Botanical Studies (2009) 50: 477-485.

SYSTEMATICS

The confirmation of Camellia formosensis (Theaceae) as an independent species based on DNA sequence analyses

Mong-Huai SU¹, Chang-Fu HSIEH², and Chih-Hua TSOU³ *

¹Department of Forestry and Nature Conservation, Chinese Culture University, Taipei 111, Taiwan
²Institute of Ecology and Evolutionary Biology, National Taiwan University, Taipei 106, Taiwan
³Institute of Plant and Microbial Biology, Academia Simica, Taipei 115, Taiwan

(Received July 16, 2008; Accepted April 13, 2009)

ABSTRACT. The endemic 'Taiwanese wild tea' represents an invaluable natural resource for the tea industry. The taxon is distributed in the central, southern and eastern regions of Taiwan with populations shrinking dramatically in the past decade. Taxonomically, it has long been treated as a lower taxon either under Camellia sinensis var. sinensis or C. sinensis var. assamica. Nevertheless, our recent morphological study indicated that this taxon is different from the other two taxa. In this report, we provide molecular evidence for reassessing its systematic status. Samples representing all the wild populations that are available in Taiwan were sequenced along with samples from the other two taxa at the introns 12-16 and 23 of the RPB2 gene of nuclear DNA. Phylogenetic analysis recovered a monophyletic group of 'Taiwanese wild tea' well separated from the other taxa. In addition, genetic distances and previous phenetic analyses all supported the identity of this species. A new combination of C. formosensis (Masamune et Suzuki) M. H. Su, C. F. Hsieh et C. H. Tsou is proposed herein.

Keywords: Camellia formosensis; Molecular taxonomy; RPB2 gene; Section Thea; Theaceae; Taiwanese wild tea.

INTRODUCTION

'Taiwanese wild tea' has been called the "tea of the gods" by indigenous people. The first written record of this plant was made by Chou (1717). This endemic Camellia taxon is an element of broad-leaved forests in the central, southern and eastern regions of Taiwan (Figure 1). Due to its special flavor, 'Taiwanese wild tea' has become an important resource in Taiwan's tea industry. A black tea strain, TTES no. 18, was developed via hybridization between the Taiwanese wild tea as the paternal plant and a cultivar from Burma as the maternal plant and it has become a popular tea item. Despite the economic potential, wild populations of Taiwanese wild tea have drastically declined due to human disturbance. Conservation of the endemic tea is, therefore, urgently required and for this purpose, taxonomical information is essential (Mace, 2004).

Various names have been given to Taiwanese wild tea, indicating the taxonomic difficulties. It was first named Thea formosensis Masamune et Suzuki (Suzuki, 1937), and later on it was referred to as Camellia sinensis (L.) O. Kuntze forma formosensis Kitamura (Kitamura, 1950;

*Corresponding authors: E-mail: chtsou@gate.sinica.edu.tw.

Hsieh et al., 1996) and C. sinensis subsp. buisanensis (Sasaki) Lu & Yang (1987). It was even merged with C. sinensis var. sinensis (Ming, 2000). Such taxonomic uncertainty mostly stems from the high morphological similarities existing among these three taxa. In our previous morphological study, however, the results showed that the Taiwanese wild tea can be well differentiated from other native and cultivated tea taxa by its glabrous ovaries and winter buds (Su et al., 2007). Lai et al. (2001) assessed the genetic relationships of tea cultivars, including the Taiwanese wild tea, based on RAPD and ISSR fingerprinting. All the Taiwanese wild tea samples were clustered together in the phylogenetic trees. Another study using RAPD and AFLP fingerprints also reached the same conclusion (Wachira et al., 2001) though only two samples of the Taiwanese wild tea were included. These studies provide the basis of the tea systematics. Nevertheless, samples of Taiwanese wild tea were too few and plants from its eastern population were not included in either of the phylogenetic studies; meanwhile the inter-relationships among central, eastern and western populations seemed complicated (Su et al., 2007).

Molecular phylogeny based on nucleotide sequences of the RPB2 gene, a single-copy gene (Denton et al., 1998) encoding the second largest subunit of RNA polymerase,

478

Botanical Studies, Vol. 50, 2009

outgroups. Leaf materials were preserved in silica gel bags immediately after sampling. Voucher specimens were deposited at the herbaria of TAI or HAST.

DNA extraction, PCR and sequencing

Total genomic DNA was extracted from dried leaf tissue, following the protocol of Vijayan and Tsou (2008). Primers used for PCR amplification and sequencing of

the RPB2 introns 12-16 and 23 are listed in Table 2. PCR

was performed in a 50 fiL final volume containing 100 ng DNA, 0.2-0.6 foM of both primers, 200 fiM of each dNTP, 0.5 U Taq polymerase and 1X buffer (Viogene, Inc.). The PCR cycle was with an initial 5 min denaturation at 95^oC, followed by 35 cycles of 1 min denaturation at 95^oC, 1 min 20 sec annealing at 58^oC to 63^oC, 1 min 20 sec extension at 72^oC, and with a final extension of 7 min at 72 。C. PCR products were purified using QIAquick® PCR Purification Kit (Qiagen Gmbh, Inc.). Automatic sequencing was conducted with an ABI PRISM® 3700 DNA Sequencer.

Data analysis

Sequences were assembled from both directions. Nucleotide sequences of introns 12-16 and 23 were aligned with CLUSTAL_X v1.83 (Thompson et al., 1997). The phylogenetic analyses were performed with the maximum parsimony (MP) and neighbor joining (NJ) methods using PAUP v0.4b10 (Swofford, 2001). In the MP analysis, the most strict consensus tree was obtained by performing a branch-and-bound searching. Indels were treated as single base changes and gaps were considered as the fifth base. In the NJ analysis, distances were measured with an 'uncorrected ("p")' index. Indels were also treated as a single base change but gaps were considered as missing. Branch supports of both phylogenetic trees were obtained by 1000 replicates of bootstrapping. Additionally, the genetic distances between taxa were calculated with the 'uncorrected ("p")' index under the distance criterion of

PAUP.

RESULTS

Introns 12-16 of RPB2 of samples examined were 954 bp long. No indels were detected. The intron 23 was 992 to 1001 bp long in all the samples except for two samples of C. sinensis var. assamica, i.e., A-Thai-2 and A-Ind-3, in which a long insertion of 284 bp was found. In the analysis, the long insertion was treated as a single mutation and a data matrix of 1923 bp for 35 OTUs was obtained.

Of 1923 bp sequences, 166 bp were variable and 75 bp were parsimoniously informative. Branch-and-bound search found 1896 equally parsimonious trees. The most strict consensus tree was identified with 233 steps. A high consistency index (CI) value 0.92 and low homoplasy index (HI) value 0.25 indicated low levels of homoplasy. MP and NJ trees were the same in topology though the branch supports might be different (Figure 2). All 13 samples of the Taiwanese wild tea formed a distinct

Figure 1. Distribution and sampling sites of Camellia formosensis. Solid circles: distribution sites based on specimens of TAI, TAIF, HAST and TNM herbaria. DNA sampling sites are marked with the running number. NT, Nantou County; YL, Yunlin County; CY, Chiayi County; KS, Kaohsiung County; PT, Pingtung County; TT, Taitung County.

provided much insight into the phylogeny of Camellia in an earlier studies (Xiao, 2001; Xiao and Parks, 2003); however, no materials from the Taiwanese wild tea were included. We thus reconstructed the phylogeny of the Taiwanese wild tea and its allies by using the same DNA segments, the RPB2 introns 12-16 and 23, and with a much more extensive sampling. The results are here documented.

MATERIALS AND METHODS

Samples

Thirty-five samples were included in the study, including 13 samples of the Taiwanese wild tea, eight samples of C. sinensis var. sinensis and 11 samples of C. sinensis var. assamica, all belonging to the section Thea (Table 1). Among the 13 samples of the Taiwanese wild tea, 11 were collected from wild populations (Figure 1) while the other two were obtained from the tea germplasm garden (Table 1). Three species―C. furfuracea (sect. Furfuracea), C. transarisanensis (sect. Theopsis), and C. brevistyla (sect. Paracamellia)——were chosen as

SU et al. ― Taxonomy of Camellia formosensis 479

Table 1. List of samples and accession numbers.


Sample code	Voucher no.	Location	Habitat^a	GenBank accession no.
Sample code	Voucher no.	Location	Habitat^a	Introns 12-16	Intron 23
C. formosensis
F-NT-1	Tsou 2133	Meiyuanshan, Taiwan	W	EU849031	EU849066
F-NT-2	Tsou 2135	Meiyuanshan, Taiwan	W	EU849032	EU849067
F-NT-3	Su 683	Derhuasia, Taiwan	G	EU849029	EU849064
F-NT-4	Su 687	Fenhuanshan, Taiwan	G	EU849030	EU849065
F-CY-1	Su 642	Sitin, Taiwan	W	EU849024	EU849059
F-KS-1	Tang 606	Yuyuoshan, Taiwan	W	EU849033	EU849068
F-KS-2	Su 269	Sasi Logging Road, Taiwan	W	EU849021	EU849056
F-PT-1	Su 645	Jenlishan, Taiwan	W	EU849025	EU849060
F-PT-2	Su 647	Jenlishan, Taiwan	W	EU849026	EU849061
F-PT-3	Su 498	Wuweishan, Taiwan	W	EU849022	EU849057
F-PT-4	Su 497	Wuweishan, Taiwan	W	EU849023	EU849058
F-TT-1	Su 655	Yungkangshan, Taiwan	W	EU849028	EU849063
F-TT-2	Su 654	Yungkangshan, Taiwan	W	EU849027	EU849062
C. sinensis var. assamica
A-Bur	Su 669	Burma	G	EU849037	EU849072
A-Chn-1	Wang s. n.	Yunnan, China	C	EU849042	EU849077
A-Chn-2	Wang 7740	Zhejiang, China	W	EU849043	EU849078
A-Ind-1	Su 667	Manipur, India	G	EU849036	EU849071
A-Ind-2	Su 670	Assam, India	G	EU849038	EU849073
A-Ind-3	Fatima s. n.	Kerala, India	C	EU849044	EU849079
A-Sri	Su 685	Sri Lanka	G	EU849040	EU849075
A-Thai-1	Su 684	Thailand	G	EU849039	EU849074
A-Thai-2	Maxwell 04-212	Lampang, Thailand	C	EU849041	EU849076
A-Tw-1	Su 609	Joefenershan, Taiwan	N	EU849034	EU849069
A-Tw-2	Su 610	Joefenershan, Taiwan	N	EU849035	EU849070
C. sinensis var. sinensis
S-Chn-1	Su 668	Fujian, China	G	EU849048	EU849083
S-Chn-2	Tsou 1957	Guangdong, China	W	EU849050	EU849085
S-Chn-3	Tsou 2052	Fujian, China	W	EU849051	EU849086
S-Chn-4	Tsou 2127	Fujian, China	W	EU849052	EU849087
S-Tw-1	Su 197	Jiantziliao, Taiwan	N	EU849045	EU849080
S-Tw-2	Su 640	Shanmei, Taiwan	C	EU849046	EU849081
S-Tw-3	Su 644	Fonshan, Taiwan	C	EU849047	EU849082
S-Tw-4	Su s. n.	Sijhih, Taiwan	C	EU849049	EU849084
C. brevistyla	Su 681	Alishan, Taiwan	W	EU849055	EU849090
C. furfuracea	Su 217	Lienhuachi, Taiwan	W	EU849053	EU849088
C. transarisanensis	Su 515	Yuanchueishan, Taiwan	W	EU849054	EU849089

^aHabitat: C, cultivated; G, germplasm garden of Yuchih Branch, Taiwan Tea Experiment Station, Taiwan; N, naturalized; W, wild. Samples of C. formosensis were indicated in Figure 1 with the running numbers.

480

Botanical Studies, Vol. 50, 2009

clade with high bootstrap values (92 and 99) (Figure 2). Thus, the monophyly of the Taiwanese wild tea was strongly supported. Within this clade, two subclades were recovered that were significantly supported. Samples from southern Taiwan, hereafter defined as the South population, clustered together. Samples from regions of Nantou (defined as the Nantou population) and Taitung (defined as the East population) formed another subclade. In contrast to the monophyly of the Taiwanese wild tea, the 19 samples of C. sinensis var. assamica and C. sinensis var. sinensis were polytomous in the cladogram (Figure 2). Since the latter two taxa have a long history of cultivation and hybridization within the tea industry, introgressions between them were expected. Meanwhile, the formation of a distinct clade by all the samples of the Taiwanese wild tea strongly demonstrated that, genetically, the Taiwanese wild tea was well separated from the other two taxa.

The averages of pairwise genetic distances between species and between con-specific populations were shown in Table 3. The average distance between samples of the Taiwanese wild tea and C. sinensis var. assamica was 0.0092, and that between the Taiwanese wild tea and C. sinensis var. sinensis was 0.0079. Both distances were about twofold of that between C. sinensis var. sinensis and C. sinensis var. assamica (0.0043). Regarding the distance between the different populations of the Taiwanese wild tea, the averages were 0.0016 between the Nantou and the East populations, 0. 0038 between the South and

Figure 2. The phylogenetic tree based on sequences of the RPB2 introns 12-16 and 23. Branch supports of MP and NJ (shown in parentheses) were generated by 1000 replicates of bootstraps.

Table 2. List of primers used in the study.

Name

Sequence (5'-3')

Target fragment

Source

C-1-11F (forward)

C-1-16R (reverse)

C-1-11F2 (forward) C-1-16R2 (reverse)

CCACTTATGGGTATCGTCTGGCTG GCCTGCTTACCCATTGCTGACTG AAGAATCTTGCATTGATGGT ATATGTATTACGTGGGGACT

RPB2 introns 12-16

Xiao, 2001

This study

C-1-10AF (forward) C-1-11R (reverse) C-1-10AF1 (forward) C-1-11R1(reverse)

CCCTCTCGAATGACTATTGG GATAGTATGTGGGACCAAGG GTAAGGTTGCAGCTCACATG CCTGTGTGACCATTGTACAT

RPB2 intron 23

Xiao, 2001

This study

Table 3. Average genetic distances among C. formosensis, C. sinensis var. assamica, and C. sinensis var. sinensis and those among the three populations of C. formosensis.

Among taxa


	C. formosensis \|	C. formosensis _\|^\|	C. sinensis var. assamica _\|^\|
	1 C. sinensis var. assamica	C. sinensis var. sinensis	C. sinensis var. sinensis
Average genetic distance	0.0092士0.0022	0.0079士0.0026	0.0043士0.0023
Among populations of C. formosensis
	Nantou population _\|^\|	East population _\|^\|	Nantou population _\|^\|
	South population	South population	East population
Average genetic distance	0.0038士0.0002	0.0033士0.0002	0.0016士0.0006

SU et al. ― Taxonomy of Camellia formosensis

481

Nantou populations, and 0.0033 between the South and East populations, and these were lower than the average distance between C. sinensis var. sinensis and C. sinensis var. assamica (0.0043).

DISCUSSION

The results of the present study demonstrated that Taiwanese wild tea is considerably different from both C. sinensis var. sinensis and C. sinensis var. assamica. This is evident from the monophyletic grouping of all the sequences of the Taiwanese wild tea in the cladogram well separated from that of the latter two taxa (Figure 2). This result is consistent with our previous observation on phenetic analysis that all the 52 specimens of the Taiwanese wild tea formed a distinct clade well separated from the 113 specimens of the other two taxa. Furthermore, the pubescences on the ovary and dormant buds along with several other minor characters distinguish the Taiwanese wild tea from the latter two taxa (Su et al., 2007). In the earlier molecular phylogenetic study with RPB2 sequences Xiao and Parks (2003) indicated that C. sinensis var. sinensis was different from C. sinensis var. assamica and the average of pairwise genetic distance of RPB2 introns 12-16 and 23 among the species of section Thea was 0.0087±0.0054 (Xiao, 2001, Appendix 6 & 7). In this study, the pairwise genetic distances―0.0092, between the Taiwanese wild tea and C. sinensis var. assamica and, 0.0079, between the Taiwanese wild tea and C. sinensis var. sinensis―are equivalent to the average species distance of the section Thea (Table 3). This clearly suggests that Taiwanese wild tea occupies the position of a distinct species. Recently Vijayan et al. (2009) also found a similar relationship in nrITS sequence analysis and supported the independent species status of Taiwanese wild tea. Therefore, we strongly advocate treating Taiwanese wild tea as an independent species

In the present study, we used three populations s e p a r a t e d g e o g r a p h i c a l l y ; h e n c e , t h e i n t r a s p e c i fi c variability of the Taiwanese wild tea was also assessed, which provided another interesting point as these three populations showed different relative relationships as suggested by the morphological phenetic study (Su et al. 2007). The present study reveals that the Nantou and the East populations are found closer than each to the South population (Figure 2). The pairwise distances 0.0016, 0.0038, and 0.0033, respectively, between the Nantou and the East, the South and the Nantou, and the South and the East populations, further confirms this relationship (Table 3) although results from phenetic analysis suggested that the Nantou population was closer to the South than to the East population (Su et al., 2007). Such incongruence between the morphological and molecular sequence analysis generally results from convergent evolution or from morphological adaptations to the environment (Wendel and Doyle, 1998). In the present case, the leaves and buds of the East population are hairy while those of the Nantou and the South populations are glabrous (Su et al.,

2007) and these traits, i.e. leaf and bud pubescences, are the major components that separated the East population from the other two populations in the phenetic trees (Su et al., 2007). It has been shown that morphological traits of vegetative parts are often influenced by the environmental factors (Jonas and Geber, 1999; Santamaria et al., 2003; Ellison et al., 2004), and the climate of eastern Taiwan is in fact quite different from that of central-west and south, especially in winter when the northeast monsoon is prevalent. This event brings strong winds to the Taitung region where the East population locates, but it diminishes when it reaches central-western and southern Taiwan after passing through the Central Mountain Range. Since a windy environment causes high evapotranspiration and leaf hairs are generally thought of as an adaptive structure to reduce evaporation in environments with water stress (Ehleringer, 1982; Woodman and Fernandes, 1991; Kenzo et al., 2008), we, therefore, assume that the pubescences on the bud and leaves of the East population of Taiwanese wild tea evolved as an adaptive trait to cope up with the windy environment. Such divergent evolutions of the East population could be the main factors causing the deviation of the East population in the phenetic tree. Nevertheless, such intraspecific variability within Taiwanese wild tea is much less than the interspecific variability observed among Taiwanese wild tea, C. sinensis var. assamica and C. sinensis var. sinensis, which is evident from both Figure 2 and Table 3.

In summary, the results of the present study strongly suggest that 'Taiwanese wild tea' (C. formosensis) is distinct from other tea producing taxa, hence, it should be treated as a separate species. The following is the taxonomic treatment of the Taiwanese wild tea as an independent species.

Camellia formosensis (Masamune et Suzuki) M. H. Su, C. F. Hsieh et C. H. Tsou comb. nov.台灣山茶

(Figures 3, 4)

Basionym: Thea formosensis Masamune et Suzuki, S. Suzuki, ed. Taiwan Zyumoku Benran. 262. 1937.― NEOTYPE (designated here): Formosa (Taiwan), Taityu (Nantou Co.), Gyoti (Yuchih), 8 Nov 1935, flowering, S. Suzuki s. n. (TAI no. 218078).

Camellia sinensis (L.) O. Kuntze forma formosensis Ki-tamura, Acta Phytotax. Geobot. 14: 59. 1950; Hsieh, Yang & Lin, Fl. Taiwan 2nd. ed. 2: 672-673. 1996— TYPE: Japan, Kanaya Tea Research Station (Introduced from Taiwan), 21 Nov 1949, flowering, S. Kitamura s.

n. (holotype: KYO).

A small, evergreen tree, up to 8 m high; trunk up to 40 cm in diameter, usually sprouting from base; bark pale-white, smooth; branches and branchlets glabrous, smooth; winter buds ovate-lanceolate, glabrous or sometimes sparsely pubescent. Leaves alternate, glabrous both sides or sometimes sparsely pubescent beneath, thin-coriaceous, oblong, 7-17 cm long, 2-6.5 cm broad, tip acuminate to cuspidate, base cuneate, margin finely serrate, midrib

482

Botanical Studies, Vol. 50, 2009

elevated on both sides, lateral veins 8-14, usually > 10; petioles 3-10 mm long, glabrous or sometimes sparsely pubescent. Flowers axillary, solitary or 2-4 in fascicles; pedicels cernuous, glabrous, 3-7 mm long; bracts 2, tiny, deciduous; sepals 5, persistent, broadly ovate, 2-3 mm long, 2-3 mm wide, glabrous outside, fully or partly pubescent inside; petals 5-7, white, suborbicular, 7-13 mm long, 6-11 mm wide, glabrous; stamens many, 5-12 mm long, glabrous; ovary 3-locular, glabrous, 2-3 mm long; style 1, 6-12 mm long, usually < 10 mm, glabrous, stigma 3-fid. Fruits globose or depressed globose, 2-3 cm across,

glabrous; seeds 1-3, globose or semi-globose, 1 cm across, glabrous, hilium obvoid, 1-2 mm across. Flowering from September to January.

Endemic to Taiwan, in the central, southern and southeastern parts of the island, including Nantou, Yunlin, Chiayi, Kaohsiung, Pingtung and Taitung Counties, at the

elevation of 800-1,800 m.

Additional specimens examined: TAIWAN. NANTOU CO.: Luku, 7 Oct 1992, C. C. Liao 719 (HAST); Meiyuenshan, elev. 1,500 m, 12 Jan 1935, A. Tanimura

Figure 3. Camellia formosensis. 1, flowering branch; 2, mature fruit (side view); 3, mature fruit (bottom view). Left scale bar is for the flowering branch, and right one is for the mature fruit.

SU et al. ― Taxonomy of Camellia formosensis

483

Yuyuoshan, 2 Jan 2005, M. S. Tang et al. 606 (TAI). PINGTUNG CO.: Akohuzi, 7 Nov 1919, E. Matsuda s. n. (TAI); Chiupaoshan, 14 Oct 2001, S. M. Ku 1442 (TAIF); same loc., elev. 1,500 m, 14 Apr 2004, M. H. Su 575 (TAI);

Jenlishan, elev. 1,000-1,100 m, 10 Jul 1993, J. C. Wang & H. T. Hung 8467 (HAST); same loc., 10 Mar 2004, M. H.

Su 544, 545 (TAI); same loc., 29 May 2004, S. W. Chung 7090 (TAIF); same loc., 27 Sep 2005, M. H. Su 645, 646, 647 (TAI); Paiwan, 7 Nov 1912, E. Matsuda s. n. (TAI); Sasi Logging Road, elev. 1,300 m, 20 Sep 2003, M. H. Su 269, 270 (TAI); Wuweishan, 1 Jan 1919, Matsudai s.

n. (TAIF); same loc., 3 Dec 1995, K. C. Yang et al. 4527 (TAIF); same loc., elev. 1,000 m, 24 Jan 2004, M. H. Su

497, 498 (TAI). TAITUNG CO.: Yungkangshan, elev. 900 m, 9 Mar 1986, S. Y. Lu 18580, 18595 (TAIF); same loc., elev. 900-1,100 m, 28 Sep 2005, M. H. Su 653, 654, 655, 656, 659, 660, 661 (TAI).

Taxonomic background: the first valid publication of Taiwanese wild tea was by Masamune and Suzuki in 1937 (Suzuki, 1937), in which the plant was named Thea formosensis, but no specimens were cited. However, the genus Thea had been treated as a synonym of the genus Camellia by Sweet (1818), and that was accepted in the International Code of Botanical Nomenclature (ICBN), Article 13.5 (Mcneill et al., 2006). Here, we propose C. formosensis as the name of a new combination for Taiwanese wild tea and treat T. formosensis as a basionym. In 1950, Kitamura published C. sinensis forma formosensis and assigned a holotype, but without citing the previous publication of T. formosensis. The holotype of C. sinensis forma formosensis displays the same traits as T. formosensis; therefore, C. sinensis forma formosensis is considered a synonym of T. formosensis. Because Masamune and Suzuki (1937) did not cite any specimens in their publication, we now designate one of Suzuki's collections (S. Suzuki s. n., 8 Nov 1935; TAI no. 218078) as the neotype for T. formosensis according to the ICBN,

Article 9.6 (Mcneill et al., 2006).

Acknowledgments. The first author appreciates the help given by the following herbaria during his visits to study

specimens: TAI, TAIF, HAST and TNM. The authors

thank Yuchih Branch of the Tea Research and Extension Station for providing valuable materials and the following people for their valuable opinions and materials: Dr. J. C. Wang (National Taiwan Normal University), Dr. C. I. Peng (Academia Sinica, Taiwan) and Dr. J. M. Hu (National Taiwan University). We also thank the anonymous reviewers and Dr. Kunjupillai Vijayan (Academia Sinica, Taiwan) for giving valuable comments and Ms. Su-ching Chang for preparing the line-drawing.

LITERATURE CITED

Chou, C.S. 1717. Chu-Lo-Hsien-Chi. The 56th year of the Kangxi Emperor, Ching Dynasty. Reprinted in 1962 by Bank of Taiwan, Taipei, pp. 286.

Figure 4. Camellia formosensis. A, flowering branch; B, dehiscent fruit.

s. n. (TAI); same loc., 8 Oct 1935, S. Sasaki s. n. (TAI); same loc., 8 Nov 1935, S. Sasaki s. n. (TAI); same loc., 9 Nov 1935, S. Suzuki s. n. (TAI); same loc., 16 Jan 1966, M. Hasimoto s. n. (TAI); same loc., 23 Apr 1966, M. T. Kao 6668, 6676 (TAi); same loc., elev. 1,250 m, 30 Mar 2005, C. H. Tsou 2132, 2134, 2137, 2139; Shuili to Yuchih, 25 Dec 1988, B. J. Wang 15069 (TAIF); Yuchih, 12 Jul 1931, S. Taniguchi s. n. (TAI); Yuchih Branch, Tea Research and Extension Station, 13 Jan 1966, M. Hasimoto s. n. (TAI); same loc., 15 Nov 2005, M. H. Su

683, 687 (TAI). YUNLIN CO.: Kulinjiao, 5 Nov 1906, U. Mori 1901 (TAIF). CHIAYI CO.: Kagi, date unknown, H.

Yamada s. n. (TAIF); Sitin, 1 Jul 1999, K. C. Yang 5624 (TNM); same loc., elev. 1,300 m, 12 Sep 2005, M. H. Su 642 (TAI). KAOHSIUNG CO.: Kakuhozan, 8 Mar 1936. S. Sasaki s. n. (TAI); Liukui, 7 Dec 1996, C. W. Huang

26 (HAST); Nanfong Logging Road, 11 Mar 1986, S. Y. Lu 18673 (TAIF); Nanfongshan, 8 Feb 1965, elev. 1,350

m, C. C. Chuang and M. T. Kao 3369 (TAI); Shanping,

10 Dec 1968, T. C. Huang 4897 (TAI); same loc., 12 May 1971, T. Kiang and M. T. Kao KT439 (TAI); same loc., 25 Dec 1989, S. Y. Lu s. n. (TAIF); same loc., 2 Dec 1996, Y. H. Lai 83 (TAIF); same loc., 25 May 2004, C. P. Lin s. n. (TAIF); Tona, 18 Apr 1986, S. Y. Lu 18948 (TAIF);

484

Botanical Studies, Vol. 50, 2009

Denton, A丄.，B. L. McConaughy, and B.D. Hall. 1998. Usefulness of RNA polymerase II coding sequences for estimation of green plant phylogeny. Mol. Biol. Evol. 15: 1082-1085.

Ehleringer, J. 1 982. The influence of water stress and temperature on leaf pubescence development in Encelia farinosa. Am. J. Bot. 69: 670-675.

Ellison, A.M., H.L. Buckley, T.E. Miller, and N.J. Gotelli. 2004. Morphological variation in Sarracenia purpurea (S arraceniaceae): geographic, environmental, and taxonomic correlates. Am. J. Bot. 91: 1930-1935.

Hsieh, C.F., L.K. Ling, and K.C. Yang. 1996. Camellia, Theaceae. In D. E. Boufford, C. F. Hsieh, T. C. Huang, H. Ohashi and Y. P. Yang (eds.), Flora of Taiwan, second edition, Vol. 2. Editorial Committee of Flora of Taiwan, Taipei, pp. 662-693.

Jonas, C.S. and M.A. Geber. 1999. Variation among populations of Clarkia Unguiculata (Onagraceae) along altitudinal and latitudinal gradients. Am. J. Bot. 86: 333-343.

Kenzo, T., R. Yoneda, M. Azani, and N. Majid. 2008. Changes in leaf water use after removal of leaf lower surface hairs on Mallotus macrostachyus (Euphorbiaceae) in a tropical secondary forest in Malaysia. J. Forest Res. 13: 137-142.

Kitamura, S. 1950. On tea and camellias. Acta Phytotaxon.

Geobot. 14: 56-63.

Lai, J.A., W.C. Yang, and J.Y. Hsiao. 2001. An assessment of genetic relationships in cultivated tea clones and native wild tea in Taiwan using RAPD and ISSR markers. Bot. Bull.

Acad. Sin. 42: 93-100.

Lu, S.Y. and Y.P. Yang. 1987. The wild tea of Taiwan. Quart. J. Chin. For. 20: 101-107.

Mace, G.M. 2004. The role of taxonomy in species conservation.

Phil. Trans. R. Soc. Lond. B. 359: 711-719.

McNeill, J., F.R. Barrie, H.M. Burdet, V. Demoulin, D.L. Hawksworth, K. Marhold, D.H. Nicolson, J. Prado, P.C. Silva, J.E. Skog, J.H. Wiersema, and N.J. Turland (eds.). 2006. International Code of Botanical Nomenclature (Vienna Code). A. R. G. Gantner Verlag KG, Ruggell, 569

pp.

Ming, T.L. 2000. Monograph of the Genus Camellia. Yunnan Science and Technology Press, Kunming, pp. 130.

Santamaria, L., J. Figuerola, J.J. Pilon, M. Mjelde, A.J. Green,

T.D. Boer, R.A. King, and R.J. Gornall. 2003. Plant performance across latitude: the role of plasticity and local adaptation in an aquatic plant. Ecology 84: 2454-2461.

Su, M.H., C.H. Tsou, and C.F. Hsieh. 2007. Morphological

comparisons of Taiwan native wild tea plant (Camellia sinensis (L.) O. Kuntze forma formosensis Kitamura) and two closely related taxa using numerical methods. Taiwania

52: 70-83.

Suzuki, S. 1937. Taiwan Zyumoku Benran, Taihoku (Taipei). 262 pp.

Swofford, D.L. 2001. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts.

Thompson, J.D., T.J. Gibson, F. Plewniak, F. Jeanmougin, and D.G. Higgins. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25: 4876-4882.

Vijayan, K. and C.H. Tsou. 2008. Phylogenetic implications of nrITS in Camellia: The need of high quality DNA and PCR amplification with Pfu-DNA polymerase. Bot. Stud. 49:

177-188.

Vijayan, K., W.-J. Zhang, and C.-H. Tsou. 2009. Molecular

taxonomy of Camellia (Theaceae) inferred from nrITS sequences. Am. J. Bot. (in press).

Wachira, F., J. Tanaka, and Y. Takeda. 2001. Genetic variation and differentiation in tea (Camellia sinensis) germplasm revealed by RAPD and AFLP variation. J. Hort. Sci. Biotech. 76: 557-563.

Wendel, J.F. and J.J. Doyle. 1998. Phylogenetic incongruence: window into genome history and molecular evolution. In

P.S. Soltis, D.E. Soltis, and J.J. Doyle (eds.), Molecular

Systematics of Plants II, Kluwer Academic Publishers,

Dordrecht, pp. 265-296.

Woodman, R. L. and G.W. Fernandes. 1991. Differential mechanical defense: herbivory, evapotranspiration, and leaf-

hairs. Oikos 60: 11-19.

Xiao, T. J. 2001. Phylogenetic relationships of the genus Camellia (Theaceae) based on the RNA polymerase II (RPB2) sequences. Ph.D. Dissertation, Department of Biology, University of North Carolina, 293 pp.

Xiao, T.J. and C.R. Parks. 2003. Molecular analysis of the genus Camellia. Intl. Camellia J. 35: 57-65.

SU et al. ― Taxonomy of Camellia formosensis

485

由DNA序列分析確認台灣山茶（山茶科）係一獨立的種

蘇夢淮¹ 謝長富² 鄒稚華³

¹中國文化大學森林暨自然保育學系
²國立台灣大學生態學與演化生物學研究所
³中央研究院植物暨微生物學研究所

台灣山茶係指台灣特有的一種野生茶樹，是茶樹育種上之重要種原。台灣山茶原生於台灣本島
中、南以及東部，但其野生族群正在急速縮減中。分類學上，台灣野生的山茶植物一直被處理為茶

(Camellia sinensis var. sinensis)或阿薩姆茶(C. sinensis var. assamica)的變種，不過我們最近形態學上的

研究顯示台灣山茶可與上述兩個分類群明顯區隔。在本研究中我們提出分子學上的証據以進一步確認台
灣山茶的分類地位。本次樣品涵蓋所有台灣山茶的野生族群以及十九個茶與阿薩姆茶的樣品。我們分析

了核 DNA中RPB2 基因的 introns 12-16 以及23 二個片段，利用最大簡約法與鄰近連接法來分析，結果
顯示分佈於台灣不同地區的野生山茶係一單源的分類群。參考遺傳距離之比較，以及先前所做的數值分
類分析，我們確認台灣山茶應處理為一個獨立的種。在此我們提出一個新的組合，將台灣山茶之學名定
為C. formosensis (Masamune et Suzuki) M. H. Su, C. F. Hsieh et C. H. Tsou ^c

關鍵詞：台灣山茶；茶組；山茶科；RPB2基因；分子分類；台灣野生茶。



	SU et al. ― Taxonomy of Camellia formosensis	485

	由DNA序列分析確認台灣山茶（山茶科）係一獨立的種蘇夢淮¹ 謝長富² 鄒稚華³ ¹中國文化大學森林暨自然保育學系 ²國立台灣大學生態學與演化生物學研究所 ³中央研究院植物暨微生物學研究所台灣山茶係指台灣特有的一種野生茶樹，是茶樹育種上之重要種原。台灣山茶原生於台灣本島中、南以及東部，但其野生族群正在急速縮減中。分類學上，台灣野生的山茶植物一直被處理為茶 (Camellia sinensis var. sinensis)或阿薩姆茶(C. sinensis var. assamica)的變種，不過我們最近形態學上的研究顯示台灣山茶可與上述兩個分類群明顯區隔。在本研究中我們提出分子學上的証據以進一步確認台灣山茶的分類地位。本次樣品涵蓋所有台灣山茶的野生族群以及十九個茶與阿薩姆茶的樣品。我們分析了核 DNA中RPB2 基因的 introns 12-16 以及23 二個片段，利用最大簡約法與鄰近連接法來分析，結果顯示分佈於台灣不同地區的野生山茶係一單源的分類群。參考遺傳距離之比較，以及先前所做的數值分類分析，我們確認台灣山茶應處理為一個獨立的種。在此我們提出一個新的組合，將台灣山茶之學名定為C. formosensis (Masamune et Suzuki) M. H. Su, C. F. Hsieh et C. H. Tsou ^c 關鍵詞：台灣山茶；茶組；山茶科；RPB2基因；分子分類；台灣野生茶。