ed with one of the MPTs in each anaiysis using s predenaturation at 960C; 30 cycles of 20 s at the TREE SCORES cornmand in PAUP'. For 960C (denaturation), 40 s at 570C (annealing), and comparison, we also performed phylogenetic 40 s at 72aC (extension); and a final 7 min exten- analysis using three different methods, sien at 720C. For seven species samples (Nl am- (1) Bayesian analysis with MrBayes version pullaria, N. hinguta, ?Vl rowanae, IVI donseri, N. 3.1.2 (Ronquist & Huelsenbeck 2003) by using neoguineensis, N pcrpuana, and N tentaculatdi, GTR+I+G model selected from the Medeltest anneal{ng temperature was changed to 58.50C. (Posada 2008) as the eptimum model fbr se- The PCR products were c]eaned using the quence evolution based on AIC criterion, Four Wizard SV Gel and PCR CIean Up System (Pro- chains were run for 1,OOO,OOO generations and megaj and were used fbr autocycle sequencing re- were sampled every 100 generations to yield a actions fbllowing the manufacturer's (Beckman posterior prebability distribution of 10,OOO trees. Coulter) instructions, Autocycle sequencing The first 2500 trees were discarded as burn in, To
products were cleaned by ethanol precipitation, check whether the MCMC chain was long enough
Both fbrward and reverse sequences were ana- to reach convergence, the trace fiies resulting
lyzed with a CEQ8000 automated sequencer from the Bayesian analysis were opened in the (Beckman Coulter), using the same primers as for Tracer program (vl.5) (Rambaut & Drummond PCR. A set ofinternal primers, AITS2R (5;-TGC-2007), and their effective sample size (ESS) sta- GTTCAAAGACTCGATGG-3') and AITS3F tistics were calculated fbr values higher than 100. (5'-GAAGAACGTAGCGAAATGCG-3'), was (2) The neighborjoining (NJ) method (Saitou designed to achieve better analysis of the ITS re- & Nei 1987) with MEGA version 5.05 ('Ilamura et rlbmura gion. The sequence ofthe ITS region from each aL 2011) by using the 3-parameter model
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
116 Acta Phytotax. Geobot. Vol. 64
TABLE 1. PLant materials examined in this study. The materials werc collected frorn the Indonesian Carnivorous PIant Soeiety (Komunitas 7innaman Karnivora IndonesiaKT7CJ), Bogor Botanical Garden (BBG, Indonesia), and Kyoto Botanical Gar- dens (KBG, Japan). All specimens were deposited in TI (*FA: Firman Alamsyah).
Species Source Voucher No. of Genbank Specimens*Voucherliving stocks accession no, of ITS
Nqpenthes acinata Tamin & M. Hotta ex Schlauer KTKIKTKIKTKIBBGKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTK]KTKIBBGKTKIKTKtKTKIKTKIKTKIKBGKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIBBGKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKTKIKBGKTKIKTKJKTKIFA-11eOOI Bllll09 Ol AB675864
IVepenthes alata Blanco FA-110002 C141!0901 AB675865 IVqpenthes alba Rid1. FA-110003 E05021001 AB675866
IVepenthes ampuUaria Jack FA-110049 KRB27100901 AB675914
IVopenthes betlii K. Kondo FA-110004 Allll09 01 AB675868
IVepenthes burbicigeae Hook. f. ex Burb FA-110005 B14110901 AB675869
IVopenthes burkei Mast. M-110006 EOS021002 AB675870
Nepenthes campanutata Sh. Kurata FA-110007 B14110902 AB675871 Nepenthes ehaniana C. Clarke et al. FA-1100e8 Bllll0902 AB675872
Nlepenthes clipeata Danser FA.LIOO09 D07110901 AB675873 Nbpenthes cqpelandii Merr, ex Maefarlane FA-110010 E05021003 AB675874
Nepenthes clanseri Jebb & Cheek FA-110050 A14110901 AB675915
Nepenthes ctenstflora Danser FA-110057 C14110902 AB675875
Nepenthes diatas Jebb & Cheek FA-110011 D071109 02 AB675876
Nepenthes distitlatoria L. FA-110012 E05021004 AB675877 AB675878 Nopenthes aphippiata Danser FA-110013 A14110902 Nepenthes,faiialiana J.H, Adam & Wilcock FA-110014 E05021005 AB675879
FA-110015 Nlepenthes fasca Danscr A141109 03 AB67588e Allll0902 AB675881 Nepenthes glabrata J.R. [[Urnbull & A.T, Middleton FA-110016 IVepenthes gracitis Korth. FA-110017 KRB27100902 AB675882 ?Vepenthes hirsuta Hook. f EA-110051 D07110903 AB675916 ?Vepenthes khasiana Hook. f FA-110018 E05021006 AB675883 Nepenthes lingulata Chi. C. Lee, Hernawati & Akhriadi FA-110019 Allll0903 AB675884
Nepenthes longijblia Nerz & Wistuba FA-110020 E05021007 AB675885 Nepenthes macrevuigaris J.R. 1ftirnbull & A.T, Middleton FA-110021 B14110903 AB67S886
Nepenthes madagascariensis Poir. FA-110058 KBG02-0521 AB769064 IVlapenthes merriUiana Macfarlane FA-110022 Allll0904 AB675887
Nepenthes mindanaoensis Sh, Kurata FA-110023 A141109 04 AB67588g Nepenthes mirabilis CLour,) Druce (Bengkulu, Sumatra) FA-1]O024 E05021008 AB675889 E05021009 AB675890 kpenthes mirabitis (Lour.) Druce (West Kalimantan) FA-110025 IVopen thes naga Akhriadi et al. FA-110026 D07110904 AB675891
Nepenthes neoguineensis Macfarlane FA-110052 A14110905 AB675917 Naj)enthes ovata Nerz & Wistuba FA-110027 Bl"10903 AB675892 Net)enthespopuanaDanser FA-1]O053 B141109 04 AB675918 NepenthespervilleiBlume FA-110028 E05021010 AB675893 ?Vepenthes platychita Chi. C. Lee FA-110029 E050210]1 AB675894 IVepenthes rojah Hook, £ FA-110030 D07110905 AB675895 NepenthesreinwardtianaMiq. FA-110031 KRB27100903 AB675896
IVepenthes rhombicaulis Sh. Kurata FA-110032 B14il09 05 AB675897 IVepenthes rowanae F,M. Bailey FA-110054 A14110906 AB675919
fVepenthes sanguinea Lindl. FA-l]O033 BL4110906 AB675898
IVepenthes smilesii Hemsl. FA-110034 CI4110903 AB67S899
Nepenthes spathulata Danser M-110035 D07110906 AB675900
Nepenthes spectabiiis Danser FA-110036 A14110907 AB675901
Nopenthes sp, Misool FA-110037 EOS021012 AB675902
Nepenthes stenoplo,lla Mast. FA-110038 A141109 08 AB675903 lventhes sumatrana (Miq.) Beck FA.110039 A14110909 AB675904 IVopenthes talangensis Nerz & Wistuba FA-IIO040 D07110907 AB675905 IVepenthes tentaculata Hook. £ FA-110055 C141109 04 AB675920
Nepenthes thai Cheek FA-110041 E05021013 AB675906 IVepenthes tobaica Danser FA-110042 B11110904 AB675907
?Vepenthes truncata Macfarlane FA-110e43 E05021014 AB675908
IVepenthes veitchii Hook, f FA-110044 A14110910 AB675909
Nepenthes ventricosa Blanco FA-110045 Allll0905 AB675910
AJqpenthes vieiUardii Hook. i FA.110059 KBG02-0522 AB76906S
IVepenthes viUvsa Hook. f, FA-110046 B14110907 AB675911
Nqpenthes vogelii Schuit. & de Vbgel FA-110047 Bllll0905 AB675912
Dionaea musciputa J. Ellis ex L. FA-110048 E0502101S AB675913
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
October 2013 ALAptrtsyAH&ITo-PhylogeneticAnalysisofNepentha ¢ eae 117
([famura 1992), Bootstrap values were calculated We obtained 57 new ITS nrDNA sequences of with 1000 replicates, 56 species of Aibpenthes and 1 ITS sequence of (3) The maximum likelihood (ML) method Dionaea muscipttta. The alignment of59 entire with the Treefinder program (Jobb et aL 2004) by ITS sequences (57 ingroup sequences plus 2 out- using GTRtG model, which was suggested based groups) provided an 856-bp-long rnatrix. Se- "propose on AIC criterion by the modeP' analysis quence length variations result{ng from inser- implemented in the Treefinder program. Boot- tions and deletions were fbund among the species
strap values were calculated with 1000 replicates. ofNlepenthes. The aligned ITSs contained 341 (39%) constant characters, 239 (29%) parsimony- Reconstruetion ofcharacter states uninfbrmative variable characters, and 276 (32%) Te study the phytogeography ofAJepenthes in parsimony-jnformative characters. The analysis Southeast Asia, we mapped the character states resulted in 324 MPTs with a length of 1130 steps fbr distribution areas onto one ofthe MPTs by us- and had consistency (CI) and retention (RI) indi- ing the MacClade program (v4.06), with acceler- ces of O,665 and O.68Z respectively. The strict ated character transfbrmation (ACCTRAN) opti- consensus tree reconstructed by the parsimony mization (Maddison & Maddison 2003). In addi- method is shown in Fig. 1. The trees obtained tion, we also mapped the character states fbr the from the NJ, ML, and Bayesian analysis methods
peristome of upper pitchers to study the evolu- were mostly consistent with the tree obtained tionary trends ofmorphological characteristics of from the parsimony method, except forbasal spe- Napenthes. The morphometric data ofthe peri- cies, which appeared in difTerent positions (Fig. 1 stomes of the upper pitchers were taken from & Fig. 2). We compiled the bootstrap values of McPherson (2009). MP,NJ,ML,andposteriorprobabilitiesofBayes- Uppcr pitchers are distinguishable from lower ian anaiysis on the MP tree (Fig, 1). pitchers, They are produced from leaves along a Using Ancistrocladus robertsoniorum and climbing stem of older plants of Nepenthes, in Dionaea mttscipula as outgroups, we recognized which the tendril attaches to the back ofthe pitch- two basal branches and seven subclades (I=Vil) er on the side where the lid meets the rear of the in the phylogenetic tree (Fig. 1). In addition, we pitcher orifice, In contrast, the lower pitchers, recognized a basal group consisting ofseven spe- which are developed from a tendril that attaches cies whose branching pattern could not be re- to the front ofthe pitcher on the same side as the solved well. The seven subclades (I-VII) were orifice, are produced by young plants of Nepen- supported by all four phylegeny reconstruction thes (Di Giusto et al. 2008, Gaume & Di Giusto methods, except fbr subclade I, The members of 2009, McPherson 2009, Moran 1996), All but subclade I were grouped with some of the basal one ofthe species oflVepenthes examined in this unresolved species in the Bayesian tree (Fig, 2). study showed this pitcher dimorphism. Nepen- Correspending to the tree topology, the two basal thes campanulata produces only one type of taxa were Nl pervillei from Seychelles and N pitcher. We included data for the peristomes of madngascariensis from Madagascar. SubcladeI these pitchers, since they develop from leaves comprised two species from outlying areas (India farther up the stem of mature plants and show (IVI khasiana) and Sri Lanka (?Vl distillatoria)); a tendril attachment as in the typical upperpitch- species from New Guinea (IVI pqpuana), and a ers of 7Vepenthes, We classified peristomes as species (N. ampullaria) distributed across Penin- having three character states: narrow (<10 mrn), sular Malaysia, Sumatra, Borneo, and New Guin- intermediate (10-20 mm), and broad (>20 mm). ea. Subclade II consisted ofa species endemic to Sulawesi (Ail gtabratdi, a species from Sulawesi Results and Borneo UV tentaculata), and a species en- demic to Borneo (Nl hinsuta). Subclade III con- Molecularplzylogeay ofNepenthes tained species restricted to the Philippines and a
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
118 Acta Phytetax, Geobot, Vbl. 64
species endemic to Borneo (IVI campanulata) in sanguinea, IV atba, and N. thai), a species en- the basal pesition. Subclade IV also contained demic to Borneo (N macrovuigaris), and a spe- species restricted to the Philippines. Subclade V cies occuring in Borneo and Sumatra (IVI rein- contained only species from Borneo. Subclade wardtiantD in the basal position, Subclade VII
Vt comprised exclusively fbur species from Pen- comprised 13 species ofNepenthes from Sumatra
insular Malaysia and Indochina (IVI smilesii, Ail and one species (N. spathulatcD also distributed in
D. tnuscipuia A. robertsonjorarn
N. perv1'"ei (Sey} N. tnadagFesceriensis (Mad) N. neoguineensis (NG) N. dansefi (NG} N. i'owanae
FIG. 1. Strict consensus tree derived frem maximum parsimeny analysis of ITS sequences of ?Vepenthes and outgroup taxa, Statistical support for each branch is shown below each branch with successive values ofMPINJfMLIBayesian, The mi- nus signs (-) on the successive statistical values indicate different topologies or polytomies. Additional information en the distribution area ofa the species is given as an abbreviation next to the taxon name,
NII-Electronic Library Service The JapaneseSocietyJapanese Society forforPlantSystematics Plant Systematics
October2013 ALAMsyAH & ITo-Phylogenctic Analysis of Nepenthaeeae 119
A. robertsoniorLmi
- O.89 D. muscipula N. pervillei N. madagascariensi.c
N. neoguineensis O,88 N. distnvatotia O.84 O,89O.83 N. khasiana '-'O179[ N, papuana O.8 N. atnpullaria N. rowanae 1,OO-o N. mirabili$2 N. mirabilisf
N. sp. AMsool O.81 89,ggE'ptt98 N, grac"is N. viefiiardii N. danseri
r1.00 N. tentacuiata r N. hirsuta II / O,60-L N. glabrata O,S8 N, campanufata N, mindanaoensis 1.00 III 1.oe-r N, copelandii O,99- N, truncata O.9S N, alataN, beLtiiN, 1.00 99'- merrilliana IV O.75O N, ventticosa i-1.00L N. burkei N, villosa
1.00 -' N, rql'ahN, r･ 1,OO veitchii
1.eo r N, stenoph)tlia ! N, chaniana O.91vr N. ciipeata V N. o.9'9'O.94-O195[ ophipaiata N. faiialiana
N. vQgelii L l N. fuscaN. 1.00 1bO platychiia N. buthidgeae N. teinwardtiana O.84 N. macrovuigaris l N. thaiN. O.84 1 VI sanguinea O.96 N. albaN. O.97 O.99 smilesii N. tongifolia N. nagaN. 1.00L adnata I.OO N. thombicadiis N. taiangensis '' N. ovataN. O.9ljL'' 1.00 iingulata VII N. ctensifiora - O.94 O.98 N. diatas 1.001"N. spectabiiis
N. spathulata
1.0 - N. sumatrana 1.00i N. lobaica
FTG. 2. The 50% majority-rule eonsensus tree derived from the Bayesian analysis of ITS sequences for IVbpenthes and outgroup taxa. Numbers below branches are Bayesian posterier probabilities.
NII-Electronic Library Service The Japanese SocietySooiety for Plant SystematicsSystematios
120 Acta Phytotax、Geobot． Vol．64
口 D ．muscip ［’1a 口A ．robert30n ’orum 口 八t．ρe 門ガ”e ’ 口 N ．madiagascarjensis ■ 八t．η eo9 正ノ加ee ’lsis ■ 八t．dati3eri 口 N ．tovatanae ロ バ1．γ’e 班a 雇 歴 国 1V．9旧 cl’18 ■ A1． sp ． Misoo［
国 八t．mi ∫abi ’isず
圜 N 、’η irabiiis2 口 N ．distitiθtorja 口 ’》．khasianθ ー ■ A1．ρapuana 圜 〜．am ρu ”a 〃a ■ t＞．gfabrθta ・ 匡ヨN ．tet］taculatθ H ■ N 、h’rSl’ta ■ 八Loa8 ηρa ρ‘”ata ■ At． mindanao θnsis 田 囲 N 、cope ’a ρ（ゴ’∫ 国 N ．trt’ρc θ維 囲 N ．a ’ata 壓 N 、beitii 国 t》、me 厂汀〃’aノ）a 障 ■ N ．レentricosa ， N ． burkei 鬮 N ．レitiOSθ 謄 N ．raJa’1 醫 N ．ve ’t〔｝hii ■ 八t’stenophylia 匿 N ．c〃ρe ∂’a ■ N ．chaniana V ■ 〜．eρノ）iPP’∂ 紬 匿 八t．faizaiiStlθ 謄 配 船 ca 巳 A1．ρ’atγch ”a 騰 N ．加 め ’（lgeae 巳 N ．v 〔igeiii 圜 N ．reinl！vardtieρδ 口 Mmacrovulgaris ・ di N ． thai 刷 es N ． eiba 駐 ノ》．sangtiine θ 図 八1．smi ’esj’ Di5セrib 蘭 t 匿 A1．’ongiiofia unordered 匿 N 、’］ag θ Qu 匸：コ 匿 A「．θdnθta
■■INe 驂 N ．rho1η bicat，μs ■ ■ Su 謄 八1．dens ’fior召 ● 〜．diatθs ■■ B。 ■ N ．3pectabUis … … 圏 Ph 騒 〜 ．taiangensis 鬮 Ma 臨 N ．’ingutata ■ Su 匿 1＞．o レ∂ ta 巳 配 s ath ｛ノ’afa 翻 poI ρ ■ N ．Sblm θtraη a eq 圄 8N ． tobθ ice
FIG 3 Character stato reconstruction ，， efNepenthes tbrdistribution areas based on l of324 MPTs using MacClade ver ．4，06 with − ACCTRAN optimization ． The squares next to the taxon namc irユdicated subclades I VII．
一 NII-ElectronicN 工 工 Eleotronio Library Service The Japanese SocietySooiety for Plant SystematicsSystematios
− October2 13 ALAMsYAH ＆ ITo Phy且ogenetic Analysis ofNepenthaceae 121
aA
配 ． M 配 M 隠 欝 廻 蜥 配 配 侃 配 配 MM
配 驫 κ 配 飆 配 配 臨 配 廻、 紺 席 庶 MM 配 諜 配 配 噺 配 嚇 配 綿 M 配 紺 侃 讌 配 細 MMMM 翫 配 M 翻 付 付 庶 庶 〜 ． 侃 撫 厚 Poristo 配． M ［dere［ 〜、 匚 コ MMM ■ 欝 ■ 庶 漁 幽 庶 κ 瀞
FIG，4． Character st 段te reoollstruction ofNepenthes for peristomes based on l of324 MPTs using MacClade ver ．4．06 with ion with comparison to the ctassification sys 亡em or 〈iepenthe v by Danser 1928 which is indica亡ed ACCTRAN optimlza し ． （ ）， ’ − by the squares containing group names next to thc taxon namcs ． The ai rows indicatc repcatcd evolution ofnarrow ， inter and ． We recogllized 亡hree character states of the ：narrow ＜ 10mm interlnediate mcdiate ， broadperistomes peristomes （ ），
− （10 20mm ）， a 冂dbroad （＞ 20111m）． The character state foreach sp ∈cies was obtained from the descriptions by McPherson （2009）．
一 NII-ElectronicN 工 工 Eleotronio Library Service The JapaneseSocietyJapanese Society for Plant Systematics
122 Acta Phytotax. Geobot. Vol. 64
Java. lack the appendages typical of most species of
Seven species were indicated as basal polyto- AJbpenthes, In addition, N pervillei also has mies on the MP phylogenetic tree, including two some unusual characteristics not found in other nonendemic species (Nl mirabiiis and Arl gracilis) species oflVbpenthes, such as nontwining tendrils distributed across Borneo, Sumatra, Sulawesi, that only slightly emerge from the upper pitcher; Peninsular Malaysia and Indochina and the Phil- similar upper and lower pitchers; black, short, ippines, New Guinea, and northern Australia fbr ovoid, truncate seeds and obconic fruit (Meim- IVI mirabilis; a species from Misool Island located berg et aL 2001, McPherson 2009). The basal po- near the west coast ofNew Guinea (Nl sp. Misool), sition oflVl pervillei was consistent with reports two species from New Caledonia (Nl vieitlardii) from previous studies ofthe phylogeny ofNbpen- and Australia (IVI rowanae), and two species from thes using the trnK intron (Meimberg et at, 2001, New Guinea ov neeguineensis and N, dunseri), Meimberg & Heubl 2006, Meimberg et aL 2006). and Waigeo Island located near the northwest The second basal taxon in our study, ?V[ madogas- coast ofNew Guinea (IVI clanseri). cariensis, was also consistent with the trnK in- tron phylogeny of Nbpenthes (Meirnberg et al.
Character state reconstruction 2001, Meimberg et al. 2006), but in a subsequent The character state reconstructions ofNkipen- study ofthe phylogeny of Nepenthes using PTRI, thes fbr distribution areas and peristomes, based N madogascariensis formed a clade together on one of the 324 MPTs are presented in Fig. 3 with an Indian species, N khasiana, and a Su- and Fig. 4, respectively, Figure 3 shows the evo- lawesian species, N. tomoriana (Meimberg & lutionary trends of distribution areas of Aiepen- Heubl 2006).
thes. which exhibits the radiation of?Vepenthes on In our study, all the New Guinean species
some islands and the migrat{on of ?Vepenthes to were included with the basal species and in sub-
adj acent islands or to mainland of Southeast Asia. clade I with species from outlying areas, includ- Figure 4 shows the evolutionary trends of the ing Australia (AC rowanae), India (IVZ khasiana), peristomes of the upper pitchers and a compari- and Sri Lanka (?Vl distillatoria), and the widely son of the characteristics of the peristomes with distributed Ail ampullaria. The positiens ofthe the classification system of the genus by Danser New Guinean species IV papuana in the same (1928). subclade as species from India (N khasianaj and Sri Lanka (IVI distiilatoricD differ from previous Discussions studies of the trnK intron phylogeny of IVlepen- thes, where the Indian species (N khasiandi was Phylogenetic relationships in the genus Nepen- recognized as the third basal branch, The posi- thes tions of the widely distributed N graeilis and AJI In our study efITS DNA analysis (Fig. 1), we mirabilis as basal species, and IVI ampullaria in obtained phylogenetic trees with relatively higher subclade I of the tree topology as determined in bootstrap supports and Bayesian posterior prob- our study were also different from reports in pre- abilities than in previous studies employing the vious studies ofthe trnK intron phylogeny ofAJe- DNA region in the chloroplast genome (Meim- penthes, where these three widely distributed
berg et al. 2001) or the coding region ofthe nucle- species were included in the same subclade as the ar genome (Meimberg & Heubl 2006), except for Bornean species (Meimberg et al. 2001, Meim- the basal positions ofthe trees, It is advantageous berg & Heubl 2006, Meimberg et al, 2006). to use nuclear ITS regions for phylogenetic analy- Despite the diffbrent positions of some spe- sls ln anglosperms. cies in subclade I and the basal species with pre- 7Vepenthes pervillei was found to be the most vious studies of Arepenthes phylogeny, these spe- basal taxon within the genus. It can be distin- cies share similar morphological characteristics guished from all other species by its seeds, which including acute leaf apex, orbicular or partly or-
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
October 2013 ALAMsyAH & ITo-Phylogenetic Analysis efNepenthaceae 123
bicular pitcher lids, and partly cylindrical lower 2006). Lineage sorting is also considered to be a pitcher form, except for IVI ampullaria, which has candidate for this incongruence. In addition to a distinctive urceolate shape (McPherson 2009). the incongruence discussed above, the basal re- Moreover, the pitcher morphology oflVl mirabiiis gion of our trees was not resolved well, and it is was shared by three species from outlying areas highly desirable to examine the phylogeny oflVe- (N khasiana, N distillatoria, and AJI vieiltardii), penthes using other DNA regions. which similarly lack putative apomorphic charac- teristics (McPherson 2009), PIrytogeograpIry qfNepenthes in Southeast Asia The positional differences of species between Meimberg et al. (2001) suggested the scenario the present ITS analysis and previous studies us- that 7Vlepenthes colonized Southeast Asia from an ing trnK intron and PTRI (Meimberg et al, 2001, ancient Indian stock, first te the Malay Peninsula,
Meimberg & Heubl 2006, Meimberg et aL 2006) and subsequently to Indochina and the Malay Ar:-
may be due to diffbrences in the markers used, chipelago, Their results showed the Indian spe- ITS is located in the nuclear genome (Baldwin et cies to be sister to the Southeast Asian species, al. I995), whereas the trnK intron is located in although they did not reach a final conclusion,
the chloroplasts, which are maternally inherited Our results show the Indian and Sri Lanka spe-
and may lead to the chloroplast capture phenom- cies fbrming a clade with the New Guinean ?Vl enon (Soltis & Kuzoff 1995). This phenomenon papuana and with the widely distributed IVI am- is considered the main reason for the incongru- pullaria and does not support the suggested sce- ence of nuclear markers and plastid phylogenies nario ofMeimberg et al. (2001). In addition, three (Meimberg & Heubl 2006, Tsitrone et al. 2003). other New Guinean species, Ail sp. Misool, IVI In contrast, ITS has another weaknesses fbr esti- neogorineensis, and N clanseri, were included in mating phylogeny. ITS is included in the rRNA the basal polytomies with other widely distribut- precursor transcript. Genes encoding rRNA and ed species, i.e., Nl gracilis and Nl miTabilis, The
spacers occur in tandem repeats that are thou- distribution area reconstruction on one ofthe 324
sands of copies long and polymorphisms in their MPTs (Fig. 3) shows that these three widely dis- sequences are often observed (Wendel et al. tributed species may have expanded from east to 1995al. Indeed, in the species of Nepenthes ex- west in Southeast Asia (east to west hypothesis), amined, we detected polymorphisms in seven Ofcourse, due to the polytomy of the basal spe- species when using the PCR protocol with an- cies, the scenario of Meimberg et al, (2001) is nealing temparature at 570C. It is critical to de- also possible based on other MPTs. They consid- cide which is an orthologous sequence for recon- ered that expansion ofNepenthes occurred from structing the phylogenetic trees. It is possible west to east in Southeast Asia, based on their tree that the incongruence of phylogenetic tree topol- topology, in which Indian species N khasiana ogies is due to using paralogous ones. The con- was sister to all of the Southeast Asian species. certed evolution (Wendel et al. 1995aj and in- That topology was not supported by the PTRI tergenomic introgression (Wendel et aL 1995b) of tree, where N khasiana formed a clade with the the nrDNA genes may also have affected this in- species of Sulawesi and Madagascar (Meimberg congruence. In the case of PTRI (Meimberg & & Heubl 2006). The east to west hypothesis was Heubl 2006), the gene is located in the nuclear ge- corroberated by similarities in morphology, alti-
nome, the same as the ITS, The incongruence tudinal distribution and habitat types of the three
betvveen ITS and PTRI trees may be due to am- widely distributed species, which were shared by plification of paralogous sequences. As men- species from New Guinea, Australia and New tioned above, ITS sometimes shows polymoF Caledonia (McPherson 2009). phisms, and PTRI also exhibits the possibility of The two species from Sulawesi, IVI glabrata paralogous genes, since IVlepenthes is suspected and N. tentaculata were included in subclade II to be of polyploid origin (Meimberg & Heubl together with All hirsuta, a species endemic to
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
124 Acta Phytotax. Geobot. Vol, 64
Borneo. This subclade is sister to subclade III- analyses also diffbred, possibly affecting the tree VII consisting of most species occurring in BoF topologies. Because the phylogenetic analysis neo, Sumatra, the Philippines, Malaysia and In- based on ITS has weaknesses, especially in dis-
dochina, This tree topology and presences of IVI criminating between orthologs and paralogs, it is tentacuiata in Sulawesi and the Bornean Islands highly desirable to base phylogenetic reconstruc- also suggests an important linkage for species of tions on other DNA regions and by using addi-
IVlepenthes distributed acress Southeast Asia tional species of fVepenthes, from east to west (Fig, 3), The flora ofSulawesi is closely linked with the Australian-New Guinean Mo rphology of the peristome phytogeographic regien, which is indicated by Some morphological characteristics in the the floral exchange between these three areas species of Nbpenthes are polymorphic, and each (van Welzen et aL 2011). Therefore, Sulawesi may ofthem has more than one fbrm or shape within a have been a bridge for the migration oflVleipenthes species, making it difficult to evaluate the evolu-
from east to west. tionary trends in morphological characteristics,
The distribution area reconstruction of IVle- In this study, we tried to analyze some morpho- penthes (Fig. 3) suggests that Nbpenthes diverged logical characteristics, such as indumentum, in- in Borneo, resulting many local species and could fiorescence, lamina, leaf apex and base, lower have also migrated to adjacent areas. NOpenthes and upper pitcher fbrms and the lid and peristome probably migrated from Borneo to the Philip- ofthe pitcher. We found theperistome ofthe up- pines. This scenario is supported by the presence per pitchers to be consistent, except in N min dnn- of N canep7anutata in the basal position of sub- aoensis, N alata, N ovata, and AC densij7ora. clade III, which includes Philippine species. In The peristome of the upper pitchers were relative- addition, subclade IV consisting of Philippines ly well correlated with the grouping ofNepenthes species was included in clades (subclade lll-VII), into seven subclades. which is sister to subclade II consisting of species Figure 4 shows the evelutionary trends in the flrom Borneo and Sulawesi. This scenario also peristomes. The peristomes ofthe two basal spe- differs from that ofMeimberg et aL (2001), which cies, as well as in most species in the basal poly- suggested migration from the Philippines to Bor- tomies and subclades I, Il, and III, were narrow, neo. In subclade VI, the two Bornean species suggesting that the narrow peristome is the ple- were iocated in basal positions as sisters to the siomorphic state in IVepenthes, In addition, all Indochina species and to all species of subclade species in subclade VI produced narrow peri-
VII that occur on Sumatra. Our results therefore stomes. Whereas the Bornean species in sub-
suggest an ancestral species migrated from Bor- clade V and most species in subclade IV and VII,
neo, diverged in Sumatra, then subsequently mi- as well as N anu)ullaria in subclade I and two
grated to Java (N spathutata is also known to be basal species (An neoguineensis and N. Toivanae), distributed in Java). It can be assumed that Bor- had intermediate and broad peristomes. The in- neo was a secondary center of diversification fbr termediate and broad peristomes likely evolved at
AJbpenthes, allowing species ofAJepenthes to ra- least seven or eight times, respectively, whereas diate within the Sunda Shelfregion. narrow peristomes evolved at least five times afl The scenario for Nbpenthes expansion esti- ter the widening of the peristome. The findings mated from our molecular analysis based on ITS suggest repeated evolution ofnarrow, intermedi- differs from that ef Meimberg et aL (2001).ate, and broad peristomes, These two scenarios were based on results from In the taxonomic study by Danser (1928). nar- different DNA regions located in different ge- row peristomes were considered te be ancestral nomes and may be due to the characteristics of in the vatigatae group, which is concordant with the difTerent DNA regions used in the phyloge- our results, suggesting that N, pervillei and 7Vl netic analyses, The species examined in both madugascariensis are basal species and that the
NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics
October2D13 ALAMsyAII&ITo-PhylogeneticAnalysisofNepenthaceae 125
tree topology of subclade I and the basal polyto- mies consist of three species from the outlying References areas (IVI distiUatoria, N khasiana, and N vieil- Akhriadi, P,, Hernawati, A. Prirnaldhi & M. Hambali. lardii), two New Guinean species (IVI dunseri and 2009. A?t)enthes naga, a new species ofNepenthace- N papuana), and two widely distributed species ae from Bukit Barisan ofSumatra. Reinwardtia 12: and IVI minabilis), All ofthese species (iVlgracilis 339-342. were assigned by Danser (1928) to the hiigatae Albert, V. A., S. E. Williams & M, W. Chase. 1992. Car- group (Fig. 4). Although ?Vl neoguineensis in the nivorous plants: phylogeny and structural evorution. Science 257: 149I-1495. basal polytomies produces intermediate peri- Alejandro, G. J. D,, R. S. Madulid & D, A, Madulid. 2008. stomes, Danser (1928) assigned it to the l)ltigatae The utility ofinternal transcribed spacer (nrDNA) se- group. Other species with narrow peristomes quence data for phylogenetic reconstruction in en- from three difTerent subclades were also assigned demic Philippine Nbpenthes L. (Nepenthaceae). by Danser (1928) to the hiigatae group. Since Philipp. ScL 45: 99-110. Baldwin, B. G., M. J. Sanderson, J. M. Porter, M. F. representatives of the Pltigatae group are scat- Wojeiechowski, C, S. Campbell & M. J. Donoghue. tered in different subclades (Fig. 4), the group 1995. The ITS region of nuclear ribosomal DNA: a with narrow peristomes is obviously polyphylet- valuable source efevidence on angiosperm phyioge- ic. ny. Ann. Missouri Bot. Gard. 82: 247-277. Intermediate peristomes can be defined as Cheek, M. & M. Jebb, 2009. IVepenthes group Montanae
one of the key characteristica of the Nbbiles (Nepenthaceae) in Indo-China, with N. thai and N bokor described as new, Kew Bull, 64: 319-325, group, while broad peristomes can be defined as Cuenoud, P., V, Savelainen, L, W, Chatrou, M, Po-'ell, R, a characteristic of the Regiae, Insignes and U}"- J. Grayer & M, W, Chase, 2002, Molecular phy]oge- ceolatae 1928). Since represen- groups (Danser netics of Caryophyllates based on nuclear 18S rDNA tatives of the insignes, Regiae, and IVbbiles and plastid rbeL, aipB, and matK DNA sequences. Amer. J. Bot. 89: 132-144. groups occur in different subclades (Fig. 4), the Clarke, C. 2006. Nbpenthes of Borneo. Natural History three groups that share intermediate and broad Publications, Kinabalu, are obviously Judging peristomes polyphyletic, Clarke, C. 2007. Nepenthes of Sumatra and Peninsular fi'om the tree topology 4), the evolutionary Technology Development Fund from MOE, and (D-1008) V. Savolainen, M. W. Chase, M. P. Powel], L. A. Al- Natonal Bioresouce Project and GRENE environmental ice, R. Evans. H. Sauquet, C. Neinhuis, T. A. Slotta, J. infbrmation from MEXT, and a Grant-in-Aid for scien- G. Rohwer, C. S. Campbell & L. W. Chatrou. 2003. tific research from JSPS to Ml. (22405012) Angiosperm phylogeny based on matK sequence in- formation. Amer. J. Bot. 90: 1758-1776. NII-Electronic Library Service TheTheJapaneseSociety Japanese Society for Plant Systematics 126 Acta Phytotax. Geobot, Vo!. 64 Jobb, G,, A. von Haeseler & K. Strimmen 2004. Rambaut, A. & A, J. Drummond, 2007. Tracer vl.4. TREEFINDER: a powerful graphical analysis envi- Availablefromhttp:fibeast.bio.ed.ac.ukfTracer ronment for molecular phylogenetics. BMC Evol. Robinson, A. S. A. S. Fleischmann, S. R, McPherson, V, Biol. 4: 18 B. Heinrich, E. P. Gironella & C. Q. Pena. 2009. A Johnson, L. E. & D, E. Soltis, 1995. Phytogenetic infer- spectacu!ar new species of ?Vepenthes L, (Nepentha- ence in Saxij}'agaceae sensu stricto and Gilia (Pole- ceae) piteher plant from centrai Palawan, Philippines. moniaceae) using matK sequences, Ann. Missouri Bot. J. Linn. Soc. 159: 195-202, Bot. Gard. 82: 149-175. Ronquist, F. & J, P. Huelsenbeck. 2003. MrBayes 3: Krutzsch, W, 1988. Pa;aeogeography and historical phy- Bayesian phylogenetic inference under mixed mod- togeography (palaeochorology) in the Neophyticum. els. Bioinformatics 19r 1572-1574. Pl. Syst. EvoL 162: 5-61. Saitou, N. & M, Nei. 1987, The Neighborjoining method: Larkin M. A., G. Blackshielcls, N. P. Brown, R. Chenna, A new method fbr reconstructing phylogenetic trees. P. A. McGettigan, H. McWilliam, F, Valentin, I. M. Molec. Biol. Evol. 4: 406-425. Wallace, A. Wilm, R. Lopez, J. D. Thompson, T. J. Soltig, D. E. & R. K. Kuzoff 1995. Discordance between Gibson & D. G. Higgins. 2007. Clustal W and Clustal nuclear and chtoroplast phylogenies in the Hkuchera X version 2.0. Bioinformatics 23: 2947-2948. group (Saxifragaceae). Evolution 49: 727-742, Lee, C. C., Hernawati & P. Akhriadi. 2006, Tiwo new spe- Swofford, D, L. 2002. PAUP': Phylogenetic analysis us- eies of Nepenthes (Nepenthaceae) from north Suma- ing parsimony C"and other methods) version 4.0blO. tra. Blumea 51: 561-568. Sinauer Associates, Sunderland. Maddison. W. P. & D. R. Maddison. 2003. MacClade 4: Tamura, K. 1992. Estimation ofthe number of nucleotide Analysis of phylogeny and character evolution ver- substitutions when there are strong transition-trans- sion 4.06. SinauerAssociates, Sunderland. version and G+C-content biases. Molec. Biol. Evol. McPherson, S. 2009. Pitcher plants of the old world. Vbl. 9I 678-687, 1 and Vol. 2. Redfern Natural History Productions, Tamura, K., D, Petersen, N. Peterson, G, Stecher, M, Nei Poole. & S. Kumar. 201I. MEGA 5: Molecular evolutionary Meimberg, H., A. Wistuba, P. Dittrich & G, Heubl. 2001, genetics analysis using maximum likelihood, evolu- Molecular phylogeny of Nepenthaceae based on cla- tionary distance, and rnaximum parsirneny methods. distic analysis of plastid trnK intron sequence data. Molec, Biol. Evol. 28: 2731-2739, Pl. BioL (StuttgarO 3: 164-175, Tsitrone, A,, M, Kirkpatrick & D, A. Levin, 2003. A tnod- Meimberg, H. & G. Heubl. 2006, Introduction ofa nucle- el for chloroplast capture. Evolution 57: T776-1782. ar marker for phylogenetic analysis ofNepenthaceae. yan den Berg, C., D. H. Goldrnan, J. V. Freudenstein, A. Pt. Biol. (Stuttgart) 8: 831-840. M, Pridgeon, K. M. Cameron & M. W. Chase. 2005. Meimberg, H. S. Thalhammer, A. Brachmann & G. Heu- An overview of the phylogenetic relationships within bl. 2006. Comparative analysis of a translocated Epidendroideae inferred from multiple DNA regions copy of the trnK intron in carnivorous family Nepen- and recircumscription of Epidendreae and Arethuse- thaceae. Molec. Phylogen. Evol. 39: 478-490. ae (Orchidaceae). Arner. J. Bot. 92: 613-624. Meimberg, H., H. Rlscher, F. G. Turini, V. Chamchum- van Welzen, P. C., J. A. N Parnell & J. W. F. Slik. 2011. roon, M. Dreyer, M. Sommmaro, G. Bringmann & G. Wallace's Line and plant distributions: two or three Heubl. 2010. Evidence for species differentiation phytogeographical areas and where to group Java? within the Ancistroctadus tectorius complex (Ancis- Biol. J. Linn. Soc, 103: 531-545, trocladaceae) in Southeast Asia: a molecular ap- Wendel, J. F., A. Schnabel & T. Seelanan. 1995a. Bidirec- proach. Pl. Syst. Evol. 284: 77-98. tional interlocus concerted evolution fo11owing allo- Moran, J. A. 1996. Pitcher dirnorphism, prey composition polyploid speciation in cotton (GosopLpium). Proc. and the mechanisms ofprey attraction in the pitcher Natl. Acad. ScL U.S.A, 92: 280-284, plant Nlezpenthes rcdi7esiana in Borneo. J. Ecol. 84: Wendel, J. F., A. Schnabel & T, Seelanan, 1995b. An un- 51S-525. usual ribosema] DNA sequence from Gossxpium Posada, D. 2008.jModelTest: phylogenetic model averag- goss)tpioides reveals ancient, cryptic, intergenomic ing. Molec. Biol, Evol, 2S: 1253-12S6, introgression, Molec, Phylogen. Evol. 4: 298-313, ReceivedDecember 29, 2012; aecqptedMby Z 2013 NII-Electronic Library Service