View of Molecular Phylogenetics of Cypripedium L. (Cypripedioideae: Orchidaceae) Based on Plastid and Nuclear DNA Sequences

144 / J. Agrobiotech., 2, 111-118 (2011)

J. Agrobiotech. Vol 2, 2011, p. 35-51.

©Universiti Sultan Zainal Abidin ISSN 1985 5133

Fatihah N. H. N. et al.

Molecular Phylogenetics of Cypripedium L. (Cypripedioideae: Orchidaceae) Based on Plastid and Nuclear DNA Sequences

Molecular Phylogenetics of Cypripedium L.

(Cypripedioideae: Orchidaceae)

Based on Plastid and Nuclear DNA Sequences

*Fatihah N. H. N.^1,2,3, Fay M. F.² and Maxted N.³

1Department of Agricultural Science, Faculty of Agriculture and Biotechnology, Universiti Sultan Zainal Abidin, Kota Campus, Jalan Sultan Mahmud, 20400 Kuala Terengganu,

Terengganu Darul Iman, MALAYSIA.

fatihah@unisza.edu.my

2Jodrell laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3DS, UNITED KINGDOM.

3School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2TT, UNITED KINGDOM.

ABSTRACT

The genus Cypripedium L. (Cypripedioideae: Orchidaceae) was in a state of taxonomic chaos for almost a century after the name was first established by Linnaeus. The beauty and scarcity of the plants has encouraged many taxonomists and botanists to investigate the relationships within Cypripedium. However, the studies often conflicted with each other and contributed towards difficulty in identification. In this study, a molecular phylogenetic analysis has been conducted on the two plastid DNA (rbcL and matK) and the nuclear DNA (ITS) in order to test the monophyly and intrageneric relationships in Cypripedium. The plastid data was combined with the nuclear ITS to further resolve the relationships among 38 investigated taxa of Cypripedium. Parsimony and Bayesian analyses indicated that the genus is monophyletic and C. irapeanum is sister to the rest of Cypripedium taxa. The phylogenetic results presented provide reliable information of intrageneric relationships in Cypripedium to be applied for future conservation studies. The present study confirmed the distinctiveness of all Endangered (C. farreri, C. fasciolatum, C. formosanum, C. franchetii, C. lichiangense and C. margaritaceum) and Critically Endangered (C. fargesii, C. henryi, C. smithii, C. wumegense and C. yunnanense) species listed in the International Union for Conservation of Nature, IUCN Red List of Threatened Species Database, thus supporting their worthiness for protection and conservation attention.

Keywords: Conservation, Cypripedium, ITS, matK, molecular phylogenetic, rbcL

ABSTRAK

Taksonomi genus Cypripedium L. (Cypripedioideae: Orchidaceae) telah berada dalam keadaan tidak stabil untuk hampir seabad lamanya selepas nama genus ini ditubuhkan oleh Linnaeus. Keindahan dan kepupusan tumbuhan ini telah mempergiatkan ramai pakar taksonomi dan ahli-ahli botani membuat penyiasatan terhadap hubungan-hubungan dalam Cypripedium. Walau bagaimana pun, kajian-kajian tersebut sering bercanggah di antara satu sama lain yang menyumbang ke arah kesukaran dalam pengenalpastian. Dalam kajian ini, suatu analisis filogenetik molekul telah dilaksanakan ke atas dua jenis plastid DNA (rbcL dan matK) dan satu DNA nukleus (ITS) bertujuan menguji monofili dan hubungan-hubungan intragenus dalam Cypripedium. Data plastid telah digabungkan dengan nuklear ITS untuk seterusnya menyelesaikan hubungan-hubungan antara 38

146 / J. Agrobiotech., 2, 111-118 (2011)

taksa Cypripedium. Analisis Parsimoni dan Bayesian menunjukkan Cypripedium adalah monofiletik dan C. irapeanum adalah kakak bagi keseluruhan taksa Cypripedium yang diuji. Hasil filogenetik menyediakan maklumat jelas hubungan-hubungan intragenus dalam Cypripedium untuk diaplikasikan dalam subjek pemuliharaan pada masa akan datang. Kajian mengesahkan sifat tersendiri semua spesis Terancam (C. farreri, C. fasciolatum, C. formosanum, C. franchetii, C. lichiangense dan C. margaritaceum) dan Sangat Terancam (C. fargesii, C. henryi, C. smithii, C. wumegense dan C. yunnanense) adalah tersenarai dalam ‘International Union for Conservation of Nature, IUCN Red List of Threatened Species Database’, seterusnya menyokong keperluan perlindungan dan perhatian terhadap pemuliharaan bagi spesis-spesis tersebut.

Kata kunci: Pemuliharaan, Cypripedium, ITS, matK, filogenetik molekul, rbcL

INTRODUCTION

The genus Cypripedium L. (Cypripedioideae: Orchidaceae) consists of about 46 species distributed in the colder climates of the Northern Hemisphere, in North America, Europe and Asia, with a few extending to the tundra in Alaska and Siberia (Pridgeon et al., 1999). Cypripedium is locally known as lady’s slippers in Europe, moccasin flowers in North America and zapatillas in Latin America. It grows in a wide range of habitats from coniferous or mixed deciduous woodlands, to marshes and grasslands. The plants are recognized by their slipper-shaped pouch flowers (modified labellums), which are modified to attract pollinating insects by deceit (Dressler, 1993). Many Cypripedium spp. are extremely cold resistant, which can survive in the snow, blooming when the snow melts. However, in the wild, many have become rare and close to extinction, due to an ever shrinking natural habitat and over collection for gardens and herbarium specimens (Nelson, 1994; Farrell, 1999).

Cypripedium has been listed on the Convention on International Trade of Endangered Species of Wild Fauna and Flora, CITES Appendix II (McGough et al., 2006). This means that unlicensed trade of endangered species is forbidden and regulated by CITES controls. The majority of range countries have banned the export of wild-collected Cypripedium species, and therefore, it is bizarre to find legally wild-collected Cypripedium plants in international trade, but, these are usually plants of the common North American species that have come from controlled collection and salvage operations (McGough et al., 2006).

The slipper orchids have been used medicinally in North America. The most frequently cited species is C. parviflorum var. pubescens, the rhizome of which has been used as a remedy for several disorders including insomnia, anxiety, headache, neuralgia, emotional tension, palpitations, tremors, irritable bowel syndrome, delirium and convulsions due to fever. It also has been reported as being used to cure worms in children by the Cherokee Indians, and as a sedative in serious illness and childbirth by some native North Americans (Cribb, 1997). However, it generally has a sedative, anodyne or antispasmodic effect, thus care has to be taken when using it. Other taxa, such as C.

parviflorum var. parviflorum, C. acaule, C. reginae and C. candidum, have also been collected for this use (Cribb, 1997). Nevertheless, the horticultural interest in slipper orchids nowadays has tended to overshadow other aspects of their use. Initial achievements on Cypripedium spp. breeding have been made by the German Werner Frosch and the American Carson E. Whitlow, who have 39 and 11 hybrids registered by the Royal Horticultural Society in the United Kingdom, respectively (Perner, 1997; Whitlow, 2006).

The genus Cypripedium was in a state of taxonomic chaos for almost a century after the name was first established by Linnaeus. According to Cox et al. (1997), great interest in these plants has attracted widespread attention among taxonomists and botanists for well over a hundred years to study the phylogeny of the species for conservation. However, the studies often conflicted with each other and have resulted in numerous taxonomic treatments with conflicting relationships at the

36 / J. Agrobiotech. 2, 2011, p. 35 - 51.

genera and species levels, including C. calceolus L., which is one of the Britain’s rarest wild flowers.

The reduction in its population is reported near to extinction, with only one surviving clump found in the wild in northern England (Fay and Cowan, 2001). Due to its extreme rarity, the plant has been considered as an important target for conservation activities, such as site management, pollination, seed germination and more recently genetic finger printing (Fay et al., 2009). However, underpinning any conservation activities and sustainable uses of Cypripedium there must be knowledge of the intrageneric relationships to differentiate the species from its allies, so effective conservation strategies can be developed.

MATERIALS AND METHODS Plant Materials

A total of 36 species and two varieties from various localities were used in this study (Table 1). Total DNA samples of Cypripedium were obtained from the DNA bank in the Jodrell Laboratory, Royal Botanic Gardens, Kew, England. Five species, (Phragmipedium longifolium, Phragmipedium schlimii, Paphiopedilum sukhakulii, Paphiopedilum delenatii and Selenipedium chica) were designated as outgroups, by using sequences obtained from the NCBI GenBank.

37 / Fatihah N.H.N. et al.

38 / J. Agrobiotech. 2, 2011, p. 35 - 51.

39 / Fatihah N.H.N. et al.

150 / J. Agrobiotech., 2, 111-118 (2011)

followed by 30 cycles at 94 °C for 1 min, annealing temperature at 52 °C for 45 s, extension at 72 °C for 2.5 min (with an extra 8 s added in each cycle) and final extension at 72 °C for 7 min.

Cycle sequencing reactions were set up containing the forward and the reverse primers in separate mixtures. Sequencing primers were the same as those used for PCR. Automatic sequencing was performed in a DNA analyzer (ABI 3730 DNA Analyzer, Applied Biosystems Inc.) according to the manufacturer’s protocol.

Table 2. Primers used for DNA amplification

Region Primer Sequence

5’ – 3’ Sources

rbcL 1F ATG TCA CCA CAA ACA GAA AC Goldman et al., 2001.

rbcL 724Rm TCG CAT GTA CCY GCA GTT GC Goldman et al., 2001.

rbcL 636F GCG TTG GAG AGA TCG TTT CT Goldman et al., 2001.

rbcL 1360R CTT CAC AAG CAG CAG CTA GTT C Goldman et al., 2001.

matK -19F CGT TCT GAC CAT ATT GCA CTA TG Molvray et al., 2000.

matK 390F CGA TCT ATT CAT TCA ATA TTT Sun et al., 2001.

matK 1326R TCT AGC ACA CGA AAG TCG AAG T Sun et al., 2001.

Phylogenetic Analyses

Phylogenetic analyses were carried out using maximum parsimony (MP) and trees were generated using PAUP* version 4.0b10 (Swofford, 2002). This was done by performing a heuristic search using 1000 replicates with the factory settings of the bisection-reconnection (TBR) branch swapping, keeping only 10 trees per replicate. The data matrix was analyzed under the equal and unordered weight criterion of Fitch parsimony (Fitch, 1971). Gaps were treated as missing values. To test support for each clade, bootstrap analysis was performed with 1000 replicates of simple taxon addition and TBR swapping, with a limit of 10 trees kept per replicate. Bootstrap percentages of 50- 70 were considered weak, 71-85 as moderate and >85 as strong (Kress et al., 2002). For each region, individual analyses were performed. This was followed by a combined analysis of both plastid DNA alone and plastid and nuclear DNA data sets. All retained trees were rooted using Selenipedium chica as an out group because Selenipedium formed the earliest branching genus among the five genera of slipper orchids (Cox et al., 1997).

A model-based phylogenetic analysis of the combined matrix using Bayesian inference (BI) was performed as implemented in MrBayes version 3·1·2 (Ronquist and Huelsenbeck, 2003).

Markov Chain Monte Carlo searches were run twice for five milliongenerations, sampling one tree in every 1000 generations. A six-parameter model of molecular evolution with invgamma distribution for rbcL and matK, gamma for ITS, and a proportion of invariant sites fit best the data sets according to the Akaike information criterion (AIC) in MrModeltest2 (Nylander, 2004). The posterior probability support for the topology was compared with the topology and support of the most parsimonious tree(s) derived from the parsimony analyses. Usually, but not always, the posterior probabilities were greater than the bootstrap support values.

RESULTS

All samples were amplified without difficulty in most cases, except for few samples extracted from old herbarium specimens where the DNA was highly degraded.

Amplification of matK resulted in the largest fragment (1400 bp in length), followed by rbcL (1342 bp in length), and the shortest was ITS (987 bp in length). Among these three regions, the plastid rbcL was the least variable with no insertion/deletion events or ‘indels’ observed, so the

40 / J. Agrobiotech. 2, 2011, p. 35 - 51.

150 / J. Agrobiotech., 2, 111-118 (2011)

alignment was straightforward. The matK sequences possessed several point mutations and indels, of which the small indels were mostly found in multiples of three nucleotides and the largest, was up to 21 bp, and thus in most cases did not result in a frame shift. Even though ITS generated the shortest fragments, it showed the most variation, where several point mutations and indels ranging from 1 bp up to 32 bp were observed.

Table 3. Tree statistics for each of the individual regions, plastid and combined data sets.

Measurement rbcL matK Plastid ITS Combined

Length of alignment (bp) 1342 1400 2742 987 3729

Characters excluded (bp) - 88 29 130 160

Remaining characters (bp) 1342 1312 2713 857 3569

Constant characters 1267

(94.40%) 1090

(83.08%) 2409

(88.80%) 372

(49.24%) 2781 (77.93) Parsimony-uninformative

Characters

(2.61%) 35 136

(10.37%) 178

(6.56%) 122

(13.68%) 300 (8.40%) Parsimony-informative

characters

(2.98%) 40 86

(6.55%) 126

(4.64%) 363

(37.08%) 488 (13.67%)

Trees retained 70 158 372 900 2184

Trees length 101 309 425 1192 1631

Consistency index 0.79 0.82 0.80 0.60 0.65

Retention index 0.88 0.82 0.83 0.74 0.76

Number of nodes with >50-89%

bootstrap support

7 11 8 6 12

Number of nodes with >90%

bootstrap support

4 8 11 15 15

In combined analysis of plastid regions, many trees were retained but the trees did not show major differences, and were largely unresolved along the spine of the tree with low bootstrap support. The matK region seemed to perform similarly or slightly better than rbcL, but both provided few potentially parsimony informative characters. Therefore, the individual trees are not shown here, and only the overall pattern of combined parsimony and Bayesian analyses is described.

The aligned combined plastid regions matrix with all 38 taxa including five outgroup taxa was 2742 bp in length, of which 29 characters were excluded from the analysis. Of the resulting 2713 characters, 2409 were constant characters, 178 were parsimony-uninformative characters and 126 were potentially parsimony-informative characters. The analysis produced 393 trees with tree length of 425 steps, consistency index (CI) of 0.80 and retention index (RI) of 0.83. One of the most parsimonious trees found is shown with bootstrap percentage (BS) and posterior probabilities (PP) from Bayesian analysis in Figure 1.

41 / Fatihah N.H.N. et al.

Fig. 1. One of the most-parsimonious trees for the combined analysis of two plastid regions in Cypripedium. Numbers above the branches are Fitch lengths and numbers below the branches are

bootstrap percentage (red) followed by Bayesian posterior probability (blue). Branches without numbers received less than 50% bootstrap support

42 / J. Agrobiotech. 2, 2011, p. 35 - 51.

In this analysis, relationships were well resolved only in the out group and some of the terminal clade within the genus Cypripedium. Branch lengths along the spine of the tree were short, suggesting that they have been subjected to low levels of character changes during the evolutionary process. The weak nodes characterized either phylogenetically isolated species that collapsed into polytomies (C. acaule, C. fasciculatum, C. wardii and C. californicum) or character conflict among major groups distributed along the spine of the tree.

From the tree, the genus Cypripedium formed a monophyletic group with strong support (BS 99, PP 1), with the single accession of C. irapeanum strongly supported as sister to the rest of Cypripedium species. This was followed by C. debile and C. plectrochilum. The support from bootstrap percentage ranged from moderate to strong for these species. However, both received strong support from posterior probabilities (PP 1).

Within the genus Cypripedium, the monophyly of some sections had strong support from both bootstrap and posterior probabilities, such as section Flabellinervia, Trigonopedia, Obtusipetala and Cypripedium. Sixteen tested taxa of section Cypripedium were closely related to each other in a clade with strong bootstrap percentage (BS 97) and posterior probabilities (PP 1). Within the clade, the North-American subclade consisting of C. parviflorum var. parviflorum, C. parviflorum var. pubescens, C. kentuckiense, C. candidum and C. montanum was separated from C. calceolus and its closest Old World relatives (C. henryi, C. segawai and C. cordigerum), was well supported by bootstrap percentage (BS 85) and posterior probabilities (PP 1), confirming the relationships among these species. Cypripedium parviflorum var. parviflorum and C. candidum were grouped together, but they received extremely low support, and therefore the relationships in this clade were not certain.

The final alignment of the ITS matrix comprised 987 bp in length, of which 486 were constant characters, 135 were parsimony-uninformative characters and 366 were potentially parsimony-informative characters. Parsimony analysis of the ITS data for all 40 taxa resulted in 1040 most parsimonious trees with a tree length of 1214 steps, CI of 0.61 and RI of 0.74. The strict consensus tree of the 1000 equally most parsimonious trees from the nuclear ITS analysis is shown with bootstrap percentage in Figure 2.

In the ITS analysis alone, the relationships of Cypripedium were also largely unresolved along the spine of the tree, but most internal clades were strongly supported. A monophyletic clade that includes all species of Cypripedium was strongly defined with bootstrap support (BS 100). Cypripedium irapeanum was sister to the rest of the genus, but, there was no support for the position of C. acaule, C. fasciculatum, C. wardii, C. californicum and C. debile.

The North American subclade of section Cypripedium, C. parviflorum var. parviflorum, C.

kentuckiense, C. parviflorum var. pubescens and C. montanum from Canada and USA were closely related to each other, and the placement of C. candidum as sister to the group was strongly supported by bootstrap percentage (BS 99). However, the position of C. parviflorum var. parviflorum and C.

kentuckiense being grouped together with strong bootstrap support (BS 99) was somewhat unexpected, as C. parviflorum var. parviflorum and its ally C. parviflorum var. pubescens are generally treated as varieties of the same species, and therefore expected to be more closely related together than to any other species. This clade represents the major contradiction found between the ITS alone and combined plastid analysis.

43 / Fatihah N.H.N. et al.

Fig. 2. Strict consensus tree of the 1000 equally most parsimonious trees from the nuclear ITS analysis. Numbers above the branches are bootstrap percentage. Branches without numbers

received less than 50% bootstrap support 44 / J. Agrobiotech. 2, 2011, p. 35 - 51.

In the combined plastid (rbcL and matK) and nuclear (ITS) data analysis, a total of 43 taxa and 3569 characters were analyzed, 2781 of which were constant, 300 were parsimony-uninformative and 488 were potentially parsimony-informative. The combined analysis with all sequences recovered 1989 trees with tree length of 1631 steps, CI of 0.65 and RI of 0.76. One of the most parsimonious trees retained is shown in Figure 3.

The topology for the internal nodes was not changed much, but the bootstrap and posterior probabilities support were generally improved compared to the individual plastid or ITS analyses, although most nodes of the spine were still not well resolved and collapsed in the strict consensus tree. The monophyly of Cypripedium and the early branching of C. debile of section Retinervia was strongly supported by both bootstrap and posterior probabilities (BS 96, PP 1, BS 93, PP 1), respectively.

Cypripedium guttatum and C. yattabeanum were clustered together with strong support (BS 100, PP 1). Parsimony analysis showed that this clade was sister to the C. japonicum-C. formosanum-C.

californicum clade, but they received extremely low support suggesting that the analysis does not assure the placement of these clades. Cypripedium japonicum and C. formosanum were always closely related to each other and their relationships were highly supported (BS 100, PP 1). The position of C.

californicum nested to C. japonicum-C. formosanum clade was defined with extremely low confidence.

Cypripedium acaule was sister to C. wardii, but the place of these two species was not supported by either parsimony or Bayesian analyses. In section Obtusipetala, C. reginae was sister to C. flavum and C. passerinum with strong support (BS 100, PP 1). Within the Trigonopedia clade, the relationships among C. margaritaceum, C. daliense and C. lichiangense were strongly supported (BS 99, PP 1). The position of C. fargesii and then C. bardolphianum as sisters to this clade with strong support (BS 100, PP 1) was in concordance with ITS analysis.

Section Cypripedium was monophyletic with stronger support (BS 100, PP 1) than in plastid tree. The distinction of the North American subclade was highly supported (BS 100, PP 1).

Within this subclade, C. parviflorum var. parviflorum and C. kentuckiense were grouped together with strong support (BS 99, PP 1), confirming the relationships between these species. Cypripedium parviflorum var. pubescens and C. candidum were well related to C. parviflorum var. parviflorum and C.

kentuckiense, with C. montanum as sister to all of them (BS 100, PP 1). This result was in agreement with the plastid analysis.

45 / Fatihah N.H.N. et al.

Fig. 3. One of the most-parsimonious trees for the combined plastid and nuclear regions in Cypripedium. Numbers above the branches are Fitch lengths and numbers below the branches are bootstrap percentage (red) followed by Bayesian posterior probability (blue). Branches without

numbers received less than 50% bootstrap support 46 / J. Agrobiotech. 2, 2011, p. 35 - 51.

DISCUSSION The Sequences

In this analysis, two plastid DNA loci, rbcL and matK were used to infer the intrageneric relationships within Cypripedium. The dataset was combined with the nuclear internal transcribed spacer (ITS) data to obtain better resolved phylogenetic trees. This was achieved to a good extent with the combined dataset producing the largest number of supported (> 50%) clades (Table 3). Many of the clades were well resolved and highly supported by either bootstrap or posterior probabilities, or both.

Of the selected loci, ITS produced the highest number of potentially parsimony- informative characters (363), in comparison with matK (86) and rbcL (40). However, the performance (in terms of CI and RI values) of the ITS region was worse than the plastid sequences. According to van den Berg et al. (2009), this could be explained by the higher number of changes per variable position (i.e. more homoplasy) observed in the ITS dataset than in the plastid dataset. Secondly, it is more likely to be affected by taxon sampling. In most cases, when the number of taxa increases, the CI and RI values are observed to decrease, irrespective of any change in information content (Kitching et al., 2003).

The matK gene produced more than twice the number of potentially parsimony- informative characters (86) when compared with rbcL (40). This result is congruent with previous studies where these plastid DNA regions were employed (Shaw et al., 2005; Asmussen et al., 2006).

As expected, the rbcL sequences produced the fewest number of potentially parsimony-informative characters (40) and resolved clades (11), whereas matK and ITS generated 19 and 21 resolved clades, respectively. Despite resulting in higher CI and RI values (0.80 and 0.83), the number of resolved clades with more than 90% support in the plastid dataset was relatively low (11), and the combined dataset produced the highest number of highly resolved clades with more parsimony-informative characters (488).

Phylogenetic Relationships of Cypripedium Species

The monophyly of Cypripedium was strongly supported in all datasets; however, to some extent, the relationships within the genus differ due to different sampling between the datasets. In this study, sequence variation in the ITS dataset alone was insufficient to resolve all the phylogenetic relationships, especially among the early branching of Cypripedium. One of the major conflicts drawn from the ITS dataset is that, the divergence of C. debile after C. irapeanum received extremely low support from parsimony analysis, and the branch to C. debile collapses in the strict consensus. This result was also confirmed by other studies in which this region was used (Kahandawala et al., unpublished data). However, the placements of C. debile just after C. irapeanum is clearly defined and supported in both plastid and combined analyses.

Cypripedium irapeanum appeared as sister to all other Cypripedium species, and this was supported by other molecular phylogenetic studies (Cox et al., 1997; Cook, 2003; Kahandawala et al., unpub. data). This taxon has been said to be the most ‘primitive’ among the others as some of its morphological features are similar to the earliest branching genus of slipper orchids, Selenipedium (Cox et al., 1997) even though Chen and Lang (1986) suggested that C. subtropicum shared many common features with Selenipedium.

No support was found for the branches of C. acaule (section Acaulia), C. californicum (section Irapeana) and C. fasciculatum (section Enantiopedilum). The positions of these species were somewhat different in other molecular phylogenetic trees; for example in Cribb (1997), C. californicum was grouped with C. irapeanum, thus supporting the morphological analysis. However, Cox et al. (1997) demonstrated that C. californicum was sister to C. formosanum in section Flabellinervia, together with C.

guttatum and C. yatabeanum in section Bifolia, although the support was weak. In this plastid analysis, C. californicum was sister to all species in section Cypripedium, whereas in the combined analysis, it 47 / Fatihah N.H.N. et al.

158 / J. Agrobiotech., 2, 111-118 (2011)

appeared to be more closely related to the species pair C. japonicum and C. formosanum in section Flabellinervia. Nonetheless, the sequence variation in both analyses is still insufficient to define the relationships of those species with confidence. The odd placement of species in phylogenetic trees is possibly an artefact of long-branch attraction (LBA) due to accumulation of homoplasy on long non- sister branches that make them appear to be sister clade (Albert, 2005). This phenomenon is reportedly susceptible in parsimony methods (Felsenstein, 1978; Hillis and Huelsenbeck, 1992;

Kuhner and Felsenstein, 1994). Sampling more taxa and characters especially of other loci might help to break up the long branches, but there are pros and cons of these as discussed in Aguinaldo et al.

(1997). In addition, lack of support was also observed for the divergence of branches along the spine of the tree. Less resolution suggests the possibility of either a rapid radiation of the major lineages or a slower rate of molecular evolution (Malcomber, 2002; Salazar et al., 2009). The rapid radiation of Cypripedium is counter to expectation based on its wide morphological variation and geographical range.

Monophyly of the Obtusipetala clade was highly supported; however, the plastid and combined analyses explained the relationships among the species in different ways. In the plastid analysis, C. reginae and C. passerinum were sisters, and both were sister to C. flavum, which was in congruent with the result of the rDNA ITS analysis by Cox et al. (1997). However, in the ITS and combined analyses, C. flavum and C. passerinum were grouped together. Morphological analyses support C. flavum and C. reginae together (Cribb, 1997; Cribb and Sandison, 1998), probably because both display similar shaped flowers. However, for better understanding of their relationships, further investigation with more characters of different loci or incorporated morphological dataset is recommended.

In section Trigonopedia, all five tested taxa (C. margaritaceum, C. lichiangense, C. daliense, C.

fargesii and C. bardolphianum) were clustered together with strong support in both parsimony and Bayesian analyses, as suggested by Cribb and Sandison (1998) who focused on morphological characters. Although C. fargesii was thought to be the closest ally of C. margaritaceum (Cribb, 1997), and has been treated as a variety of C. margaritaceum, this study revealed that C. margaritaceum was in fact more closely related to C. lichiangense than to C. fargesii with moderate support, and was confirmed by Cox et al. (1997) with higher bootstrap support (BS 100). Cypripedium daliense, first introduced as a synonym of C. margaritaceum by Chen and Wu in 1991 was probably a misidentification (Cribb, 1997), but there was no strong support to verify the status of C. daliense as it could be a closely related taxa to both C. margaritaceum and C. lichiangense or either one as they are all appeared together in an unresolved clade in both ITS and combined analyses.

The section Cypripedium clade in the combined dataset was always well defined with higher/slightly higher bootstrap and posterior probabilities support when compared with plastid alone. The North American subclade, C. parviflorum var. parviflorum and C. parviflorum var. pubescens were not clustered together, although they were often treated as varieties of the same species (Cribb, 1997; Anon, 2009). According to Cribb (1997), the status of these orchids has been under debate since they were first described as varieties, being raised to specific rank in 1929, but then reduced to varietal rank within C. calceolus in 1940. Recently, Sheviak (1992; 1994) provided morphological evidence and treated both taxa as varieties. Considering the highly supported relationships of C.

parviflorum var. parviflorum-C. kentuckiense in both ITS and the combined analyses, C. parviflorum var.

parviflorum and C. parviflorum var. pubescens should probably be treated as two different species. In this case, incorporating multiple sampling and more molecular loci might help elucidate patterns of relationships among the tested taxa more clearly.

The association between C. calceolus and C. shanxiense as sister taxa lacked support from morphological data. Cypripedium calceolus was grouped with C. parviflorum and C. montanum into one clade, but the relationships were not well resolved (Cribb, 1997; Cribb and Sandison, 1998).

Cypripedium shanxiense was difficult to find in the wild because of its dull color, and consequently it has been under-collected (Cribb, 1997). To date, this species was not included in previous molecular studies (Cox, 1995; Cox et al., 1997); therefore the relationship to C. calceolus needs further investigation.

48 / J. Agrobiotech. 2, 2011, p. 35 - 51.

Implications for Conservation and Reintroduction

As much of the population decline in the wild is due to over collection and habitat destruction, serious situations arise when setting priorities for conservation. In this study, the most important implication for conservation that has resulted from the phylogenetic analysis is not concerned with each Cypripedium species as a separate entity, but rather the fact that the species should be conserved together as a group or an evolutionary significant unit (ESU) (Moritz, 1994; Cracraft, 1997; Waples, 1998) to assure minimal loss of genetic diversity. For example, C. kentuckiense, which was recently described as a rare orchid in southern regions of the USA displays high genetic and morphological similarity with its ally C. parviflorum var. parviflorum (Case et al., 1998). As their relationships are strongly supported, we think that these two species should be conserved together. Although manipulation of C. kentuckiense populations may endanger them, efficient management techniques could be developed for C. parviflorum var. parviflorum, which is apparently more common and less vulnerable to extinction (Case et al., 1998). The techniques could then be evaluated for transfer to C.

kentuckiense populations. Their closest relatives (C. parviflorum var. pubescens, C. candidum and C.

montanum), are also worthy of conservation because they may carry some genetic markers that can be found in rare species or absent in the other groups. Therefore, as many as species grouped with species that have become rare in the wild (with strong support from phylogenetic analysis) have some conservation value and warrant protection.

Misidentification is more likely to occur when one considers morphological data alone in classification. One such example is reintroduction of one of Britain’s rarest wild flower, C. calceolus, to the wild. This species has similar morphological characters with C. parviflorum which made them difficult to identify, and for that reason, C. parviflorum was mistakenly incorporated in the UK conservation programmes (Fay et al., 2009). However, the distinction of C. parviflorum has been identified by this study and other molecular studies (Cox, 1995; Cribb, 1997; Cox et al., 1997), and this allowed the introduced plant to be identified and excluded from the programme. The ITS and combined trees generated from this study show that the association between C. calceolus and C.

shanxiense is strongly supported by bootstrap and posterior probabilities, confirming the relationships between these two species; therefore, C. shanxiense could be a good candidate to be considered for conservation along with C. calceolus.

CONCLUSIONS

This study indicates that the combination of certain regions of the plastid genome (rbcL and matK) and the nuclear genome (ITS) has the potential to reinforce previous hypotheses of the monophyly of the genus Cypripedium and to provide insight into the intrageneric relationships of Cypripedium.

Many relationships were supported by both parsimony and Bayesian analyses in the combined dataset, and even strengthen those uncovered with individual or plastid datasets alone. Lack of support for the divergence of branches along the spine of the tree suggests the possibility of either a rapid radiation of lineage divergences or a slower rate of molecular evolution (Malcomber, 2002;

Salazar et al., 2009). The rapid radiation of Cypripedium is contrary to expectation based on its marked morphological variation and wide geographical range. However, issues regarding incomplete sampling (of data and/or taxa) should be taken into account as these can cloud the interpretation of phylogeny.

ACKNOWLEDGEMENTS

We would like to thank the Royal Botanic Gardens, Kew for providing plant materials and the sequencing facilities.

49 / Fatihah N.H.N. et al.

160 / J. Agrobiotech., 2, 111-118 (2011)

REFERENCES

Aguinaldo, A. M. A, Turbeville, J. M., Linford, L. S., Rivera, M. C., Garey, J. R., Raff, R. A., Lake, J.

A. 1997. Evidence for a clade of nematodes, arthropods and other moulting animals. Nature 387: 489-493.

Albert, V. A. 2005. Parsimony phylogeny and genomics. Oxford University Press, New York. 61 pp.

Anon. 2009. World Checklist of Monocotyledons. http://apps.kew.org/wcsp/qsearch.do. Retrieved on 14^th August 2009.

Asmussen, C. B., Dransfield, J., Deickmann, V., Barfod, A. S., Pintaud, J-C. & Baker, W. J. 2006. A new subfamily classification of the palm family (Aracaceae): Evidence from plastid DNA phylogeny. Botanical Journal of the Linnaean Society 151: 15-38.

Case, M. A., Mlodozeniec, H. T., Wallace, L. E. & Weldy, T. W. 1998. Conservation genetics and taxonomic status of the rare Kentucky lady’s slipper: Cypripedium kentuckiense (Orchidaceae).

American Journal of Botany 85(12): 1779-1786.

Chen, S. C. & Lang, K. Y. 1986. Cypripedium subtropicum, a new species related to Selenipedium. Acta Phytotax. Sin. 24(4): 317-322.

Cook, P. 2003. Systematics, Genome Size and Plastid Microsatellite Profiling in Cypripedium with a Focus on Cypripedium calceolus. Jodrell Laboratory, Royal Botanic Gardens, Kew. 521 pp.

Cox, A. 1995. Molecular Systematics of Cypripedium: Phylogenetic Utility of the 5s rDNA Spacer.

PhD Thesis, University of Reading, UK.

Cox, A. V., Pridgeon, A. M., Albert, V. A. & Chase, M. W. 1997. Phylogenetics of the slipper orchids (Cypripedioideae, Orchidaceae): Nuclear rDNA ITS sequences. Pl. Syst. Evol. 208: 197-223.

Cracraft, J. 1997. Species concepts in systematic and conservation biology: An ornithological viewpoint. In Species: The Units of Biological Diversity. M. F. Claridge, H. A. Dawah and M. R.

Wilson (eds.). Systematics Association Special Volume Series 54. Chapman & Hall, London. p. 325-339.

Cribb, P. 1997. The genus Cypripedium. Timber Press, Portland. 294 pp.

Cribb, P. & Sandison, M. S. 1998. A preliminary assessment of the conservation status of Cypripedium species in the wild. Botanical Journal of the Linnean Society 126: 183-190.

Dressler, R. L. 1993. Phylogeny and the Classification of the Orchid Family. Cambridge University Press.

Farrell, L. 1999. Cypripedium calceolus L. (Orchidaceae): Lady’s slipper orchid. In British Red Data Books.

1. Vascular Plants. M. J. Wigginton (ed.). Peterborough: Joint Nature Conservancy Council, 114 pp.

Fay, M. F., Bone, R., Cook, P., Kahandawala, I., Greensmith, J., Harris, S., Pedersen, H. Ǽ., Ingrouille, M. J. & Lexer, C. 2009. Genetic diversity in Cypripedium calceolus (Orchidaceae) with a focus on north-western Europe, as revealed by plastid DNA length polymorphisms.

Annals of Botany 104: 517-525.

Fay, M. F. & Cowan, R. S. 2001. Plastid microsatellites in Cypripedium calceolus (Orchidaceae): Genetic fingerprints from herbarium specimens. Lindleyana 16(3): 151-156.

Felsenstein, J. 1978. Cases in which parsimony and compatibility methods will be positively misleading. Systematic Zoology 27: 401-410.

Fitch, W. M. 1971. Toward defining the course of evolution: Minimum change for a specific tree topology. Systematic Zoology 20: 406-416.

Goldman, D. H., Freudenstein, J. V., Kores, P. J., Molvray, M., Jarrel, D. C., Whitten, W. M., Cameron, K. M., Jansen, R. K.. & Chase, M. W. 2001. Phylogenetics of Arethuseae (Orchidaceae) based on plastid matK and rbcL sequences. Systematic Botany 26: 670-695.

Hillis, D. & Huelsenbeck, J. 1992. Signal, noise, and reliability in molecular phylogenetic analyses. J.

Hered. 83: 189-195.

Kitching, I. J., Forey, P. L., Humphries, C. J. & Williams, D. M. 2003. Cladistics - The Theory and Practice of Parsimony Analysis, 2^nd Edition. Oxford University Press Inc., New York, p. 96-99.

50 / J. Agrobiotech. 2, 2011, p. 35 - 51.

Kress, W. J., Prince, L. M. & William, K. J. 2002. The phylogeny and a new classification of the gingers (Zingiberaceae): Evidence from molecular data. American Journal of Botany 89: 1682- 1696.

Kuhner, M. & Felsenstein, J. 1994. A simulation comparison of phylogeny algorithms under equal and unequal evolutionary rates. Mol. Biol. Evol. 11: 459-468.

Malcomber, S. T. 2002. Phylogeny of Gaertnera Lam. (Rubiaceae) based on multiple DNA markers:

Evidence of a rapid radiation in a widespread, morphologically diverse genus. Evolution 56(1): 42-57.

McGough, H. N., Roberts, D. L., Brodie, C. & Kowalczyk, J. 2006. CITES and slipper orchids: An introduction to slipper orchids covered by the Convention on Internal Trade in Endangered Species. The Board of Trustees, Royal Botanic Gardens, Kew.

Molvray, M., Kores, P. J. & Chase, M. W. 2000. Polyphyly of mycoheterotrophic orchids and functional influences on floral and molecular characters. In Monocots: Systematic and Evolution.

K. L. Wilson and D. A. Morrison (eds.). Melbourne: CSIRO Publishing. p. 441-448.

Moritz, C. 1994. Defining evolutionarily significant units for conservation. Trends in Ecology and Evolution 9: 373-375.

Nelson, E. C. 1994. Historical data from specimens in the herbarium, National Botanic Gardens, Glasnevin, Dublin (DBN), especially on Cypripedium calceolus. BSBI News 67: 21-22.

Nylander, J. A. A. 2004. MrModeltest2 (Program distributed by the author). Evolutionary Biology Centre, Uppsala University.

Perner, H. 1997. Cultivation in the Genus Cypripedium. P. Cribb (ed.). Timber Press, Portland. p. 60-96.

Pridgeon, A. M., Cribb, P. J., Chase, M. W. & Rasmussen, F. N. 1999. Genera Orchidacearum: General introduction, Apostosioideae, Cypripedioideae. Oxford University Press, Great Clarendon Street, Oxford, UK. 195 pp.

Ronquist, F. & Huelsenbeck, J. P. 2003. MRBAYES 3·1·2: Bayesian Phylogenetic inference under mixed models. Bioinformatics 19: 1572-1574.

Salazar, G. A, Cabrera, L. I., Madrinan, S. & Chase, W. M. 2009. Phylogenetic relationships of Cranichidinae and Prescottiinae (Orchidaceae, Cranichideae) inferred from plastid and nuclear DNA sequences. Annals of Botany 104: 403-416.

Shaw, J., Lickey, E., Beck, J. T., Farmer, S. B., Liu, W., Miller, J., Siripun, K. C., Winder, C. T., Schilling, E. E. & Small, R. L. 2005. The tortoise and the hare II: Relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany 92: 142-166.

Sheviak, C. J. 1992. Natural hybridization between Cypripedium montanum and its yellow-flowered relatives. American Orchid Society Bulletin 61: 546-559.

Sheviak, C. J. 1994. Cypripedium parviflorum Salisb. I: The small-flowered varieties. American Orchid Society Bulletin 63(3): 664-669.

Sun, H., McLewin, W. & Fay, M. F. 2001. Molecular phylogeny of Helleborus (Ranunculaceae), with an emphasis on the East Asia-Mediterranean disjunction. Taxon 50: 1001-1018.

Swofford, D. L. 2002. PAUP* - Phylogenetic Analysis Using Parsimony (*and Other Methods) Version 4.0b10. Sunderland, MA: Sinauer Associates.

Van den Berg, C., Higgins, W. E., Dressler, R. L., Whitten, W. M., Soto-Arenas, M. A. & Chase, M.

W. 2009. A phylogenetic study of Laeliinae (Orchidaceae) based on combined nuclear and plastid DNA sequences. Annals of Botany 104: 417-430.

Waples, R. S. 1998. Evolutionary significant units, distinct population segments and the endangered species act: Reply to Penncock and Dimmick. Conservation Biology 12: 718-721.

Whitlow, C. E. 2006. Registered Cypripedium hybrids. http://www.c-we.com/cyp.

haven/cyp1tbl.htm. Accessed on 10^th August 2009.

51 / Fatihah N.H.N. et al.