Botanical Studies (2012) 53: 265-274.

SYSTEMATICS

DNA barcoding of Nyssaceae (Cornales) and taxonomic issues

Nian WANG^1,2,6, Frederic M.B. JACQUES³, Richard I. MILNE⁴, Chang-Qin ZHANG^1,2 *, and

Jun-Bo YANG⁵

¹Kunming Botanic Garden, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, China

²Key Laboratory of Resource Plants and Biotechnology, Kunming Institute of Botany, Chinese Academy of Sciences, Kun-
ming650204, China

³Key Laboratory of Tropical Forest Ecology, Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences,
Mengla 666303, China

⁴Institute of Molecular Plant Sciences, University of Edinburgh, Edinburgh EH9 3JH, UK

⁵The Germplasm Bank of Wild Species in Southwest China, Kunming Institute of Botany, Chinese Academy of Sciences,
Kunming 650201, China

(Received June 2, 2011; Accepted December 22, 2011)

ABSTRACT. DNA barcoding, as a tool for species discrimination, has been efficiently used in animals. However, there are still debates on which DNA region (s) can be adopted as the standard barcode (s) for land plants. In the present study, we evaluated the four proposed barcoding loci (matK, rbcL, trnH-psbA and ITS) on nine species of Nyssaceae. The results showed that ITS was the best performing single locus, although matK + rbcL might be used as the core barcodes for land plants. The chloroplast regions have low resolution compared with ITS. The low efficiency of these candidate barcodes in Nyssaceae might be caused by a poor taxonomy, especially within the genus Nyssa. The results also indicated that species status of N. shangszeen-sis, N. sinensis, N. shweliensis and N. wenshanensis requires to be reevaluated based on more morphological characters combined with rapidly evolving loci.

Keywords: DNA barcoding; ITS; matK; Nyssaceae; rbcL; trnH-psbA.

INTRODUCTION

DNA barcoding refers to the use of standardized DNA sequence as a tag for rapid and accurate species identification (Herbert et al., 2003, 2004; Savolainen et al., 2005; Kress et al., 2007; Sundberg et al., 2010). DNA barcoding based on a fragment of cytochrome c oxidase I gene (COI) was successfully used to discriminate animal species (Herbert et al., 2004; Kress et al., 2005; Chase et al., 2005; Ward et al., 2005; Fazekas et al., 2008; Lakra et al., 2011). However, the application of DNA barcoding to plants has been impeded due to problems such as difficulties with amplification and sequencing, and relatively lower diver-gence between than within species (Kress et al., 2005; Liu et al., 2010). Hence, a multi-locus approach based on the chloroplast genome has been proposed, and is increasingly accepted as an effective strategy for barcoding land plants

(Newmaster et al., 2008; Kress and Erickson, 2007; CBOL Plant Working Group, 2009). As the Barcoding of Life Database has an increasing number of data entries (Ratnas-ingham and Hebert, 2007), barcoding undoubtedly plays an increasingly important role when it comes to species that are difficult to identify. Although the power of DNA barcoding is challenged by taxonomically closely related species (Newmaster et al., 2008; Newmaster and Ragupa-thy, 2009), it remains promising in distinguishing cryptic species (Ragupathy et al., 2009).

Nyssa L. is a genus in which few characters separate many of the species, leading to taxonomic ambiguity, especially for those distributed in eastern Asia. Since current classification is mainly based on morphological characters and the geographical ranges of taxa, views on its taxonomy differ between authors. Initially, two species were proposed for eastern Asia, N. sinensis and N. javanica (Eyde, 1963). Subsequently, five further species were described, i.e. N. shweliensis (Airy Shaw, 1969), N. shangszeensis, N. leptophylla and N. wenshanensis (Fang and Song, 1975), and N. yunnanensis (Wu and Fan, 1977). All seven species were recognized by Qin and Phengklai (2007); however, Wen and Stuessy (1993) treated N. yunnanensis, N. wen-shanensis and N. leptophylla as part of N. javanica species complex (and hence synonymous) with N. javanica, based

⁶Present Address: School of Biological and Chemical Sciences, Queen Mary University of London, Mile End Road, London E1 4NS, United Kingdom. E-mail: wangnian@ mail.kib.ac.cn.

*Corresponding author: E-mail: zhangchangqin@mail.kib. ac.cn; Tel: +86-871-5223630; Fax: +86-871-5216345.

266

Botanical Studies, Vol. 53, 2012

on morphological similarities. Hence classifications disagree on how many species are present in East Asia, and molecular data provides a means of assessing the accuracy of the two taxonomic treatments.

Given the uncertainty regarding the number of taxo-nomic species in eastern Asian Nyssa, a DNA-based approach seems appropriate. Therefore, the objectives of this study are to: 1) test the efficiency of the four loci (matK, rbcL, trnH-psbA and ITS) as barcodes for Nyssaceae; and 2) use data from these to reevaluate the taxonomic issues.

MATERIALS AND METHODS

Plant materials

To gain a more comprehensive understanding, we sampled the species as recorded in Flora of China. Material of N. leptophylla Fang et Soong could not be obtained because insufficient information on its distribution (Sun, 2008) prevented us from locating it in the field. From the other six species, between three and five accessions were collected per species (Figure 1; Table S1). In addition, three accessions of one of the American species, N. sylvat-ica, were obtained from the Arnold Arboretum of Harvard University. Three and four accessions each of species from two other genera of Nyssaceae, Camptotheca acuminata and Davidia involucrata were also included (Table S1). Hence a total of 32 accessions from nine species were used to evaluate the candidate barcodes.

Material of N. yunnanensis and N. javanica was collected from one locality each (Table S1) and could be assigned to these species based on morphology. However, past taxo-nomic confusion combined with morphological similarity made identification of the other four species uncertain. Each of N. shangszeensis and N. wenshanensis have only ever been collected from their type localities (Table S1), and in each case no other Nyssa species has ever been recorded from the vicinity of their type locality. Our own

field observations confirmed the presence of a species matching the type description, and no other, at each type locality. Therefore, material from the type locality of each species was assigned to that species in each case. Similarly, material from a locality where N. sinensis (and no other species) and N. shweliensis (and no other species) had previously been recorded was examined, confirmed to match descriptions of each species, and therefore assigned to N. sinensis and N. shweliensis respectively. From each wild population of each species, each individual sampled was at least 30 meters from the nearest sampled accession. In addition, one cultivated accession of N. sinensis from Kunming was examined.

DNA extraction, PCR amplification and sequencing

Genomic DNA was extracted from 30-50 mg of silica-gel dried leaves of all accessions using a modified CTAB (cetyl trimethyl ammonium bromide) method (Doyle and Doyle, 1987). The extracted total DNA was diluted to a final concentration of 50-80 ng/[il for subsequent use. Am-plifications of the target gene regions were performed via the polymerase chain reaction (PCR) in a GeneAmp PCR System 9700 DNA Thermal Cycler (PerkinElmer, USA) or Eppendorf (Hamburg, Germany).

Nuclear ribosomal ITS and three chloroplast genes (matK, rbcL and psbA-trnH) were amplified (White et al. 1990; Tate and Simpson, 2003; Dunning and Savolainen, 2010; Ivanova et al., 2008). The PCR amplifications were carried out in 25 [il mixtures contained 0.3 U AmpliTaq polymerase, 10 x buffer, 1.5 mmol/L MgCl?, 0.2 mmol/L dNTP, 0.5 imol/L primer and 30-50 ng genomic DNA. The PCR conditions consisted of an initial denaturation at 95°C for 2-3 min, followed by 32-34 cycles of 1 min at 94°C, 50 sec at 52°C, 1.5 mm at 72°C, and ended with an extension step of 7 min at 72°C. The PCR products were purified using the GELase Agarose (Epicentre Technologies, Madison, WI, USA) according to the manufacturer's protocol for sequencing PCR reactions. Sequencing reactions were performed using PRISM Dye Terminator Cycle Sequencing Ready Reaction kit (Applied Biosystems, Foster City, Calif.). The sequencing products were run on an ABI 3700 automated sequencer (Perkin Elmer). All sequences were deposited in GenBank with the accession numbers listed in Table S1.

Data analysis

Sequences from each DNA region were aligned using the CLUSTAL X version 1.83 (Thompson et al., 1997) then refined manually. The inter- and intraspecific divergences of each barcoding region was computed by calculating Kimura 2-parameter (K2P) distances in MEGA 4 (Tamura et al., 2007). The significance of divergence was then assessed by Wilcoxon signed rank and Wilcoxon two-sample tests (http://www.fon.hum.uva.nl/Service/Statis-tics.html). The neighbor-joining (NJ) method was used to construct phylogenetic trees in PAUP 4.0b10 (Swofford et

Figure 1. The localities of Nyssa species used in this study.

WANG et al. ― DNA barcoding of Nyssaceae

267

al., 2002) due to its accuracy for smaller data sets and its computational speed (Tamura et al., 2004). The K2P model was adopted because it performs best for low value genetic distances, and is therefore popularly used for species-level analysis (Nei and Kumar, 2000). Branch support for NJ was assessed with 1000 bootstrap replicates.

RESULTS

Successful PCR amplification rate and sequence quality

The amplification success rates vary among the four

loci, with 92% in ITS, 89.5% in rbcL, 86.9% in trnH-psbA and 80.5% in matK. High quality bidirectional sequences were obtained from rbcL and trnH-psbA. The remaining two loci required more manual editing in Nyssa species. The greatest problems in obtaining bidirectional sequences with few ambiguous bases were encountered with matK; this was in part attributable to mononucleotide repeats disrupting individual sequencing reads.

Genetic divergence between and within species

The Wilcoxon signed-rank test demonstrated that ITS exhibited the highest divergence at the interspecific level

Supplementary Table 1. Detailed information on samples used in this study.


Specimen	Locality	Voucher No.		GenBank Accession No.
Specimen	Locality	Voucher No.	ITS	matK	rbcL	trnH-psbA
N. yunnanensis 1	Puwen, Yunnan	SBL 2006030901	JQ280773	JQ280869	JQ280837	JQ280805
N. yunnanensis 2	Puwen, Yunnan	SBL 2006030303	JQ280774	JQ280870	JQ280838	JQ280806
N. yunnanensis 3	Puwen, Yunnan	SBL 2006030302	JQ280775	JQ280871	JQ280839	JQ280807
N. yunnanensis 4	Puwen, Yunnan	SBL 2006030304	JQ280776	JQ280872	JQ280840	JQ280808
N. javanica 1	Mengsong, Yunnan	SBL 2007040302	JQ280777	JQ280873	JQ280841	JQ280809
N. javanica 2	Mengsong, Yunnan	SBL 2005040601	JQ280778	JQ280874	JQ280842	JQ280810
N. javanica 3	Mengsong, Yunnan	SBL 2005032703	JQ280779	JQ280875	JQ280843	JQ280811
N. sinensis 1	Pingbian, Yunnan	SBL 2007041802	JQ280759	JQ280855	JQ280823	JQ280791
N. sinensis 2	Pingbian, Yunnan	SBL 2007041803	JQ280760	JQ280856	JQ280824	JQ280792
N. sinensis 3	Kunming, Yunnan	SBL 2007041801	JQ280758	JQ280854	JQ280822	JQ280790
N. shangszeensis 1	Shangsi, Guangxi	SBL 2007041001	JQ280764	JQ280860	JQ280828	JQ280796
N. shangszeensis 2	Shangsi, Guangxi	SBL 2007041002	JQ280765	JQ280861	JQ280829	JQ280797
N. shangszeensis 3	Shangsi, Guangxi	SBL 2007041003	JQ280767	JQ280863	JQ280831	JQ280799
N. shangszeensis 4	Shangsi, Guangxi	SBL 2007041004	JQ280766	JQ280862	JQ280830	JQ280798
N. shweliensis 1	Honghe, Yunnan	SBL 2007040207	JQ280768	JQ280864	JQ280832	JQ280800
N. shweliensis 2	Honghe, Yunnan	SBL 2007040213	JQ280769	JQ280865	JQ280833	JQ280801
N. shweliensis 3	Honghe, Yunnan	SBL 2007040318	JQ280770	JQ280866	JQ280834	JQ280802
N. shweliensis 4	Honghe, Yunnan	SBL 2007040201	JQ280772	JQ280868	JQ280836	JQ280804
N. shweliensis 5	Honghe, Yunnan	SBL 2007040204	JQ280771	JQ280867	JQ280835	JQ280803
N. wenshanensis 1	Wenshan, Yunnan	SBL 2007041304	JQ280761	JQ280857	JQ280825	JQ280793
N. wenshanensis 2	Wenshan, Yunnan	SBL 2007041303	JQ280762	JQ280858	JQ280826	JQ280794
N. wenshanensis 3	Wenshan, Yunnan	SBL 2007041302	JQ280763	JQ280859	JQ280827	JQ280795
N. sylvatica 1	Arnold Arboretum	SBL 2007041901	JQ280755	JQ280851	JQ280819	JQ280787
N. sylvatica 2	Arnold Arboretum	SBL 2007041902	JQ280756	JQ280852	JQ280820	JQ280788
N. sylvatica 3	Arnold Arboretum	SBL 2007041903	JQ280757	JQ280853	JQ280821	JQ280789
C. acuminata 1	Puwen, Yunnan	SBL 2007033001	JQ280783	JQ280879	JQ280847	JQ280815
C. acuminata 2	Puwen, Yunnan	SBL 2007033002	JQ280784	JQ280880	JQ280848	JQ280816
C. acuminata 3	Puwen, Yunnan	SBL 2007033003	JQ280785	JQ280881	JQ280849	JQ280817
C. acuminata 4	Puwen, Yunnan	SBL 2007033004	JQ280786	JQ280882	JQ280850	JQ280818
D. involucrata 1	Punwen, Yunnan	SBL 20070418011	JQ280780	JQ280876	JQ280844	JQ280812
D. involucrata 2	Punwen, Yunnan	SBL 20070418021	JQ280781	JQ280877	JQ280845	JQ280813
D. involucrata 3	Kunming, Yunnan	SBL 20070418031	JQ280782	JQ280878	JQ280846	JQ280814

268

Botanical Studies, Vol. 53, 2012

(P < 1.757x10^-07) (both between and within genera; Table 1). Within Nyssa, the next highest divergences were for rbcL, trnH-psbA and matK respectively, though all were an order of magnitude less than for ITS (Table 1). Between genera, matK gave the highest distances after ITS, followed by trnH-psbA and rbcL (Table 1). The four loci showed no significant differences in intraspecific divergence (Table S2), with no divergence detected at all for ITS (Table 1). For each gene, the interspecific divergence significantly exceeded the corresponding intraspecific divergence (Table S3).

Taxon discrimination

Even compared to the three cpDNA genes combined, ITS has the highest discriminatory power (Table 3). All species other than N. shweliensis are monophyletic in the ITS NJ trees, with >99% bootstrap support for monophyly of N. sylvatica, N. wenshanensis, N. yunnanensis, Camp-totheca acuminata and Davidia involucrata, 85% and 65% support for N. javanica and N. sinensis respectively, but no support for N. shangszeensis. However, N. shweliensis, appears polyphyletic, though this is not supported.

Of the three cpDNA genes (Table 3), rbcL (44.4%) ranks second to ITS in discriminatory power, followed by matK (33.3%) and trnH-psbA (22.2%) (Table 3). Combin-ing these three gives a similar discriminatory power to rbcL alone, whereas combining one or more cpDNA genes with ITS gives a lower power than ITS alone, or a similar score in the case of ITS + matK (Table 3).

The three combined cpDNA genes distinguish four groupings within Nyssa: N. sylvatica (99% support), the N. yunnanensis plus two accessions of N. javanica (92%), N. sinensis (57%) and a clade of N. shweliensis, N. shang-

szeensis, and N. wenshanensis (54%). However, a third accession of N. javanica is positioned among the latter three clades, making that species appear paraphyletic for cpDNA.

DISCUSSION

Ease of amplification of candidate genes for DNA barcodes in Nyssa

DNA bacording aims to identify species rapidly and accurately by adopting a standardized DNA locus as a tag. Finding an ideal region in plants, i.e. which is sufficiently variable to discriminate among all the species and conserved enough to be less variable within species, is a real challenge (Kress et al., 2007). Thus, many studies have been carried out to evaluate different loci and test their efficiency as DNA barcodes (Kress et al., 2007; Chen et al., 2010; Liu et al., 2010).

PCR and sequencing success are key criteria for DNA barcoding (Kress and Erickson, 2007; Hollingsworth et al., 2009). ITS, rbcL and trnH-psbA performed well regarding this aspect. The locus matK could not be sequenced for some accessions of Nyssa species because of the downstream of mononucleotide repeats which can disrupt individual sequencing reads. Similar problems have been reported before for matK (Sass et al., 2007; Fazekas et al., 2008; Hollingsworth et al., 2009), making it a difficult locus to work with for land plant barcoding.

Discriminatory power of single barcodes

The power to discriminate species is a crucial criterion to select suitable DNA barcodes (CBOL Plant Working

Table 1. Analysis of interspecific and intraspecific divergence of four DNA barcodes for the nine taxonomic species.


Potential barcodes	Interspecific distance¹		Intraspecific distance¹
Potential barcodes	Within Nyssa	Between genera	Within species
matK	0.0016 ± 0.0016	0.02 89 ± 0.0037	0.0003 ± 0.0007
rbcL	0.0025 ± 0.0021	0.0099 ± 0.0013	0.0002 ± 0.0007
trnH-psbA	0.0021 ± 0.0023	0.0591 ± 0.0125	0.0006 ± 0.0012
ITS	0.0242 ± 0.0152	0.1631 ± 0.0292	0.0000 ± 0.0000

¹ Interspecific distance and intraspecific distance were calculated based on Kimura 2-parameter (Nei and Kumar, 2000).

Supplementary Table 2. Wilcoxon signed-rank tests of intraspecific divergence among loci for the nine taxonomic species.


W+	W-	Relative ranks		n	P value	Result
W+	W-	W+	W-	n	P value	Result
matK	rbcL	7	3	4	< 0.625	matK = rbcL
matK	trnH-psbA	0	3	2	< 0.5	matK = trnH-psbA
matK	ITS	10	0	4	< 0.125	matK = ITS
rbcL	ITS	3	0	2	< 0.5	rbcL = ITS
trnH-psbA	ITS	10	0	4	< 0.125	trnH-psbA = ITS
trnH-psbA	rbcL	10	0	4	< 0.125	trnH-psbA = rbcL

WANG et al. ― DNA barcoding of Nyssaceae

269

Group 2009). Given this criterion, ITS performed best as a single barcode out of the four loci tested in this study. Eight out of nine taxonomic species (88.9%) were well defined. This locus was first promoted as a plant barcode by Kress et al. (2005), due to its high amount of interspecific variation. It functioned well in some taxa and was highly recommended for barcoding applications (Sass et al., 2007). However, several characteristics of ITS make it

a potentially problematic barcode, such as gene duplication, incomplete concerted evolution, pseudogenes and the presence of paralogs (King and Roalson, 2008; Starr et al., 2009). These problems were not encountered in the present study.

The remaining loci showed incomplete taxon discrimination with matK identifying three species (33%); rbcL discriminated four (44%); trnH-psbA two (22%). Two loci,

Table 2. Wilcoxon signed-rank tests of interspecific divergence among loci for the nine taxonomic species.


W+	W-	Relative ranks¹		n	P value	Result¹
W+	W-	W+	W-	n	P value	Result¹
matK	rbcL	288	63	26	< 0.0044	matK > rbcL
matK	trnH-psbA	36	429	30	< 5.547x10^-05	matK < trnH-psbA
matK	ITS	0	666	36	< 1.757x10^-07	matK < ITS
rbcL	ITS	0	666	36	< 1.757x10^-07	rbcL < ITS
trnH-psbA	ITS	0	666	36	< 1.757x10^-07	trnH-psbA < ITS
trnH-psbA	rbcL	469	92	33	< 0.0008	trnH-psbA > rbcL

¹The symbols "W+" and "W-" represent the sum of all of the positive values and the sum of all of the negative values in the Signed Rank column, respectively. Symbol ">" is used if the interspecific divergence for a locus significantly exceeds that of another locus.

Table 3. Bootstrap values of nine species based on single or combined DNA loci.


No. of accessions smatK	3 65	4 X	3 X	3 X	4 X	5 X	3 X	4 100	3 100	- 33.3
rbcL	88	X	X	69	X	X	X	97	99	44.4
trnH-psbA	X	X	X	X	X	X	X	100	100	22.2
ITS	99	100	86	100	78	X	99	100	100	88.9
matK + rbcL	99	X	X	54	X	X	X	100	100	44.4
matK + trnH-psbA	81	X	X	X	X	X	X	100	100	33.3
matK + ITS	100	100	91	64	73	X	99	100	100	88.9
rbcL + trnH-psbA	93	X	X	55	X	X	X	100	100	44.4
rbcL + ITS	100	100	X	87	85	X	98	100	100	77.8
trnH-psbA + ITS	100	100	X	64	76	X	99	100	100	77.8
matK + rbcL + trnH-psbA	99	X	X	54	X	X	X	100	100	44.4
matK + rbcL + ITS	100	100	X	86	83	X	99	100	100	77.8
matK + trnH-psbA + ITS	100	100	X	63	70	X	99	100	100	77.8
rbcL + trnH-psbA + ITS	100	100	X	86	85	X	98	100	100	77.8
matK + rbcL + trnH-psbA + ITS	100	100	X	86	80	X	99	100	100	77.8

Values in all but the last column represent percentage bootstrap support for monophyly of all accessions for that taxon that were examined; x indicates that the taxon was not monophyletic for that marker/set of markers. Discrimination (%) represents the percentage of species recognized as monophyletic with >50% bootstrap support by each locus or combination of loci.

270

Botanical Studies, Vol. 53, 2012

matK and rbcL, were proposed as the core barcodes for land plants (CBOL Plant Working Group, 2009). However, their evolution may be too slow for Nyssa, with a resulting incomplete taxon assignment (Table 3).

The trnH-psbA intergenic spacer is one of the most variable regions of the plastid genome (Shaw et al., 2007). However, much of its variability occurs as insertions and deletions, which makes sequence alignment a difficult task (Ford et al., 2009). Although this is considered disadvantageous for barcode application (CBOL Plant Working Group, 2009), the insertion/deletion mutations were also seen as an advantage of trnH-psbA due to their diagnostic nature (Kress and Erickson, 2007). In the present study, trnH-psbA is not proposed as a satisfactory barcode for Nyssaceae due to its low species discriminatory power (Table 3).

Overall, ITS had greater discriminatroy power than all three cpDNA genes combined, successfully separating almost all species examined.

A multi-locus barcoding system for Nyssaceae

The multi-locus barcoding system has been suggested for land plants because no single locus is satisfactory on its own. Within this kind of system, a locus with low evolution rate discriminates individuals at the genus level

and above, while more rapidly evolving loci delineates species within genera (Newmaster et al., 2006; Kress and Erickson, 2007). Combining barcoding markers has shown benefits for species delimitation (Fazekas et al., 2008) because multi-locus combinations can result in increased robustness, as revealed by high clade support values. The combinations of rbcL + matK and rbcL + trnH-psbA were recommended as two-locus systems for land plants (CBOL Plant Working Group, 2009; Kress and Erickson, 2007). Within Nyssaceae, however, even the three genes combined (rbcL + matK + trnH-psbA) had limited discriminatory power, separating only three clades among the six eastern Asian Nyssa species examined (Figure 3). Combining any or all of these with ITS did not improve upon the discrimination obtained from ITS alone.

Supplementary Table 3. Wilcoxon two-sample test based on interspecific versus intraspecific Kimura 2-distances of the four loci.


	Loci	Wilcoxon two-sample test
	matK	#A = 451, #B = 43, W = 3414.5, P < 6.503e-16
	rbcL	#A = 447, #B = 43, W = 2643.5, P < 4.634e-19
	trnH-psbA	#A = 453, #B = 43, W = 4231.5, P < 6.771e-13
	ITS	#A = 456, #B = 43, W = 1010.5, P < 4.643e-27

Figure 2. Neighbor-joining tree based on Kimura 2-parameter using ITS. The branch support was assessed with 1000 replicates. The support values above 50% were shown.

WANG et al. ― DNA barcoding of Nyssaceae

271

Taxonomic implications for Nyssa

The present taxonomy of Nyssa, based on a limited number of morphological characters, is confusing and controversial (Sun, 2008). Based on 18 morphological

characters collected from herbarium specimens, Wen and Stuessy (1993) supported the recognition of four Nyssa species present in eastern Asia: N. sinensis, N. javanica, N. shangszeensis and N. shweliensis. Because the differences among N. javanica, N. yunnanensis, N. wenshan-ensis and N. leptophylla appear minor, mainly regarding several morphological characters (shape, size and textures of leaves, number of veins, length of petiole, pubescence on leaves and infloresences, length of pedicels, and size of fruits), they were treated as the N. javanica complex (Wen and Stuessy, 1993). However, our DNA results contradict this view: based on ITS, N. javanica is clearly distinct from N. yunnanensis and N. wenshanensis; indeed, N. javanica and N. yunnanensis are by some margin the most distinct species according to ITS data. The other four species formed a well-supported clade, within which the monophyly of N. sinensis, N. wenshanensis and N. shangszeensis, but not N. shweliensis, was supported. However, ITS data alone did not provide strong support for species status for these taxa: only N. wenshanensis had strong (99%) bootstrap support for its monophyly, and the interspecific divergence values among these species were 0.7% or less (Table S4). If a threshold of 1.0% genetic distance is taken as the minimum to differentiate species (following Ratnasingham and Hebert, 2007), then N. wen-shanensis, N. shangszeensis and N. shweliensis were not

Supplementary Table 4. The pair-wise distance between the nine taxonomic species revealed by ITS.


N. sylvatica
N. sinensis	0.024
N. wenshanensis	0.026	0.007
N. shangszeensis	0.022	0.002	0.007
N. shweliensis	0.018	0.005	0.007	0.004
N. yunnanensis	0.035	0.045	0.047	0.043	0.039
N. javanica	0.026	0.037	0.039	0.035	0.031	0.009
D. involucrata	0.123	0.136	0.138	0.134	0.129	0.133	0.129
C. camptotheca	0.178	0.194	0.196	0.192	0.189	0.182	0.175 0.219

Figure 3. Neighbor-joining tree based on Kimura 2-parameter using combined chloroplast loci (matK + rbcL + trnH-psbA). The branch support was assessed with 1000 replicates. The support values above 50% were shown.

272

Botanical Studies, Vol. 53, 2012

recognized as distinct species from N. sinensis. However, distance thresholds undoubtedly differ between organisms (Pettengill and Neel, 2010), so this 1% threshold can only serve as an approximate guideline, and other explanations for low divergence must be considered.

Three possible explanations are: first, inaccurate taxonomy (i.e. they are not truly distinct); second, very recent speciation (where insufficient time has passed to accumulate sequence differentiation); and third, past or present hybridization. Hybridization appears unlikely because no two Nyssa species are known to occur in sympatry in eastern Asia, except for N. javanica and N. sinensis (N. Wang, personal observation). However, one accession of N. javanica has a cpDNA type that is in a different position on the NJ tree from the other two (Figure 3), despite this species being monophyletic for ITS (Figure 2). This could indicate possible chloroplast capture via ancient introgres-sion; however, the difference within N. javanica is only one bp, so homoplasy is also a possible explanation.

Recent speciation is plausible for Nyssa, irrespective of whether current taxonomy is accurate. Although Nyssa is an old genus with a long Tertiary history (Edye, 1963; Wen and Stuessy, 1993), this is not incompatible with recent speciation in eastern Asia. For example, Rhododendron is 60 million years old (Milne, 2004), but contains clades of species that are not distinguishable based on ITS sequences (e.g. R. decorum and R. irroratum; Wang, 2010) or particular cpDNA sequences (Milne et al., 2010). Furthermore, Tertiary relict floras contain numerous examples of clades of morphologically similar species within which are nested clades of more distinctive species (Milne and Abbott, 2002); Nyssa fits this pattern. Hence it is likely that N. sinensis, N. shangszeensis, N. shweliensis and N. wenshanensis diverged recently, perhaps as a result of quaternary climatic oscillations.

With this being the case, recognition of these four species as distinct from one another must depend upon other factors, such as morphology. Although all four were recognized by both Wen (1993) and Qin and Phengklai (2007),

our data indicates that some re-evaluation of these taxa might be necessary. The non-monophyly of N. shweliensis for ITS might indicate that this comprises two cryptic species, but the cpDNA provides no evidence either way so this possibility in particular requires further investigation.

Regarding biogeography, both cpDNA and ITS data indicate that East Asian Nyssa do not form a monophyletic group, because N. javanica and N. yunnanensis are basal in both ITS and cpDNA analyses, with the American N. sylvatica sister to the remaining species. This correlates well with the analysis of Eyde (1963) who concluded based on fossil morphology that N. sinensis was more closely-related to N. sylvatica than N. javanica. Clearly the inclusion of only one non-Asian species in our analysis limits what can be inferred, but the simplest explanation for these results would be that Nyssa originated in eastern Asia and that subsequently one (or more) lineages moved into the Americas.

Conclusion

Our data indicates that Nyssa in eastern Asia comprises three relatively old lineages, i.e. N. javanica, N. yunnanen-sis and all other species (bearing in mind that relationships of N. leptophylla remain uncertain), and that recent diversification within the latter clade has given rise to much of the morphological diversity now evident in eastern Asian Nyssa. Our data hence rejects the grouping together of N. yunnanensis, N. wenshanensis and N. javanica as a species complex, and instead indicates that N. wenshanensis, N. shangszeensis, N. shweliensis, N. sinensis and possibly N. leptophylla might form a complex, either of recently diverged species or of taxa that are not truly distinct at specific level.

In spite of potential taxonomical problems in Nyssa, among the studied loci, ITS appeared to be the best barcode candidate for Nyssaceae. To get a whole picture of the taxonomy of Nyssa, a detailed morphological study of Nyssa combined with more rapidly evolving loci is needed.

Acknowledgements. We sincerely thank Yu-Xiao Zhang and Zhao-Ming Cai, at Kunming Institute of Botany, Chinese Academy of Sciences (CAS), for their help with laboratory work and data analysis. The research was funded by the Natural Science Foundation of China (No. 30770139), the Bureau of Science and Technology of Yunnan, China (No. 2009BB001), the Large-Scale Scientific Facilities of the Chinese Academy of Sciences (No. 2009-LSF-GBOWS-01), GBOWS and by CAS Young Scientists Fel-lowship (2009YB1-13) and NSFC Research Fellowship for International Young Scientists (40950110338) to F. M. B. Jacques.

LITERATURE CITED

Airy Shaw, H.K. 1969. An overlooked Chinese Nyssa. Kew Bull. 23: 311.

CBOL Plant Working Group. 2009. A DNA barcode for land plants. Proc. Nat. Acad. Sci. USA 106: 12794-12797.

Chase, M.W., N. Salamin, M. Wilkinson, J.M. Dunwell, R.P. Ke-sanakurthi, N. Haidar, and V. Savolainen. 2005. Land plants and DNA barcodes: short-term and long-term goals. Phil. Trans. R. Soc. Lond. B. 360: 1889-1895.

Chen, S.L., H. Yao, J.P. Han, C. Liu, J.Y. Song, L.C. Shi, Y.J. Zhu, X.Y. Ma, T. Gao, X.H. Pang, K. Luo, Y. Li, X.W. Li,

X.C. Jia, Y.L. Lin, and C. Leon. 2010. Validation of the

ITS2 region as a novel DNA barcode for identifying medicinal plant species. PloS One. 5: 1-8.

Cronquist, A. 1981. An Integerated System of Classification of Flowering Plants. Columbia University, New York.

Doyle, J.J. and J.L. Doyle. 1987. A rapid DNA isolation proce-dure for small quantities of fresh leaf tissue. Phytochem. Bull. 19: 11-15.

Dunning, L.T. and V Savolainen. 2010. Broad-scale amplifica-

WANG et al. ― DNA barcoding of Nyssaceae

273

tion of matK for DNA barcoding plants, a technical note.

Bot. J. Linn. Soc. 164: 1-9.

Eyde, R.H. 1963. Morphological and paleobotanical studies of the Nyssaceae. A survey of the modern species and their

fruits. J. Arnold Arbor. 44: 328-376.

Fang, W.P. and Z.P. Song. 1975. Praecursors Florae Nyssace-arum Sinensium. Acta. Bot. Sin. 2: 83-89.

Fazekas, A.J., K.S. Burgess, P.R. Kesanakurti, S.W. Graham, S.G. Newmaster, B.C. Husband, D.M. Percy, M. Hajiba-baei, and S.C.H. Barrett. 2008. Multiple multilocus DNA

barcodes from the plastid genome discriminate plant species

equally well. PLoS One 3: e2802. Ford, C.S., K.L. Ayres, N. Toomey, N. Haider, J. Van A. Stahl,

L.J. Kelly, N. Wikstrom, P.M. Hollingsworth, R.J. Duff, S. Hoot, R.S. Cowan, M.W. Chase, and M.J. Wilkinson. 2009. Selection of candidate coding DNA barcoding regions for use on land plants. Bot. J. Linn. Soc. 159: 1-11.

Hebert, P.D.N., N.A. Cywinska, S.L. Ball, and J.R. Ward. 2003. Biological identifications through DNA barcodes. Phil. Trans. R. Soc. Lond. B. 270: 313-321.

Hebert, P.D.N., M.Y. Stoeckle, T.S. Zemlak, and C.M. Francis.

2004. Identification of birds through DNA barcodes. PLoS

Biol. 2: 1657-1663.

He, Z., J.Q. Li, and H.C. Wang. 2004. Karyomorphology of Da-vidia involucrata and Camptotheca acuminata, with special reference to their systematic positions. Bot. J. Linn. Soc.

144: 193-198.

Hollingsworth, M.L., A.A. Clark, L.L. Forrest, J. Richardson,

R.T. Pennington, D.G. Long, R. Cowan, M.W. Chase, M.

Gaudeul, and P.M. Hollingsworth. 2009. Selecting barcod-ing loci for plants: evaluation of seven candidate loci with species-level sampling in three divergent groups of land

plants. Mol. Ecol. Res. 9: 439-457.

King, M.G. and E.H. Roalson. 2008. Exploring evolutionary dynamics of nrDNA in Carex subgenus Vignea (Cyperaceae). Syst. Bot. 33: 514-524.

Ivanova, N.V., N.J. Fazekas, and P.D.N. Hebert. 2008. Semi-

automated, membrane-based protocol for DNA isolation

from plants. Plant Mol. Biol. Rep. 26: 186-198. Kress, W.J., K.J. Wurdack, E.A. Zimmer, L.A. Weigt, and D.H.

Janzen. 2005. Use of DNA barcodes to identify flowering plants. Proc. Nat. Acad. Sci. USA 102: 8369-8374.

Kress, J.W. and D.L. Erickson. 2007. A two-locus global DNA

barcode for land plants: the coding rbcL gene complements the non-coding trnH-psbA spacer region. PLoS One 2:e508.

Lakra, W.S., M.S. Verma, M. Goswami, K.K. Lal, V. Mohindra, P. Punia, A. Gopalakrishnan, K.V. Singh, R.D. Ward, and P. Hebert. 2011. DNA barcoding Indian marine fishes. Mol. Ecol. Res. 11: 60-71.

Liu, Y., H.F. Yan, T. Cao, and X.J. Ge. 2010. Evaluation of 10

plant barcodes in Bryophyta (Mosses). J. Syst. Evol. 48: 3646.

Meier, R., G.Y. Zhang, and F. Ali. 2008. The use of mean instead

of smallest interspecific distances exaggerates the size of the "barcoding gap" and leads to misidentification. Syst.

Biol. 57: 809-813.

Milne, R.I. 2004. Phylogeny and biogeography of Rhododendron subsection Pontica, a group with a Tertiary relict distribution. Mol. Phyloge. Evol. 33: 389-401.

Milne, R.I., C. Davies, R. Prickett, L.H. Inns, and D.F. Chamberlain. 2010. Phylogeny of Rhododendron subgenus Hy-menanthes based on chloroplast DNA markers: between-lineage hybridisation during adaptive radiation? Plant Syst. Evol. 285: 234-244.

Milne, R.I. and R.J. Abbott. 2002. The origin and evolution of tertiary relict floras. Adv. Bot. Res. 38: 282-309.

Nei, M. and S. Kumar. 2000. Molecular Evolution and Phyloge-netics. Oxford University Press, Oxford.

Newmaster, S.G. and S. Ragupathy. 2009. Testing plant barcod-ing in a sister species complex of pantropical Acacia (Mi-mosoideae, Fabaceae). Mol. Ecol. Res. 9: 172-180.

Newmaster, S.G., A.J. Fazekas, R.A.D. Steeves, and J. Janovec. 2008. Testing candidate plant barcode regions in the Myris-

ticaceae. Mol. Ecol. Res. 8: 480-490.

Pennisi, E. 2007. Wanted: a barcode for plants. Science 318:

Pettengill, J.B. and M.C. Neel. 2010. An evaluation of candidate plant DNA barcodes and assignment methods in diagnosing 29 species in the genus Agalinis (Orobanchaceae). Amer. J.

Bot. 97: 1391-1406.

Qin, H.N. and C. Phengklai. 2007. Nyssaceae. In Z.Y. Wu, P.H. Raven and D.Y. Hong (eds.), Flora of China 13. Science Press, Beijing; Missouri Botanical Garden Press, St. Louis,

pp. 300-303.

Ragupathy, S., S.G. Newmaster, V. Balasubramaniam, and M. Murugesan. 2009. DNA barcoding discriminates a new cryptic grass species revealed in an ethnobotany study by the hill tribes of the Western Ghats in southern India. Mol. Ecol. Res. 9: 439-457.

Ratnasingham, S. and P.D.N. Hebert. 2007. BOLD: the Barcode of Life Data System (http://wwwbarcodinglifeorg). Mol. Ecol. Notes 7: 355-364.

Sass, C., D.P. Little, D.W. Stevenson, and C.D. Specht. 2007. DNA barcoding in the Cycadales: testing the potential of proposed barcoding markers for species identification of cycads. PLoS ONE 2: e1154.

Savolainen, V., R.S. Cowan, A.P. Vogler, G.K. Roderick, and R. Lane. 2005. Towards writing the encyclopedia of life: an introduction to DNA barcoding. Phil. Trans. R. Soc. Lond. B. 360: 1850-1811.

Shaw, J., E.B. Lickey, E.E. Schilling, and R.L. Small. 2007.

tortoise and the hare III. Amer. J. Bot. 94: 275-288.

Starr, J.R., R.F.C. Naczi, and B.N. Chouinard. 2009. Plant DNA barcodes and species resolution in sedges (Carex, Cyper-

aceae). Mol. Ecol. Res. 9: 151-163.

274

Botanical Studies, Vol. 53, 2012

Sun, B.L. 2008. Taxonomy of Nyssa L. (Nyssaceae) and conservation biology of Nyssa yunnanensis, a critically endangered species. Ph.D. Dissertation. Kunming: Kunming Institute of Botany, Chinese Academy of Sciences.

Sundberg, P., E.T. Vodoti, and M. Strand. 2010. DNA barcod-ing should accompany taxonomy - the case of Cerebratulus spp. (Nemertea). Mol. Ecol. Res. 10: 274-281.

Swofford, D.L. 2002. PAUP: phylogenetic analysis using parsimony, version 4.0b10. Sinauer, Sunderland, Massachusetts, USA.

Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4:molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24: 1596-1599.

Tamura, K., M. Nei, and S. Kumar. 2004. Prospects for inferring very large phylogenies by using the neighbor-joining method. Proc. Nat. Acad. Sci. USA 101: 11030-11035.

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. Nucl. Acids Res. 24: 4876-4882.

Tate, J.A. and B.B. Simpson. 2003. Paraphyly of Tarasa (Malvaceae) and diverse origins of the polyploid species. Syst. Bot. 28: 723-737.

Wang, N. 2010. Natural hybrid origin of Rhododendron irrora-tum Franch. (Ericaceae) in Yunnan, China. MSc Thesis. Kunming: Kunming Institute of Botany, Chinese Academy of Sciences.

Ward, R.D., T.S. Zemlak, B.H. Innes, P.R. Last, and P.D.N. He-

bert. 2005. DNA barcoding Australia's fish species. Phil.

Trans. R. Soc. Lond. B. 360: 1847-1857.

Wen, J. and T.F. Stuessy. 1993. The phylogeny and biogeography of Nyssa (Cornaceae). Syst. Bot. 18: 68-79.

Wu, C.Y. and J.R. Fan. 1977. Nyssaceae. In C.Y. Wu (ed.), Flora Yunnanica, Vol. 1. Science Press, Beijing, pp. 288-293.

Wu, C.Y., A.M. Lu, Y.C. Tang, Z.D. Chen, and D.Z. Li (eds.).

2003. The families and genera of angiosperms in China: A comprehensive analysis. Science Press, Beijing.

DNA條碼在藍果樹科（Nyssaceae)植物中的應用及分類探討

王年^1,2 Frédéric M.B. JACQUES³ Richard I. MILNE⁴ 張長芹楊俊波

¹中國科學院昆明植物硏究所

²中國科學院植物硏究所植物資源與生物技術重點實驗室

³中國科學院地理與古植物硏究所古植物與孢！粉學硏究室

⁴英國愛丁堡大學植物分子生物學硏究室

⁵中國科學院昆明植物硏究所中國西南野生植物種植資源庫

DNA條碼，作為物種鑒別工具，在動物中得到廣泛應用。然而對於植物，其標準DNA條碼仍
沒確定。在本研究中，我們評價了四個候選條碼(matK, rbcL, trnH-psbA和ITS)在九種藍果樹科
(Nyssaceae)植物中的應用。結果表明，作為陸地植物核心條碼matK + rbcL在該科中表現不佳而ITS
表現最好。與ITS相比，這幾個葉綠體基因的鑒別率較低，可能由於該科藍果樹屬(Nyssa)存在種間
分類問題。研究還表明，上思藍果樹 (N. shangszeensis)、中華藍果樹 (N. sinensis)、瑞麗藍果樹(N.
shweliensis)以及文山藍果樹 (N. wennshhannennsis) 是否成立有待結合形態學及進化速率較快的基因進一
步確定。

關鍵詞：DNA條碼；ITS ； matK ；藍果樹科；rbcL ； trnH-psbA 。



	272	Botanical Studies, Vol. 53, 2012

	recognized as distinct species from N. sinensis. However, distance thresholds undoubtedly differ between organisms (Pettengill and Neel, 2010), so this 1% threshold can only serve as an approximate guideline, and other explanations for low divergence must be considered. Three possible explanations are: first, inaccurate taxonomy (i.e. they are not truly distinct); second, very recent speciation (where insufficient time has passed to accumulate sequence differentiation); and third, past or present hybridization. Hybridization appears unlikely because no two Nyssa species are known to occur in sympatry in eastern Asia, except for N. javanica and N. sinensis (N. Wang, personal observation). However, one accession of N. javanica has a cpDNA type that is in a different position on the NJ tree from the other two (Figure 3), despite this species being monophyletic for ITS (Figure 2). This could indicate possible chloroplast capture via ancient introgres-sion; however, the difference within N. javanica is only one bp, so homoplasy is also a possible explanation. Recent speciation is plausible for Nyssa, irrespective of whether current taxonomy is accurate. Although Nyssa is an old genus with a long Tertiary history (Edye, 1963; Wen and Stuessy, 1993), this is not incompatible with recent speciation in eastern Asia. For example, Rhododendron is 60 million years old (Milne, 2004), but contains clades of species that are not distinguishable based on ITS sequences (e.g. R. decorum and R. irroratum; Wang, 2010) or particular cpDNA sequences (Milne et al., 2010). Furthermore, Tertiary relict floras contain numerous examples of clades of morphologically similar species within which are nested clades of more distinctive species (Milne and Abbott, 2002); Nyssa fits this pattern. Hence it is likely that N. sinensis, N. shangszeensis, N. shweliensis and N. wenshanensis diverged recently, perhaps as a result of quaternary climatic oscillations. With this being the case, recognition of these four species as distinct from one another must depend upon other factors, such as morphology. Although all four were recognized by both Wen (1993) and Qin and Phengklai (2007), our data indicates that some re-evaluation of these taxa might be necessary. The non-monophyly of N. shweliensis for ITS might indicate that this comprises two cryptic species, but the cpDNA provides no evidence either way so this possibility in particular requires further investigation. Regarding biogeography, both cpDNA and ITS data indicate that East Asian Nyssa do not form a monophyletic group, because N. javanica and N. yunnanensis are basal in both ITS and cpDNA analyses, with the American N. sylvatica sister to the remaining species. This correlates well with the analysis of Eyde (1963) who concluded based on fossil morphology that N. sinensis was more closely-related to N. sylvatica than N. javanica. Clearly the inclusion of only one non-Asian species in our analysis limits what can be inferred, but the simplest explanation for these results would be that Nyssa originated in eastern Asia and that subsequently one (or more) lineages moved into the Americas.	Conclusion Our data indicates that Nyssa in eastern Asia comprises three relatively old lineages, i.e. N. javanica, N. yunnanen-sis and all other species (bearing in mind that relationships of N. leptophylla remain uncertain), and that recent diversification within the latter clade has given rise to much of the morphological diversity now evident in eastern Asian Nyssa. Our data hence rejects the grouping together of N. yunnanensis, N. wenshanensis and N. javanica as a species complex, and instead indicates that N. wenshanensis, N. shangszeensis, N. shweliensis, N. sinensis and possibly N. leptophylla might form a complex, either of recently diverged species or of taxa that are not truly distinct at specific level. In spite of potential taxonomical problems in Nyssa, among the studied loci, ITS appeared to be the best barcode candidate for Nyssaceae. To get a whole picture of the taxonomy of Nyssa, a detailed morphological study of Nyssa combined with more rapidly evolving loci is needed. Acknowledgements. We sincerely thank Yu-Xiao Zhang and Zhao-Ming Cai, at Kunming Institute of Botany, Chinese Academy of Sciences (CAS), for their help with laboratory work and data analysis. The research was funded by the Natural Science Foundation of China (No. 30770139), the Bureau of Science and Technology of Yunnan, China (No. 2009BB001), the Large-Scale Scientific Facilities of the Chinese Academy of Sciences (No. 2009-LSF-GBOWS-01), GBOWS and by CAS Young Scientists Fel-lowship (2009YB1-13) and NSFC Research Fellowship for International Young Scientists (40950110338) to F. M. B. Jacques. LITERATURE CITED Airy Shaw, H.K. 1969. An overlooked Chinese Nyssa. Kew Bull. 23: 311. CBOL Plant Working Group. 2009. A DNA barcode for land plants. Proc. Nat. Acad. Sci. USA 106: 12794-12797. Chase, M.W., N. Salamin, M. Wilkinson, J.M. Dunwell, R.P. Ke-sanakurthi, N. Haidar, and V. Savolainen. 2005. Land plants and DNA barcodes: short-term and long-term goals. Phil. Trans. R. Soc. Lond. B. 360: 1889-1895. Chen, S.L., H. Yao, J.P. Han, C. Liu, J.Y. Song, L.C. Shi, Y.J. Zhu, X.Y. Ma, T. Gao, X.H. Pang, K. Luo, Y. Li, X.W. Li, X.C. Jia, Y.L. Lin, and C. Leon. 2010. Validation of the ITS2 region as a novel DNA barcode for identifying medicinal plant species. PloS One. 5: 1-8. Cronquist, A. 1981. An Integerated System of Classification of Flowering Plants. Columbia University, New York. Doyle, J.J. and J.L. Doyle. 1987. A rapid DNA isolation proce-dure for small quantities of fresh leaf tissue. Phytochem. Bull. 19: 11-15. Dunning, L.T. and V Savolainen. 2010. Broad-scale amplifica-