Molecular Phylogenetics and Evolution 77 (2014) 216–222

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev

A new phylogenetic analysis sheds new light on the relationships in the Calanthe alliance (Orchidaceae) in China Jun-Wen Zhai a,b,c,h, Guo-Qiang Zhang c, Lin Li b, Meina Wang c, Li-Jun Chen c, Shih-Wen Chung d, Francisco Jiménez Rodríguez e, Javier Francisco-Ortega f,g, Si-Ren Lan a,h, Fu-Wu Xing b,⇑, Zhong-Jian Liu a,c,⇑ a

College of Landscape Architecture, Fujian Agriculture and Forestry University, Fuzhou 350002, China South China Botanical Garden, Chinese Academy of Sciences, Guangzhou 510650, China Shenzhen Key Laboratory for Orchid Conservation and Utilization, The National Orchid Conservation Center of China and The Orchid Conservation & Research Center of Shenzhen, Shenzhen 518114, China d Department of Botany, Taiwan Forestry Research Institute, Taipei 10066, Taiwan e Jardín Botánico Nacional, Apartado Postal 21-9, Santo Domingo, Dominican Republic f Department of Biological Sciences, Florida International University, Miami, FL 33199, USA g Kushlan Tropical Science Institute, Fairchild Tropical Botanical Garden, Coral Gables, Miami, FL 33156, USA h Orchid Conservation and Research Center of Fujian Agriculture and Forestry University, Fuzhou 350002, China b c

a r t i c l e

i n f o

Article history: Received 9 October 2013 Revised 21 February 2014 Accepted 3 April 2014 Available online 18 April 2014 Keywords: The Calanthe alliance Calanthe Cephalantheropsis Phaius Orchidaceae Phylogenetics

a b s t r a c t The taxonomy of the Calanthe alliance (Epidendroideae, Orchidaceae), consisting of Calanthe, Cephalantheropsis, and Phaius, has been difficult for orchidologists to understand because of the presence of common morphological features. In this study, in addition to morphological and geographical analyses, maximum parsimony and Bayesian inference analyses were performed based on nucleotide sequences of the nuclear internal transcribed spacer and cpDNA genes of 88 taxa representing the major clades of the Calanthe alliance in China. The results indicated that Cephalantheropsis is monophyletic, while both Phaius and Calanthe are polyphyletic. In Phaius, a total of three species, P. flavus, P. columnaris, and P. takeoi, were segregated to form a new genus, Paraphaius. In Calanthe, subgenus Preptanthe and sect. Styloglossum were both categorized as distinct genera from Calanthe. Our results also confirm that Calanthe delavayi and C. calanthoides are members of Calanthe. Previous studies assigned C. delavayi to Phaius and C. calanthoides to Ghiesbrechtia. Five sections, namely, Alpinocalanthe, Puberula, Ghiesbrechtia, Tricarinata, and Calanthe, three of which are new taxa, were recognized in Calanthe. Therefore, we propose that the Calanthe alliance is composed of six genera: Calanthe, Cephalantheropsis, Paraphaius, Phaius, Preptanthe and Styloglossum. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Calanthe, Cephalantheropsis, and Phaius form an independent alliance within Epidendroideae (Orchidaceae), which can be easily distinguished from other taxa in this subfamily. This alliance is characterized by plicate leaves, similar sepals and petals, and eight waxy pollinia forming two groups. However, it is difficult for orchidologists to determine the phylogenetic and taxonomic relationships within this alliance. Cephalantheropsis is characterized by its labellum, which does not have spurs and is separate from

⇑ Corresponding authors. Address: Shenzhen Key Laboratory for Orchid Conservation and Utilization, The National Orchid Conservation Center of China and The Orchid Conservation & Research Center of Shenzhen, Shenzhen 518114, China (Z.-J. Liu). E-mail addresses: [email protected] (F.-W. Xing), [email protected] (Z.-J. Liu). http://dx.doi.org/10.1016/j.ympev.2014.04.005 1055-7903/Ó 2014 Elsevier Inc. All rights reserved.

the column, and by its pollinium, which grows directly on the globose viscidium. The labellum of Phaius also grows at the base of the column but is not adnate with column wings and usually has a spur, and its pollinium is generally attached by short caudicles. Calanthe is characterized by its labellum, which is most often adnate to column wings forming a tube and spurred base, and its pollinium has conspicuous or inconspicuous caudicles, usually adhering to a sticky viscidium (Chen et al., 1999, 2009; Perner and Cribb, 2002; Kurzweil, 2010; Pridgeon et al., 2005). However, the aforementioned features are ambiguous and have poor taxonomic value to enable distinctions among the genera or infrageneric taxa of the Calanthe alliance. One of the most important characteristics of Calanthe is that its lip is adnate to the column wing, forming a tube. However, this feature is not shared by some species of the sect. Styloglossum or any species under the subgenus Preptanthe. Sect. Styloglossum has been treated as a section of

J.-W. Zhai et al. / Molecular Phylogenetics and Evolution 77 (2014) 216–222

subgenus Calanthe from genus Styloglossum (Breda, 1827), which was widely accepted by most orchid specialists (Chen et al., 1999, 2009). Caducous bracts are an important feature distinguishing the sect. Styloglossum from the other taxa within Calanthe. However, some species of Cephalantheropsis and Phaius also possess this feature. Morphologically, sect. Styloglossum has a closer relationship to several species under Cephalantheropsis and Phaius because its stems and flowers do not open widely. Subgenus Preptanthe, which is characterized by swollen pseudobulbs, labella free from column wing, and deciduous leaves, was established as an independent genus Preptanthe (Reichenbach, 1853), but later came to be regarded as a subgenus (Schlechter, 1912). Smith (1915) treated Preptanthe as a section, but this concept has not been widely accepted (Chen et al., 1999, 2009). The ambiguous species Calanthe delavayi has always been considered a member of Calanthe (Chen et al., 1999). However, Perner and Cribb (2002) transferred this species to Phaius based on the fact that its labellum surrounds the column and is separated from the column wing, and this treatment was accepted by Chen et al. (2009). Cephalantheropsis species were usually treated as Phaius or Calanthe species before the establishment of this genus by Guillaumin (1960). For example, Calanthe dolichopoda, identified by Fukuyama (1935), and Phaius longipes, distinguished by Holttum (1947), were both later considered as synonyms of Cephalantheropsis longipes (Ormerod, 1998). Phaius includes approximately 40 species worldwide, and nine species are restricted to China (Chen et al., 1999; Kurzweil, 2010; Huang et al., 2012). This genus can be divided into two types based on plant morphology: bract caducous or persistent. However, a genetic study has not been conducted. Chen et al. (2010) performed a molecular phylogenetic analysis of the Calanthe group using one cpDNA marker, rpcL-trnL. The results of this study confirmed the traditional classification of three genera. However, a limited number of samples were used for the study, i.e., five species of Phaius, one species of Cephalantheropsis, and three species of Calanthe as an outgroup (Calanthe alismaefolia, Calanthe sieboldii, and Calanthe triplicata, which belonged to sect. Calanthe in the previous classification). Species of other sections or subgenera were not considered. Thus, the study by Chen et al. (2010) had limited value for efforts to determine the phylogenetic relationships among the three genera. In the current study, we sampled three species of Cephalantheropsis, six species of Phaius, and 41 species of Calanthe, 38 species of which belong to subgenus Calanthe and the remaining three of which belong to the subgenus Preptanthe. Calanthe comprised 32 species of sect. Calanthe (34 samples), six species of sect. Styloglossum, and one species of sect. Ghiesbrechtia. Four molecular markers were used for the phylogenetic reconstruction: the Internal Transcribed Spacer (ITS) of the nuclear ribosomal DNA and the trnL-F, atpI-atpH, and matK regions of the chloroplast genomes.

2. Materials and methods 2.1. Materials Specimens were collected from either cultivated or wild plants. We selected more than 50 species of the Calanthe alliance following the current classification of Pridgeon et al. (2005) and Chen et al. (2009), including three species of Cephalantheropsis, six species of Phaius, and 41 species of Calanthe (including three species of subgenus Preptanthe and 38 species of subgenus Calanthe). Calanthe comprised six species of sect. Styloglossum and 32 species of sect. Calanthe. These species were mainly from China except for Calanthe rubens from Vietnam and Calanthe calanthoides from Valle Nuevo. Acanthephippium sylhetense, Ania hongkongensis,

217

Bletilla striata, Collabium chinense, Cymbidium kanran, Cymbidium goeringii, Geodorum recurvum, Mischobulbum cordifolium, Nephelaphyllum tenuiflorum, Oberonia japonica, Oberonia kwangsiensis, and Tainia dunnii were chosen to represent outgroups. Table S1 presents a list of the taxa analyzed in this study, including information on voucher specimens and Genbank accessions. 2.2. Methods 2.2.1. DNA extraction Total DNA was extracted from fresh or silica-gel dried leaves using a modified version of the CTAB method (Doyle and Doyle, 1987), with samples crushed by a MiniBeadbeater-16 (Biospec). 2.2.2. PCR amplification and sequencing Table S2 lists primers (Mike et al., 1999; Reeves et al., 2001; Sulaiman et al., 2003; Liu et al., 2011) used for PCR and subsequent cycle sequencing protocols. Polymerase chain reactions (PCRs) were performed in a reaction mix (30 ll) containing total DNA (1 ll), primers (2 ll each), 2 MightyAmp Buffer (Ver.2) (15 ll), MightyAmp DNA polymerase (Takara Bio) (1 ll), and H2O (9 ll). The PCR program consisted of an initial 2 min pre-melt stage at 98 °C, followed by 35 cycles of 20 s at 98 °C (denaturation), 30 s at 45–55 °C, and 50–110 s at 68 °C, followed by a final 8 min extension at 68 °C. The PCR products were run on 1.5% agarose gels to check the quality of amplified DNA. Target products were excised from these gels, purified using extraction kits (OMEGA BIO TEK), and sequenced by Invitrogen (Shanghai). 2.2.3. Data analysis Both forward and reverse sequences were edited and assembled with DNASTAR (http://www.dnastar.com/). DNA sequences were aligned with MEGA5.05 (Tamura et al., 2011) under the Muscle model (Edgar, 2004) and manually adjusted to account for obvious or missing inserts. Aligned sequences are available from the corresponding authors upon request. The homogeneity between the ITS data and the combined chloroplast DNA (cpDNA) dataset (i.e., trnL-F, atpI-atpH, matK) was tested using the incongruence length difference (ILD) test (Farris et al., 1995), as implemented in PAUP* version 4.0b10 (Swofford, 2002), and following procedures indicated by Li et al. (2011). Maximum Parsimony (MP) and Bayesian Inference (BI) methods were applied in phylogenetic analyses. Nucleotides that could not be identified were treated as unknown (?), and gaps were edited as missing data. Maximum Parsimony inference was conducted with PAUP* 4.0b10 (Swofford, 2002). All characters were equally weighted and heuristic searches were conducted with 1000 random additional replicates involving TBR branch swapping. Bootstrap values were calculated from 1000 pseudo-replicates, each with 100 random additions. For the Bayesian analyses, MrModeltest 3.7 (Posada and Crandall, 1998) was used to determine the nucleotide substitution model. The selected nucleotide model was: ITS, SYM + G; cpDNA and combined dataset, GTR + I + G. The first 2.5 million generations were discarded as burn-in periods before sufficiently stationary generations, and the remaining trees were used to calculate posterior probabilities. Bayesian searches were conducted by Markov Chain Monte Carlo algorithms with two independent sets of four chains, each run for ten million generations, and sampling was carried out every 100 generations by MrBayes 3.1.2 (Ronquist and Huelsenbeck, 2003). Genetic distance analysis was performed using MEGA5.05. The data were divided into 10 groups corresponding to 10 sections/ genera within the Calanthe alliance. The mean distance among groups was computed using 1000 bootstrap replications and the Kimura 2-Parameter model (K2P) following the procedures outlined by Liu et al. (2011).

218

J.-W. Zhai et al. / Molecular Phylogenetics and Evolution 77 (2014) 216–222

3. Results 3.1. ITS analysis Eighty-eight taxa were included in the ITS data matrix, 12 of which were part of the outgroup. Table 1 presents the numbers of variable and parsimony informative sites, as well as tree statistics after MP analyses. Details for the best-fit models (needed for BI analyses) are presented in Table 2. The results obtained from the BI analyses (Fig. S1) showed that the Calanthe alliance is divided into five clades. The first divergent clade contained three species of subgenus Preptanthe (PP 96%): C. rubens, C. cardioglossa, and C. labrosa. The second divergent clade comprised three species of Phaius: P. mishmensis, P. tankervilleae, and P. wallichii (PP100%). The third clade was composed of two sister groups of: (1) Cephalantheropsis (C. longipes, C. obcordata, and C. halconensis) and (2) sect. Styloglossum of Calanthe (C. angustifolia, C. clavata, C. densiflora, C. lyroglossa, and C. speciosa). The fourth clade was made up of three species of Phaius (P. columnaris, P. flavus, and P. takeoi). The last clade included the entire sect. Ghiesbreghtia of Calanthe (C. calanthoides) and the rest of the species of sect. Calanthe. The MP phylogram (Fig. S2) does not reference significance here because of its low bootstrap percentage. 3.2. cpDNA analysis

nucleotide sequences from both the nuclear and cpDNA markers. Table 1 presents the numbers of variable and parsimony informative sites and tree statistics for MP analyses, and the best-fit molecular evolutionary models are presented in Table 2. The ILD test for the nrDNA and combined cpDNA data resulted in P = 0.01, indicating incongruence between the two datasets. We speculate the differences between these two data matrices are the result of frequent natural hybridization within Calanthe. Good examples are described taxa such as Calanthe  alboliacina, Calanthe  subhamata, Calanthe  varians and Calanthe  dominyi (Karasawa and Ishida, 1998; Chen et al., 2009). The results of Bayesian analysis (Fig. 1) showed that the Calanthe alliance can be divided into nine clades (PP = 100%, except for one clade supported with a PP value of 92%). Six of these clades were congruent with the cpDNA analysis and were also well supported (PP = 100%). Sect. Calanthe was divided into three clades; however, C. puberula formed an independent lineage (PP = 100%), in agreement with the ITS analysis. The topology after the MP analysis (Fig. 1) is congruent with that produced by the Bayesian analysis, although most of the nodes were not supported as well. In addition, the MP analyses did not support the two groups identified in sect. Calanthe.

4. Discussion 4.1. An updated phylogeny of the Calanthe alliance

Sixty-nine samples including 13 species identified as outgroups were sampled for the cpDNA analysis. Nucleotide sequences of the three plastid markers were combined in a single data matrix. Table 1 presents the numbers of variable and parsimony informative sites, as well as tree statistics for MP analyses. Details pertinent to the best-fit model of molecular evolution can be found in Table 2. Eight major clades were recovered after BI analyses (Fig. S3). The first four divergent clades were congruent with those recovered in the ITS topology, with a PP of 100%. The fifth clade included C. delavayi, and the sixth only consisted of the Neotropical species C. calanthoides (PP 100%). All species of sect. Calanthe were clustered in a single lineage, which was divided into two distinct clades (PP = 100%). The first clade included C. alismaefolia, C. argenteostriata, C. davidii, C. herbacea, C. puberula, C. sinica, C. sylvatica, C. tricarinata, C. triplicata, C. wenshanensis, and C. odora, and the other clade included C. arisanensis, C. aristulifera, C. bingtaoi, C. graciliflora, C. hancockii, C. henryi, C. lechangensis, C. mannii, C. plantaginea, C. sieboldii and C. tsoongiana. The tree obtained after MP analysis recovered the same major lineages, and its topology was virtually identical to that produced by the BI searches (Fig. S2).

This study presents an updated phylogeny of the Calanthe alliance. We examined an extensive sampling of species, and our results are based on the phylogenetic analysis of markers from two different unlinked genomes. The results of this study supported Cephalantheropsis as monophyletic, while Calanthe and Phaius were recognized as polyphyletic. Compared with previous treatments (Chen et al., 1999; Perner and Cribb, 2002; Chen et al., 2010), this study shed new light on the Calanthe alliance. Our research demonstrates that a taxonomic interpretation of the Calanthe alliance is necessary. 4.2. Phaius Lour In this study, Phaius species were identified to form two clades (Fig. 1). The first clade comprised P. tankervilleae, P. wallichii, and P. mishmensis, a sister lineage to subgenus Preptanthe of Calanthe. The second clade with P. takeoi, P. columnaris, and P. flavus is closely related to Cephalantheropsis, sect. Styloglossum of Calanthe, C. delavayi, and C. calanthoides.

3.3. Combined analysis Sixty-nine samples including 13 outgroup species were included in the analysis of the data matrix that combined Table 1 Statistics from the analyses of the various dataset. Information

No. of taxa Aligned length No. variable characters No. informative characters (%) Tree length Consistency index (CI) Retention index (RI) Rescaled consistency index (RC)

Calanthe trees ITS

cpDNA

Combined

88 666 400 71 (10.66%) 1052 0.5656 0.8692 0.4916

69 3851 1410 876 (22.74%) 2829 0.6557 0.7837 0.5139

69 4644 2038 1279 (27.54%) 4038 0.6647 0.8096 0.5381

4.2.1. Phaius tankervilleae, P. wallichii, and P. mishmensis versus P. takeoi, P. columnaris, and P. flavus This study is the first to report the divergence within Phaius (excluding C. delavayi). Morphologically, the P. tankervilleae – P. wallichii – P. mishmensis clade has caducous floral bracts and eight pollinia in two groups that are separated from each other. The second clade has persistent floral bracts and pollinia in two groups, but the pollinia are attached by caudicles to a sticky substance. P. columnaris was previously recorded as having caducous floral bracts (Chen et al., 1999, 2009), which were confirmed to be persistent after careful observation (Jun-Wen Zhai, unpublished information). Therefore, we propose that Phaius is restricted to the lineage P. tankervilleae – P. wallichii – P. mishmensis, including the type species P. tankervilleae. Furthermore, we suggest a new genus, Paraphaius, to encompass the lineage P. takeoi – P. columnaris – P. flavus. Phaius flavus was chosen as the basionym for the type species of this new genus.

219

J.-W. Zhai et al. / Molecular Phylogenetics and Evolution 77 (2014) 216–222 Table 2 Best-fit model and parameter for each dataset of Orchidaceae. Region

AIC select model

Base frequencies

Substitution model(rate matrix)

A

C

G

T

A–C

A–G

A–T

C–G

C–T

G–T

ITS cpDNA Combined

TrN + I + G GTR + I + G GTR + I + G

0.2307 0.2868 0.2579

0.2377 0.1826 0.2111

0.2592 0.2360 0.2378

0.2724 0.2946 0.2933

1.0000 1.4600 1.3345

3.3453 3.0809 2.3857

1.0000 0.5896 0.6980

1.0000 0.8706 0.4229

3.9282 4.8115 2.6901

1.0000 1.0000 1.0000

I

G

0.1093 0.4756 0.2120

1.3333 0.7061 0.4854

Fig. 1. Phylogram obtained from BI analysis of the combined nrDNA ITS and cpDNA data.

4.2.2. Celanthe delavayi Celanthe delavayi has been widely considered to be a link between Calanthe and Phaius. Morphologically, it is similar to Calanthe with its rather small individual, basal leaves, inconspicuous pseudobulbs, and elongated column. However, it also

resembles Phaius because its labellum embraces its column. Chen et al. (1999) treated it as a member of sect. Calanthe, but in their recent treatment for the Flora of China project (Chen et al., 2009); they accepted a taxonomic placement within Phaius as proposed by Perner and Cribb (2002).

220

J.-W. Zhai et al. / Molecular Phylogenetics and Evolution 77 (2014) 216–222

Furthermore, Genetic distance clearly demonstrated that this taxon has a closer relationship with Calanthe than with Phaius (Table S3). Therefore, we also suggest that C. delavayi should be kept within Calanthe rather than within Phaius. We propose a new unispecific section, Alpinocalanthe, to accommodate this taxon because of its unique morphological features and independent phylogenetic placement. 4.3. Cephalantheropsis Guill Three species of Cephalantheropsis were grouped in one clade (PP = 100%/BP = 98%). This clade is a sister to the clade of Calanthe sect. Styloglossum. Cephalantheropsis halconensis from southern Taiwan was recently recognized as an independent species (Chen et al., 2009). Species with caducous bracts, as well as nodded and erect stems, are apparently quite similar to those belonging to Cephalantheropsis and Calanthe sect. Styloglossum. However, Cephalantheropsis is distinct because of its labellum without a sac or spur, its inflorescence arising from nodes on the stem, and the callose mid lobe of the labellum. Thus, we suggest that Cephalantheropsis should maintain its taxonomic status as a separate genus. 4.4. Calanthe R. Br In the previous taxonomic system, which is mainly based on Schlechter’s (1912) treatment, Calanthe is divided into two subgenera, namely subgenus Preptanthe (Rchb.f., 1853), corresponding to Bentham’s genus Vestitae, with large fleshy pseudobulbs and deciduous leaves, and subgenus Eu-calanthe R.Br., which contains species with smaller pseudobulbs and persistent leaves. Most recent studies (Chen et al., 1999, 2009; Pridgeon et al., 2005) have accepted these two subgenera, but some taxonomic rearrangements have been proposed among the six sections of subgenus Calanthe based on morphological characteristics. However, Calanthe is confirmed as a phylogenetically relevant classification, and our molecular phylogenetic study identified several distinct groups within Calanthe. 4.4.1. Calanthe Subgenus Preptanthe Preptanthe was established as a genus by Reichenbach (1852) based on Calanthe vestita Lindl. However, Schlechter (1912) recognized it as a subgenus within Calanthe, a taxonomic placement that has been accepted in most subsequent studies. This subgenus is distinguished by its deciduous leaves in dry seasons, its large, swollen and fleshy pseudobulbs, and its column wing free from labellum. This subgenus contains approximately 10 species worldwide, which are mainly distributed in tropical Asia and the Pacific islands. Our phylogenetic study showed that these species make up a strongly supported clade (PP = 100%/BP = 91%). Our results also showed that the Preptanthe clade is closely related to Phaius and Cephalantheropsis, rather than to Calanthe. Therefore, retaining this taxon as part of Calanthe will make the latter paraphyletic. Therefore, we suggest that Preptanthe should be considered as an independent genus within the Calanthe alliance.

However, Cephalantheropsis can be distinguished by a labellum without sacs or spurs, cylindrical and erect stems, and inflorescence arising from nodes in the lower part of the stem. Consequently, we suggest that sect. Styloglossum should be recognized as a separate genus, Styloglossum. 4.4.3. Calanthe sect. Ghiebrechtia The unispecific Calanthe sect. Ghiebrechtia was established by Richard and Galeotti (1845) to recognize the contributions of Ghiebrecht to plant classification. This taxon is based on Ghiebrechtia calanthoides. However, its systematic placement has been controversial. Schlechter (1912) and Seidenfaden (1975) recognized Ghiebrechtia as an independent genus, but Hamer (1974) transferred it to Calanthe as C. calanthoides, which was accepted by Cribb and Thomas (1993) and Pridgeon et al. (2005). Based on our molecular analysis, we recognize C. calanthoides as a section under Calanthe. 4.4.4. Calanthe sect. Calanthe Calanthe sect. Calanthe is the largest infrageneric group of the genus, comprising approximately 140 species worldwide, 50 of which are found in China. In this study, all of the 23 species formed an independent clade that contains three lineages. The first lineage contains only C. puberula, and this taxon is sister to a clade that contains the rest of the species of this section. Morphologically, C. puberula is identified by its small individual, elongated petals, and its lip without lamellae or calli. We suggest establishing a new unispecific sect. Puberula to accommodate this lineage. The remaining species formed the two other lineages of this section, namely (1) triplicata-sylvatica-davidii and (2) gracilifloratsoongiana. The labellum of species belonging to the former clade has warty calli and a two-lobed mid-lobe. In contrast, the labellum of those clustered in the graciliflora-tsoongiana clade has ridges or lamellae and a mid-lobe. Thus, we suggest that these lineages should be recognized as two sections within the genus. Calanthe triplicata is the type species of sect. Calanthe. The new section, Tricarinata (with C. tricarinata as the type species), is proposed to accommodate the taxa belonging to the graciliflora-tsoongiana clade. Researchers have speculated that cryptic species are present within some species in this group. The widely distributed species C. sylvatica from Guangdong, Guangxi, and Taiwan is present in different clades. The same situation was found in C. triplicata from Hainan and Taiwan and C. tsoongiana from Zhejiang and Guizhou (Fig. S3). Cryptic species are often found throughout the animal and plant kingdoms (Collins and Paskewitz, 1996; Hebert et al., 2004; Lahaye et al., 2008). Furthermore, cryptic species are also reported in some orchid genera. Bower and Brown (2009) identified five undescribed cryptic taxa in the highly pollinator-specific Chiloglottis gunnii complex based on evidence of specific pollinators. Geographical isolation, different pollinators, and frequent hybridization are speculated as the potential causes for the cryptic taxa in Calanthe. 4.5. Morphological evolution

4.4.2. Calanthe sect. Styloglossum Styloglossum was established by Breda (1827) as an independent genus, but Schlechter (1912) regarded it as a section within Calanthe subgenus Calanthe, and this was accepted by most subsequent studies (Chen et al., 1999, 2009; Pridgeon et al., 2005). Sect. Styloglossum has caducous floral bracts, flowers that do not open widely, and inflorescence arising from the upper node of the pseudobulb. However, these morphological features place sect. Styloglossum near Cephalantheropsis rather than Calanthe, as confirmed by our molecular analysis. Indeed, species of this section formed a sister clade to Cephalantheropsis (PP = 100%/BP = 98%).

The phylogram shows that taxa with persistent floral bracts (Preptanthe) were the first to diverge from the Calanthe alliance, followed by the taxa with caducous bracts (Phaius and Cephalantheropsis), and then by those with persistent bracts. Thus, these data support multiple independent instances of the evolution of floral bracts in the Calanthe alliance. The trait of the column wing adnate to the labellum base has also evolved independently several times within this alliance. Interestingly, this highly homoplastic trait is one of the most relevant traits distinguishing the genera of this alliance. The presence and length of the spurs are important

J.-W. Zhai et al. / Molecular Phylogenetics and Evolution 77 (2014) 216–222

characteristics in orchid taxonomy. However, these features apparently have no critical significance in the Calanthe alliance. For instance, some species (e.g., C. tsoongiana and C. tricarinata) in sect. Tricarinata have no spur at the base of the labellum, some species (e.g., C. mannii and C. hancockii) have spurs less than 4 mm long, and some species (e.g., C. plantagenea and Calanthe sieboldii) have spurs more than 1 cm long. These traits are assumed to result from adaptation to pollinators from different habitats. In contrast, the existence of appendages on the labellum and their respective shapes and additional features, as well as the condition of having a two-lobed mid-lobe, has clear taxonomic value within the Calanthe alliance. The newly established sect. Alpinocalanthe is recognized by its labellum with hairy ridges and emarginated midlobe. In contrast, the new sect. Puberula does not have any appendages, and its mid-lobe is dentate or apically fringed. The newly ranged sect. Calanthe is characterized by labellum with warty calli and a mid-lobe that is two-lobed apically, while species of the new sect. Tricarinata are characterized by ridges or lamellae on the labellum and a mid-lobe with a single lobe on the apex. 5. Conclusions This study clarified, for the first time, the phylogenetic relationships among the three genera of the Calanthe alliance. Cephalantheropsis is confirmed as a monophyletic genus, while Phaius and Calanthe are both polyphyletic. A new genus, Paraphaius, is established containing the species P. flavus, P. columnaris, and P. takeoi. Subgenus Preptanthe and sect. Styloglossum of Calanthe are recognized as independent genera. C. delavayi is improperly classified under Phaius, and should be reconsidered to be a member of Calanthe with C. delavayi, and the new section, Alpinocalanthe, was established for this species only. C. calanthoides is confirmed as a member of Calanthe, and sect. Ghiesbrechtia is maintained under the same genus. Previous sect. Calanthe is demonstrated to be polyphyletic in our study and is divided into three sections, namely, Puberula, Calanthe, and Tricarinata, according to molecular and morphological characteristics. In conclusion, six genera are recognized in the Calanthe alliance, namely, Preptanthe, Phaius, Cephalantheropsis, Styloglossum, Paraphaius, and Calanthe, and three new sections are established in the newly defined Calanthe. 6. Taxonomic treatment

Key to the genera and sections of the Calanthe alliance 1. Flower bracts persistent 2. Labellum adnate to the column wing more or less..............................................................Calanthe 3. Labellum adnate to the column wing at the base, embracing the column.................................................sect. Alpinocalanthe 3. Labellum adnate to the column wings and forming a tube, lip flat or reflexed 4. Labellum disk without lamellae or calli.....................................................sect. Puberula 4. Labellum disk with lamellae or calli 5. Labellum entire with a prominent middle vein................................................sect. Ghiesbrechtia 5. Labellum usually 2-4-lobed 6. Labellum without warty appendages, mid-lobe entire.....................................................sect. Tricarinata 6. Labellum with warty calli, mid-lobe 2lobed....................................................sect. Calanthe

221

2. Labellum separate from the column wing, sometimes adnate to the column foot 7. Leaves deciduous in the dry season, inflorescence pubescent.........................................Preptanthe 7. Leaves persistent, inflorescence glabrous................................................Paraphaius 1. Flower bracts caducous 8. Labellum without spur or sac, vesiculate on mid-lobe or disk.........................Cephalantheropsis 8. Labellum with spur or sac, not vesiculate on mid-lobe or disk 9. Labellum separate from the column wing; flowers big...............................................Phaius 9. Labellum adnate to the column wings, forming a tube; flowers small..........................................................Styloglossum

1. Paraphaius J.W. Zhai et F.W. Xing, nov. gen. Type species: Paraphaius flavus (Blume) J.W. Zhai, Z.J. Liu et F.W. Xing Diagnosis: Flower bracts persistent; flower yellowish white, ridges on lip glabrous, spur conic, shorter than 0.8 cm; pollinia 8, in 2 groups, apart. Description: Pseudobulb stem-like, cylindric. Leaves on upper part of pseudobulb. Inflorescence arising from basal or lower nodes of pseudobulb; floral bracts persistent. Flowers not opening widely, yellow to mid-yellow. Sepals subsimilar, oblong-obovate, glabrous. Petals oblong-elliptic; lip 3-lobed; disk with 3 or 4 ridges; spur conic. Column densely white pubescent ventrally; pollinia 8, in 2 groups, apart. 1.1 Paraphaius flavus (Blume) J. W. Zhai, Z. J. Liu et F. W. Xing, nov. comb Basonym: Limodorum flavum Blume, Bijdr. 375. 1825. Synonym: P. flavus (Blume) Lindley, Gen. Sp. Orchid. Pl. 128. 1831. 1.2 Paraphaius takeoi (Hayata) J. W. Zhai, Z. J. Liu et F. W. Xing, nov. comb. Basonym: Calanthe takeoi Hayata, Icon. Pl. Formosan. 9: 111. 1920. Synonym: P. takeoi (Hayata) H. J. Su, Quart. J. Exp. Forest. Natl. Taiwan Univ. 3(4): 77. 1989. 1.3 Paraphaius columnaris (C. Z. Tang & S. J. Cheng) J. W. Zhai, Z. J. Liu et F. W. Xing, nov. comb. Basonym: P. columnaris C. Z. Tang & S. J. Cheng, Bull. Bot. Res., Harbin 5(2): 141. 1985. 2. Calanthe sect. Alpinocalanthe J. W. Zhai, Z. J. Liu et F. W. Xing, nov. sect. Type species: C. delavayi Finet. Diagnosis: Plant small; flower bracts persistent; labellum adnate to column wings at base, column slender; labellum slightly 3-lobed, surrounding column; labellum disk with three shortly hairy ridges. Monotypic species section 3. Calanthe sect. Puberula J.W. Zhai, Z.J. Liu et F. W. Xing, nov. sect. Type species: C. puberula Lindl. Diagnosis: Plant small; petal linear; labellum disk without lamellae or calli, mid-lobe dentate or fringed apically. 4. Calanthe sect. Tricarinata J. W. Zhai, Z. J. Liu et F. W. Xing, nov. sect.

222

J.-W. Zhai et al. / Molecular Phylogenetics and Evolution 77 (2014) 216–222

Type species: C. tricarinata Lindl. Diagnosis: Labellum 3-lobed, disk with mid-lobe without waxy lamellae or calli, not lobed. Containing approximately 10 species. Acknowledgments This study was funded by the National Natural Science Foundation of China (31370231) and the Special Fund for Forestry Scientific Research in the Public Interest - China (201204604). The authors would like to thank Huai-Zhen TIAN, Feng GUO, BingMou WANG, Ai-Qun HU, Xin-Sheng QIN, Yi-Bo LUO, Hong LIU, Chen REN, Zhi-Jian YIN, Ze-Jing MU, Xiao-Lang DU, Ce-Hong LI, Yu-Song HUANG, Yao-Hua HUANG, Ri-Hong JIANG, and Zong-Xin REN for collecting samples or making suggestions on the experimental design. The authors are also grateful to Yi-bo LUO, Peter Bernhardt, Pankaj Kumar, and Lizbeth Gale for assistance with the literature as well as to Xing WU and Lin Chen for growing plant materials. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ympev.2014.04. 005. References Bower, C.C., Brown, G.R., 2009. Pollinator specificity, cryptic species and geographical patterns in pollinator responses tosexually deceptive orchids in the genus Chiloglottis: the Chiloglottis gunnii complex. Aust. J. Bot. 57, 37–55. Breda, R., 1827. Genera et species Orchidearum et Asclepiadearum. t.7. Chen, S.C., Tsi, Z.H., Lang, K.Y., Zhu, G.H., 1999. Flora Reipublicae Popularis Sinicae 18. Science Press, Beijing. Chen, S.C., Liu, Z.J., Zhu, G.H., Lang, K.Y., Ji, Z.H., Luo, Y.B., Jin, X.H., Cribb, P.J., Wood, J.J., Gale, S.W., Ormerod, P., Vermeulen, J.J., Wood, H.P., Clayton, D., Bell, A., 2009. Orchidaceae. In: Wu, Z.Y., Raven, P.H., Hong, D. (Eds.), Flora of China, vol. 25. Science Press, Beijing & Missouri Botanical Garden Press, St. Louis, pp. 1–9. Chen, C.L., Liaoa, F.S., Cheng, S.H., 2010. Phylogenetic Analysis in the Genera Phaius and Cephalantheropsis Using rpl32-trnL Marker. In: Blanchard, M.G., et al., (Eds.). Proc. 1st Int’l Orchid Symposium, Acta Hort. 878, ISHS. Collins, F.H., Paskewitz, S.M., 1996. A review of the use of ribosomal DNA (rDNA) to differentiate among cryptic Anopheles species. Insect Mol. Biol. 5, 1–9. Thomas, P., Cribb, S., 1993. The genus Calanthe in tropical America. Orquidea (Mexico), 13: 227–232. Doyle, J.J., Doyle, J.L., 1987. A rapid isolation procedure from small quantities of fresh leaf tissue. Phytochem. Bull. 19, 11–15. Edgar, R.C., 2004. Muscle: multiple sequence alignment with high accuracy and high throughput. Nucl. Acids Res. 32 (5), 1792–1797.

Farris, J.S., Källersjö, M., Kluge, A., Bult, G.C., 1995. Constructing a significance test for incongruence. Syst. Biol. 44, 570–572. Fukuyama, N., 1935. Studia Orchidacearum Japonicarum. Bot. Mag. Tokyo 49, 269. Guillaumin, André, 1960. Bulletin du Muséum d’Histoire Naturelle sér. 2 32 (2), 188–189. Hamer, F., 1974. Orquideas de El Salvador 1, 90. Hebert, P.D.N., Penton, E.H., Burns, J.M., 2004. Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. PNAS 101 (41), 14812–14817. Holttum, R.E., 1947. New species of vascular plants from the Malayan Peninsula. Garden Bull. Singapore 11, 286. Huang, Y. et al., 2012. A newly recorded species of Phaius (Orchidaceae) from China. Ghihaia 13, 227–232. Karasawa, K., Ishida, G., 1998. Calanthe. Yasaka Shobo Inc, Japan. Kurzweil, H., 2010. A precursory study of the Calanthe group (Orchidaceae) in Thailand. Adansonia, sér. 32 (1), : 57–107. Lahaye, R., van der Bank, M., Bogarin, D., et al., 2008. DNA barcoding the floras of biodiversity hotspots. PNAS 105 (8), 2923–2928. Li, J.H., Liu, Z.J., Salazar, G.A., et al., 2011. Molecular phylogeny of Cypripedium (Orchidaceae: Cypripedioideae) inferred from multiple nuclear and chloroplast regions. Mol. Phylogenet. Evol. 61, 308–320. Liu, Z.J., Chen, L.J., Chen, S.C., Cai, J., Tsai, W.C., Hsiao, Y.Y., Ma, X.Y., Zhang, G.Q., 2011. Paraholcoglossum and Tsiorchis, two new orchid genera established by molecular and morphological analyses of the Holcoglossum alliance. PLoS ONE 6 (10), e24864. http://dx.doi.org/10.1371/joural.pone.0024864. Mike, T., Lena, S., Joachim, W.K., 1999. The phylogenetic relationships and evolution of the Canarian laurel forest endemic Ixanthusviscosus (Aiton) Griseb. (Gentianaceae): evidence from matK and ITS sequences, and floral morphology and anatomy. Plant Syst. Evol. 218, 299–317. Ormerod, P., 1998. A review of Cephalantheropsis. Orchid Digest 62, 155–159. Perner, H., Cribb, P.J., 2002. Orchid wealth. Alpine Gardener 70, 293. Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14, 817–818. Pridgeon, A.M., Cribb, P.J., Chase, M.W., Rasmussen, F.N., 2005. Genera Orchidacearum: Epidendroideae (part one), vol. 4. Oxford University Press Inc., New York, pp. 89–115. Reeves, G., Chase, M.W., Goldblatt, P., Rudall, P., Fay, M.F., et al., 2001. Molecular systematics of Iridaceae: evidence from four plastid DNA regions. Am. J. Bot. 88 (11), 2074–2087. Reichenbach, H.G., 1853. Flore des Serres et des Jardins de I’ Europe 8, 245–247. Richard, M.M.A., Galeotti, H., 1845. Annales des sciences naturelles botanique 3 (3), 28. Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19, 1572–1574. Schlechter, R., 1912. Die Orchidaceen von Deutsch-Neu-Guinea. Repertorium Species Novarum Regni Vegtabilis, Beihefte 1, 376–393. Sulaiman, S.F., Culham, A., Harborne, J.B., 2003. Molecular phylogeny of Fabaceae based on rbcL sequence data: with special emphasis on the tribe Mimoseae (Mimosoideae). Asia Pac. J. Mol. Biol. Biotechnol. 11 (1), 9–35. Swofford, D.L., 2002. PAUP⁄: Phylogenetic Analysis using Parsimony (⁄ and Other Methods), Version 4.0b10. Sinauer, Sunderland. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., et al., 2011. MEGA5:molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and Maximum Parsimony methods. Mol. Biol. Evol. http://dx.doi.org/10.1093/molbev/msr121.