全 文 ：Journal of Systematics and Evolution 46 (4): 563–572 (2008) doi: 10.3724/SP.J.1002.2008.06197
(formerly Acta Phytotaxonomica Sinica) http://www.plantsystematics.com
Phylogenetic analysis of Paeonia sect. Moutan (Paeoniaceae) based on
multiple DNA fragments and morphological data
1,2Xuan ZHAO 1,2,3Zhi-Qin ZHOU* 2Qi-Bing LIN 2Kai-Yu PAN 1Ming-Yang LI
1 (College of Horticulture and Landscape Architecture, Southwest University, Chongqing 400716, China)
2 (State Key Laboratory of Systematic and Evolutionary Botany, Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China)
3 (Citrus Research Institute, Chinese Academy of Agricultural Sciences, Chongqing 400712, China)
Abstract Tree peony, being crowned the title “King of Flowers” in China, is of great medicinal, ornamental, and
economic values. In the present study, the phylogeny of the wild tree peony species (section Moutan, Paeonia,
Paeoniaceae), represented by twelve accessions collected from all eight species in the section, was investigated
based on the DNA sequence in five DNA fragments from both nuclear (Adh1A, Adh2 and GPAT) and chloroplast
(trnS-trnG and rps16-trnQ) genomes, as well as morphological characters. Both maximum parsimony (MP) and
Bayesian inference of phylogeny (BI) trees were reconstructed based on the combined data of the DNA sequences
and morphological data, respectively. The MP and BI trees have the similar topology, and the sect. Moutan clearly
branched into two clades. One clade consists of two species, P. delavayi and P. ludlowii, corresponding to the
subsect. Delavayanae, and another clade is composed of other six species. Within the second clade, the six species
can be divided into three subclades consisting of P. rockii and P. decomposita, P. jishanensis and P. qiui, P.
suffruticosa and P. ostii, respectively. Among the three subclades, P. jishanensis/P. qiui is most closely related to
P. suffruticosa/P. ostii. These results provide up to date the clearest picture of the phylogeny of wild tree peony
species in the sect. Moutan.
Key words chloroplast DNA, morphological character, nuclear DNA, Paeonia sect. Moutan, phylogeny.
The cultivated tree peonies have been used as
medicine for over 2000 years in China and 500 years
in Europe, respectively (Foster & Yue, 1992), and
now cultivated all over China and the temperate
regions of the world for their medicinal, ornamental,
and economic values (Hong & Pan, 1999a; Lan et al.,
2002).
Tree peony belongs to the section Moutan DC. of
the genus Paeonia L. (Paeoniaceae) (Stern, 1946; Pan,
1979). In addition to this woody group, there are two
herbaceous sections in this genus, sect. Paeonia and
sect. Onaepia (Stern, 1946; Pan, 1995). The genus
Paeonia consists of more than 30 species, distributed
widely in the temperate region of the world (Pan,
1995). The section Onaepia has only two species
endemic to North America, whereas sect. Paeonia
comprises about 22 species found in Europe,
North-west Africa and Asia, spreading from Portugal
and Morocco to Japan (Stern, 1946; Pan, 1995). The
sect. Moutan, following a recent classification system
of Hong & Pan (1999a), contains eight species and
three of them each have two subspecies, distributed in
southwestern, central and northern region of China
(Hong & Pan, 1999a).
In the past decade, important progress has been
made in the taxonomy of tree peony (Xi, 1984; Hong
et al., 1988; Hong & Pan, 1999a; Zhou, 2006). How-
ever, although the interspecific relationships of wild
tree peonies have been previously investigated using
morphological data (Hong, 1997a; Zhou et al., 2003),
molecular marker (Zou et al., 1999), and DNA frag-
ments from different genomes (Sang et al., 1995,
1997a, b; Sang & Zhang, 1999; Ferguson & Sang,
2001; Tank & Sang, 2001; Lin et al., 2004; Zhao et
al., 2004), the species phylogeny of the sect. Moutan
is still poorly understood. Especially, the interspecific
relationship among the six species (P. suffruticosa, P.
ostii, P. jishanesis, P. qiui, P. rockii, and P. decompo-
sita) remained unresolved (Hong & Pan, 1999b; Lin et
al., 2004; Zhao et al., 2004).
Using DNA molecular marker and gene se-
quences from both nuclear and cytoplasmic genomes
has had an enormous impact on studies of plant
phylogenetics and systematics (Sang, 2002). How-
ever, the problem of the reconstruction of phylogeny
of taxa at lower taxonomic level, especially the
closely related species, remains unresolved (Sang,
2002; Grob et al., 2004). For examples, nuclear ribo-
somal DNA regions, especially the internal tran-
scribed spacers (ITS1 and ITS2) are applied exten-
sively for phylogeny reconstruction at lower
———————————
Received: 5 December 2006 Accepted: 26 March 2007
* Authors for correspondence. E-mail: zqzhoubj@yahoo.com.
Journal of Systematics and Evolution Vol. 46 No. 4 2008 564
taxonomic levels. ITS sequences, however, have their
own problems when applied to phylogenetic studies,
such as extensive length variations between copies,
paralogy problems, and/or lack of resolving power
(Grob et al., 2004). The mitochondrial nad1 intron 2
has been successfully used in studying population
structure and phylogenetic relationships among
closely related taxa (Gugerli et al., 2001). But for their
low substitution rates and high levels of gene rear-
rangements, plant mitochondrial genes have been
regarded as less useful for systematic studies than
animal mtDNA (Sanjur et al., 2002). The chloroplast
genome shares many features with animal mtDNA,
such as the conserved gene order and their high levels
of sequence variation in the noncoding parts of the
genome (Provan et al., 2001). Therefore, the cpDNA
sequences were widely used in systematic studies.
Single- or low-copy nuclear genes have high rates of
substitution. The phylogenetic utility of this kind of
genes has been investigated in various plant taxa, such
as granule-bound starch synthase (GBSSI or waxy) in
Poaceae (Mason-Gamer et al., 1998), vicilin in Ster-
culiaceae (Whitlock & Baum, 1999), malate synthase
in Arecaceae (Lewis & Doyle, 2001), and Alcohol
dehydrogenase (Adh) in Oryza (Ge et al., 1999; Guo
& Ge, 2005). Especially, the single-copy nuclear gene
glyceraldehyde 3-phosphate dehydrogenase (G3pdh)
was successfully used in the phylogeographic study of
Cassva and its close relatives (Olsen & Schaal, 1999).
Furthermore, just as Zhang & Hewitt (2003) has
pointed out that because of its general higher poly-
morphism and the ease of designing primer in the
flanking coding sequences (normally conserved
among closely related taxa), intron will be probably
the most useful nuclear markers in the near future. The
potential utility of the intron of low-copy nuclear
genes in phylogenetic reconstruction at lower taxo-
nomic levels has been explained completely (Grob et
al., 2004).
Adh is a metabolic enzyme responsible for the
interconversion of ethanol and acetaldehyde (Freeling
& Bennett, 1985). Adh gene family usually has two or
three loci in a broad array of angiosperm species
(Clegg et al., 1997). The DNA divergence and mo-
lecular evolution of this gene family have been inves-
tigated in a broad array of plants, such as Brassicaceae
(Koch et al., 2000), Gossypium (Small et al., 1998),
Paeonia (Sang et al., 1997b), Pinus banksiana (Perry
& Furnier, 1996), and Poaceae (grass) (Gaut, 1999;
Ge et al., 1999; Guo & Ge, 2005). In the genus Paeo-
nia, Adh gene family has three members, Adh1A,
Adh1B and Adh2, among them Adh1B was merely
found in sect. Moutan (Sang et al., 1997b). Sang et al.
(1997b) reconstructed the phylogeny of 11 putative
nonhybrid species and investigated the Adh gene
evolution in the genus Paeonia. Their study showed
that Adh gene is better than ITS and matK genes in
resolving the interspecific relations among the species
in the genus Paeonia. However, in this study the
species of section Moutan were poorly sampled.
Similar to Adh gene, Glycerol-3-phosphate acyltrans-
ferase (GPAT) is an essential enzyme and has been
used for phylogenetic analysis, which is utilized in the
catalysis of the initial step of glycerolipid synthesis in
the cells of all higher organisms (Nishida et al., 1993).
In plant cell, there are three types of GPATs that differ
in their subcellular location (chloroplasts, mitochon-
dria and cytoplasm) and substrate specificity (Tank &
Sang, 2001). Among the genes coding for different
GPATs, the nuclear-encoded chloroplast-expressed
GPAT gene has been found to be single-copy in
several distantly related angiosperm families (Tank &
Sang, 2001). In the genus Paeonia, Tank & Sang
(2001) investigated the DNA variation in nuclear-
encoded chloroplast-expressed GPAT gene, and suc-
cessfully reconstructed the phylogeny of 13 Paeonia
species. Their study revealed that the Paeonia GPAT
gene contains a large intron of more than 2 kb with a
high level of DNA variation rate (Tank & Sang, 2001).
Although many genes have been used effectively
to address the systematic problems, the information
from single gene usually was insufficient to resolve
the problem of interspecific relationship of sect.
Moutan (Sang et al., 1997a, b; Sang & Zhang, 1999;
Lin et al., 2004; Zhao et al., 2004). In contrast, com-
bined analysis of multiple data sets has become an
effective way to increase the resolving power and the
reliability of phylogenetic reconstruction (Hillis &
Huelsenbeck, 1995; Wendel & Doyle, 1998; Ge et al.,
1999; Cronn et al., 2002; Guo & Ge, 2005; James et
al., 2006).
In the present study, we reconstruct the phylog-
eny of the sect. Moutan using five gene fragments
from both nuclear (Adh1A, Adh2, and GPAT) and
chloroplast (trnS-trnG and rps16-trnQ) genomes, as
well as morphological characters. Our objective is to
clarify unresolved problems concerning interpecific
relationship within Paeonia sect. Moutan.
1 Material and methods
1.1 Material
In this study, twelve accessions represent eight
wild species of Paeonia sect. Moutan. Paeonia
ZHAO et al.: Phylogenetic analysis of Paeonia sect. Moutan 565
lactiflora was sampled as outgroup, which belongs to
Paeonia sect. Paeonia. All vouchers are deposited at
the Herbarium, Institute of Botany, Chinese Academy
of Sciences, Beijing, China (PE). The details of the
sampled species, accession numbers or vouchers and
origins are listed in Table 1. Some sequences of GPAT
gene (AY016249), Adh gene (AF009042, AF009043,
AF009044, AF009049, AF009058, AF009060,
AF009061 and AF009068), and rps16-trnQ gene
(DQ313804) were downloaded from GenBank. All the
sequences produced in this paper are deposited in
GenBank (GenBank Accession Number: EF520815–
EF520868).
1.2 Methods
1.2.1 Morphological data The data set of mor-
phology was taken from the matrix presented by Zhou
et al. (2003). In the present study, we selected 12
samples, including yinpingmudan-HNSX, yinping-
mudan-AHCH, jishanensis-ShXYA, jishanesis-SXJS,
decomposita-MEK, rotundiloba-MX, ostii-HNNX,
qiui-HBBK, taibaishanica-ShXMX, rockii-HNNX,
ludlo- wii-ML1, and delavayi-XZLZ.
1.2.2 DNA isolation, amplification, and sequenc-
ing Total DNA was isolated from silica-gel dried
leaves using the CTAB method as described by Doyle
and Doyle (1987). Amplifications were performed in a
Peltier Thermal Cycler (PTC-200, PE). The nuclear
Adh and GPAT gene were amplified and sequenced as
described by Lin et al. (2004) and Zhao et al. (2004),
respectively. PCR cycling parameters for the two
chloroplast genes (trnS-trnG and rps16-trnQ) were
similar, with an initial 4 min at 70 ℃, followed by 2
cycles of 1 min at 94 ℃, 20 s at 52 ℃, and 1.5 min
at 72 ℃; after that, 35 cycles of 20 s at 94 ℃, 20 s at
52 ℃, and 1.5 min at 72 ℃ were conducted, with a
final extension time of 10 min at 72 ℃. All the
amplifying and internal sequencing primers are listed
in Table 2.
PCR products were electrophoresed on and ex-
cised from 1.5% agarose gel, then purified using DNA
Purification Kit (Pharmacia). Sequencing was done on
an ABI 377 (Applied Biosystems, Foster City, CA) or
MegaBASE 1000 automatic DNA sequencer (Amer-
sham Biosciences, Buckinghamshire, UK).
1.2.3 Data analysis Morphological characters
were treated following Zhou et al (2003).
Sequences were aligned using ClustalX version
1.81 (Thompson et al., 1997) and refined manually.
The partition-homogeneity test, implemented in
PAUP* version 4.0b10 (Swofford, 2002), was used to
evaluate the congruence between different data sets
(Farris et al., 1995). Replicates were analyzed using
the parsimony criterion, and branch swapping using
tree bisection-reconnection (TBR) was performed and
one tree held at each step during the stepwise addition.
However, the test was not used as a criterion to decide
whether we are allowed to combine data sets. Because
there is no agreement on how to treat different data
set, whether analyze them as a combined data set or
analyze them separately one by one (Kluge, 1989;
Miyamoto & Fitch, 1995; Farris et al., 1995);

Table 1 Source of materials
No. Taxon Voucher/ID Origin
1 Paeonia ludlowii (Taylor & Stem) D. Y. Hong
(大花黄牡丹)
S. L. Zhou (周世良) H02124 Nyingchi, Xizang, China (西藏林芝)
2 P. delavayi Franch. (滇牡丹) S. L. Zhou (周世良) H02123 Nyingchi, Xizang, China (西藏林芝)
3 P. decomposita Hand.-Mazz. ssp. decomposita
(四川牡丹)
Z. Q. Zhou (周志钦) H03001 Barkam, Sichuan, China (四川马尔康)
4 P. decomposita ssp. rotundiloba D. Y. Hong
(圆裂四川牡丹)
D. Y. Hong et al. (洪德元等) H02016 Pengzhou, Sichuan, China (四川彭州)
5 P. rockii (S. G. Haw & Lauener) T. Hong & J.
J. Li ex D. Y. Hong ssp. rockii (紫斑牡丹)
D. Y. Hong et al. (洪德元等) H02121 (H97015) Neixiang, Henan, China (河南内乡)
6 P. rockii ssp. taibaishanica D. Y. Hong
(太白紫斑牡丹)
D. Y. Hong et al. (洪德元等) H02122 (H97058) Mt. Taibai, Shaanxi, China (陕西太白山)
7 P. ostii T. Hong & J. X. Zhang (凤丹) D. Y. Hong & K. Y. Pan (洪德元, 潘开玉)
H02106
Lushi, Henan, China (河南卢氏)
8

P. suffruticosa ssp. yinpingmudan D. Y. Hong,
K.Y. Pan & Z. W. Xie (银屏牡丹)
K. Y. Pan & Z. W. Xie (潘开玉, 谢中稳)
H02117 (H9701)
D. Y. Hong (洪德元) H02118 (H97010)
Chaohu, Anhui, China (安徽巢湖)

Song Xian, Henan, China (河南嵩县)
9 P. qiui Y. L. Pei & D. Y. Hong (卵叶牡丹) Z. Q. Zhou (周志钦) H02080 Lanzhou, Gansu, China (甘肃兰州)
10 P. jishanensis T. Hong & W. Z. Zhao
(矮牡丹)
Z. Q. Zhou (周志钦) H02077
D. Y. Hong et al. (洪德元等) H02119 (H97066)
Lanzhou, Gansu, China (甘肃兰州)
Yan’an, Shaanxi, China (陕西延安)
Journal of Systematics and Evolution Vol. 46 No. 4 2008 566
Table 2 The primers used for PCR amplification and sequencing
Primer Sequence
rps16-trnQ
rps16
trnQ
trnQ -r*

5′- CGTTGCTTTCTACCACATCG
5′- TTACTCGGAGGTTCGAATCC
5′- CCCTCCCTCACTTCATATTG
trnS-trnG
trnS
trnG
trnG-r*

5′- GCCGCTTTAGTCCACTCAGC
5′- GAACGAATCACACTTTTACCAC
5′- TCGTCAGGGAACTTAACGAG
* represents the internal primer for sequence.

moreover, it is legitimate to combine data sets al-
though different sources of data may yield alternative
phylogenetic results (Farris, 1997; Soltis et al., 1997).
Maximum parsimony (MP) analyses were con-
ducted using PAUP* version 4.0b10 (Swofford,
2002). All characters were equally weighted, gaps
were treated as missing, and character states were
treated as unordered. Heuristic search was performed
with MULPARS option, tree bisection-reconnection
(TBR) branch swapping, and random stepwise addi-
tion with 1000 replicates. Topological robustness was
assessed by bootstrap analysis with 1000 replicates
using simple taxon addition (Felsenstein, 1985).
An appropriate nucleotide substitution model was
determined using Modeltest version 3.06 (Posada &
Crandall, 1998) for each data set. The models were
chosen according to the Hierarchical Likelihood Ratio
Test (LRT) and then used for subsequent Bayesian
analysis. Bayesian inference (BI) was conducted using
MrBayes version 3.1.2 (Huelsenbeck & Ronquist,
2001). One cold and three incrementally heated
Markov Chain Monte Carlo (MCMC) chains were run
for 1 million generations, with trees sampled every
100 generation, using random tree as its starting point
and a temperature parameter value of 0.2 (the default
setting of MrBayes). For each data set, MCMC runs
were repeated twice as a safeguard against spurious
results. The first 1000 trees were discarded as burn-in,
and the remaining trees were used to construct Bayes-
ian trees. Examination of the log-likelihoods and the
observed consistency between runs suggested that the
burn-in periods were sufficiently long enough for
chains to have become stationary.
2 Results
2.1 Sequence characteristics
In the present study, the amplified sequences of
Adh1 and Adh2 include four introns and five exons.
The resulting fragment of Adh1A ranged in length
from 1141 to 1167 bp with an aligned length of 1191
bp, among which 93 sites are variable and 23 sites are
parsimony informative. The sequences of Adh2 vary
from 1095 to 1149 bp in length. The final alignment
of Adh2 is 1185 bp, with 135 sites variable and 31
sites informative. The amplified segment of GPAT
gene is a big intron between exon 5 and 6. The se-
quence length varies from 1848 to 1916 bp, and the
final alignment of GPAT is 1962 bp, including 144
variable sites and 54 informative sites. Among the
three nuclear gene segments, the informative site of
GPAT is the highest (2.75%), Adh2 is 2.62%, and
Adh1A is the lowest (1.93%). Chloroplast genes
rps16-trnQ and trnS-trnG have less variable and
informative sites than the nuclear genes, with the
informative sites 1.06% and 0.97%, respectively. The
characteristics of each sampled sequence are presented
in Table 3.
Each data set was identified to fit a correspond-
ing model using Modeltest ver. 3.06 (Posada & Cran-
dall, 1998). Nuclear genes Adh1A, Adh2, and GPAT
have the same evolutionary model HKY+G, while the
models of chloroplast genes rps16-trnQ and trnS-trnG
are F81.
2.2 Phylogenetic analysis
Due to limited information and low resolution of
single gene data set or morphological data set (data
not shown), in the present study, we only analyzed
data in combined data sets. Based on the information
of chloroplast genes, the phylogenetic analysis of the
combined data set (rps16-trnQ and trnS-trnG,
P=1.000 for partition homogeneity tests) suggests that
these species can be divided into two clades: one clade
includes P. ludlowii and P. delavayi; the second clade
consists of other six species (data not shown). The
Bayesian analysis generates a similar topology, with
only a few differences in bootstrapping support (BS)
and Bayesian posterior probability for some clades
(data not shown).
Three nuclear genes (hereafter represented as
data1, P<0.001), three nuclear genes plus two chloro-
plast genes (hereafter represented as data2, P<0.001),
and all DNA sequence data with morphological data
(hereafter represented as data3, P<0.001), were ana-
lyzed one by one (Table 4). The data1 generated two
most parsimonious trees (Fig. 1), the data2 generated
one (Fig. 2), and one most parsimonious tree (Fig. 3)
was obtained from the data3.
The topologies of the trees generated from the
different combined data sets are significantly congru-
ent with each other. Paeonia sect. Moutan is divided
into two major clades, consistent with the above
results based on combined chloroplast sequence data.
The first clade consists of P. ludlowii and P. delavayi;
ZHAO et al.: Phylogenetic analysis of Paeonia sect. Moutan 567
Table 3 Informative parameters for each DNA fragment
DNA fragment Aligned length variable site (%) Informative site (%) GC (%) Model Nst Rate
rps16-trnQ 1229 40 (3.25) 13 (1.06) 29.85 F81 1 equal
trnS-trnG 512 6 (1.17) 5 (0.97) 30.01 F81 1 equal
Adh1A 1191 116 (9.74) 23 (1.93) 39.75 HKY+G 2 gamma
Adh2 1185 166 (14.00) 31 (2.62) 40.68 HKY+G 2 gamma
GPAT 1962 198 (10.09) 54 (2.75) 35.18 HKY+G 2 gamma
morphological character 25 24 (96.00) 22 (88.00) / / / /
Summary 7309 664 (9.08) 194(2.65) / / / /


Table 4 Results conducted by partition homogeneity test for various
combinations of the data sets
Combined data sets PHT value
rps16-trnQ + trnS-trnG 1.000
Adh1A + Adh2 + GPAT 0.001
Adh1A + Adh2 + GPAT + rps16-trnQ + trnS-trnG 0.001
Adh1A + Adh2 + GPAT+ rps16-trnQ + trnS-trnG +
morphology
0.001

the second clade includes other six species. Within the
second clade, P. rockii is closely related to P. decom-
posita; while other four species form another sub-
clade, and P. jishanensis/P. qiui and P. suffruticosa/P.
ostii was sister to each other. Although the topologies
are congruent between the three different combined
data sets, there are some differences among them: (1)
in P. jishanensis/P. qiui clade, two samples of P.
jishanensis grouped in the data2 and data3, but not in
data1; (2) within P. rockii/P. decomposita clade, two
subspecies of P. rockii grouped in the data1 and the
data3, but not in the data2. In addition, it is also
apparently that the BS value is increasing with the
addition of separate data set to combined data set. For
each node of the data1, the average BS is 82% (Fig.
1). The average BS for each node of phylogeny de-
rived from the data2 was 85% (Fig. 2), while that of
data3 get an average support of 91.4% (Fig. 3). The
trend of Bayesian posterior probability is the same as
that of BS.



Fig. 1. The 50% majority-rule consensus tree of two MP trees based on Adh1A, Adh2 and GPAT (Tree length = 559, CI = 0.8927, RI = 0.7196). The
numbers near branches are bootstrap percentages followed by Bayesian posterior probabilities.
Journal of Systematics and Evolution Vol. 46 No. 4 2008 568


Fig. 2. Single MP tree based on rps16-trnQ, trnS-trnG, Adh1A, Adh2 and GPAT (Tree length = 613, CI = 0.8907, RI = 0.7220). The numbers near
branches are bootstrap percentages followed by Bayesian posterior probabilities.




Fig. 3. One MP tree based on rps16-trnQ, trnS-trnG, Adh1A, Adh2, GPAT and morphological data (Tree length = 679, CI = 0.8616, RI = 0.6846).
The numbers above branches are bootstrap percentages followed by Bayesian posterior probabilities.
ZHAO et al.: Phylogenetic analysis of Paeonia sect. Moutan 569
3 Discussion
3.1 Phylogenetic relationships among species of
Paeonia sect. Moutan
In previous studies, due to the incomplete sam-
pling or limited informative characters (Zou et al.,
1999; Yu et al., 1998; Tank & Sang, 2001), the phy-
logenetic relationships of Paeonia sect. Moutan is
ambiguous. Recently, both morphological (Zhou et
al., 2003) and molecular (Lin et al., 2004; Zhao et al.,
2004) evidence was used to investigate the phyloge-
netic relationship of Paeonia sect. Moutan based on
the new classification system proposed by Hong and
Pan (1999a). These studies made some impressive
advances, but interspecific relationship of Paeonia
sect. Moutan is still not resolved because of the lim-
ited phylogenetic information (Sang et al., 1997a, b;
Zhou et al., 2003; Lin et al., 2004; Zhao et al., 2004).
For the first time, in this study, we use both multiple
genes and morphological characters to reconstruct the
phylogeny of Paeonia sect. Moutan.
There are eight wild species in Paeonia sect.
Moutan, distributed in a wide range from Southwest to
North China (Pan, 1995; Hong & Pan, 1999a). Within
the eight species, P. ludlowii and P. delavayi are
distributed in the west of Mt. Daxueshan, while other
six species grow in the east of Mt. Daxueshan (Hong
& Pan, 1999a). Based on the morphological charac-
ters, P. ludlowii is similar to P. delavayi, their flowers
are terminal and axillary, disc fleshy and short. In
contrast, the flowers of other six species are solitary,
disc leathery (Hong, 1997b; Hong & Pan, 1999a).
Recently, phylogenetic analyses suggested that Paeo-
nia sect. Moutan is composed of two clades, one clade
includes P. ludlowii and P. delavayi, another clade
includes other six species (Zhou et al., 2003; Lin et
al., 2004; Zhao et al., 2004). These results are corre-
spondent to the treatment of two subsections by Stern
(1946). In the present study, our results (Figs. 1–3)
support the two-subsection treatment, namely, sub-
sect. Delavayanae and subsect. Vaginatae.
The phylogenetic relationship of subsect. Vagi-
natae has been in dispute in the past several decades.
After analyzing the morphological characters, Hong
and Pan (1999b) treated five species as the complex P.
suffruticosa, including P. rockii, P. jishanensis, P.
qiui, P. suffruticosa, and P. ostii, due to the entirety
disc leathery, whereas P. decomposita is distinguished
from them by its partially disc leathery. Zhou et al.
(2003) used twenty-five morphological characters to
study the phylogenetic relationships of eight wild
species of Paeonia sect. Moutan, the result indicated
that P. decomposita grouped with the complex P.
suffruticosa and located at the basal position of this
clade, thus supported the treatment of the complex P.
suffruticosa. But in the recent DNA sequence analy-
ses, Lin et al. (2004) and Zhao et al. (2004) found that
P. decomposita was grouped with P. rockii, and other
four species formed a clade with higher BS. Other
previous studies based on molecular markers also got
the similar results (Sang et al., 1997a, b; Sang &
Zhang, 1999; Zou et al., 1999; Tank & Sang, 2001). In
the present study, the combined analysis of molecular
and morphological data supports that P. decomposita
is grouped with P. rockii, and other four species
formed another clade (Fig. 3). Therefore, we can
conclude that P. decomposita is closely related to P.
rockii comparing to other four species in subsect.
Vaginatae.
Paeonia decomposita and P. rockii have two
subspecies, respectively. In previous studies, the
relationships among the four subspecies conflicted
with each other (Hong et al., 1996; Zhou et al., 2003;
Lin et al., 2004; Zhao et al., 2004). Our results suggest
that P. decomposita ssp. rotundiloba was sister to P.
decomposita ssp. decomposita and two subspecies of
P. rockii (Figs. 1–3). Therefore, we can conclude that
P. decomposita ssp. decomposita is closely related to
P. rockii. Comparing the morphological characters, P.
decomposita ssp. rotundiloba has 2–5 carpels, usually
3 or 4, while P. decomposita ssp. decomposita has a
stable number (5) of carpel (Hong & Pan, 1999a).
Therefore, both morphological and molecular evi-
dence supports that P. decomposita ssp. decomposita
is closely related to P. rockii.
In summary, Paeonia sect. Moutan can be di-
vided into two subsections, the subsect. Delavayanae
and subsect. Vaginatae. Subsection Delavayanae
consists of two species, P. delavayi and P. ludlowii;
subsect. Vaginatae includes other six species. Within
the subsection Vaginatae, P. rockii is closely related
to P. decomposita; while other four species form a
clade, and P. jishanensis/P. qiui and P. suffruticosa/P.
ostii were sister to each other.
3.2 The significance of the analysis based on com-
bined data
The primary aim of phylogenetic analysis is to
infer the relationship among species, namely species
tree. As we know, gene tree derived from single gene
is difficult to give a correct species tree for the limited
information at lower taxonomic group (Doyle & Gaut,
2000). In addition, different data sets usually generate
different phylogenetic relationships (Doyle & Gaut,
2000), such as among different molecular data sets or
between molecular data set and morphological
Journal of Systematics and Evolution Vol. 46 No. 4 2008 570
character data. Therefore, using the combined genes
data set to infer phylogenetic relationship has become
a widely accepted trend (Hillis & Huelsenbeck, 1995;
Wendel & Doyle, 1998; Ge et al., 1999; Cronn et al.,
2002; Guo & Ge, 2005; James et al., 2006).
Up to now there are three commonly used meth-
ods to deal with the multiple data sets. Firstly, always
combine the data (Kluge, 1989). In particular, all of
the available taxa (e.g. fossil and living) should be
combined in a phylogenetic analysis, as well as all of
the available characters. Kluge (1989) believes the
total evidence solution is sought because it maximizes
the “informativeness” and “explanatory power” of the
character data used in the analysis. Secondly, analyze
different data set separatly (Miyamoto & Fitch, 1995).
That means analyzing the different data independ-
ently. Thirdly, conditional data combination. Namely,
partitions were subjected to a statistical test of “ho-
mogeneity”: Heterogeneous data partitions are those
that result in significantly different estimates about
phylogeny (differences beyond that expected from
sampling error) when analyzed separately (Farris et
al., 1995). If the test result is non-significant, then the
data can be combined. The three methods have their
respective advantages, and the third method, condi-
tional data combination, essentially was the interme-
diate between other two methods. Since the criterion
of the PHT value does not have a common agreement
and the theoretical basis of PHT is not mature (Cun-
ningham, 1997); moreover, it is believed it is legiti-
mate to combine data sets although different sources
of data may yield alternative phylogenetic results
(Farris, 1997; Soltis et al., 1997). Therefore, in the
present study, we combined the data and analyzed
them together even some combined data sets have a
lower value of partition-homogeneity test.
In this study, the information of single gene and
combined chloroplast genes is too limited to clarify
the phylogenetic relationship among all the species in
Paeonia sect. Moutan. With the addition of separate
data set to the combined data set, the number of trees
reconstructed reduced to only single one, and some
ambiguous clade was resolved clearly (Figs. 1–3). At
the same time, the average support of BS and Bayes-
ian posterior probability for each node is improved
apparently. Consequently, based on the comparison of
the phylogenetic relationship resulted from different
combined data sets, it is apparent that the combined
data sets of all partitions give the most robust system-
atic relationships among species of Paeonia sect.
Moutan.
Acknowledgements Comments and suggestions
from anonymous reviewers are sincerely appreciated.
This study was supported by grants from Chinese
Academy of Sciences (KSCX2-SW-108), and the
National Natural Science Foundation of China
(30130030).
References
Clegg MT, Cummings MP, Durbin ML. 1997. The evolution of
plant nuclear genes. Proceedings of the National Academy
of Sciences USA 94: 7791–7798.
Cronn RC, Small RL, Haselkorn T, Wendel JF. 2002. Rapid
diversification of the cotton genus (Gossypium:
Malvaceae) revealed by analysis of sixteen nuclear and
chloroplast genes. American Journal of Botany 89:
707–725.
Cunningham C. 1997. Can three incongruence tests predict
when data should be combined? Molecular Biology and
Evolution 14: 733–740.
Doyle JJ, Doyle JL. 1987. A rapid DNA isolation procedure for
small quantities of fresh leaf tissue. Phytochemical
Bulletin 19: 11–15.
Doyle JJ, Gaut BS. 2000. Evolution of genes and taxa: a primer.
Plant Molecular Biology 42: 1–23.
Farris JS. 1997. Combinability vs congruence. Cladistics 13:
170.
Farris JS, Kallersjo M, Kluge AG, Bult C. 1995. Testing
significance of incongruence. Cladistics 10: 315–319.
Felsenstein J. 1985. Confidence limits on phylogeny: an
approach using the bootstrap. Evolution 39: 783–791.
Ferguson D, Sang T. 2001. Speciation through homoploid
hybridization between allotetraploids in peonies
(Paeonia). Proceedings of the National Academy of
Sciences USA 98: 3915–3919.
Foster S, Yue CX. 1992. Herbal emissaries: Bringing Chinese
herbs to the West. Rochester, VT: Healing Arts Press.
200–207.
Freeling M, Bennett DC. 1985. Maize Adh1. Annual Review of
Genetics 19: 297–323.
Gaut BS, Peek AS, Morton BR, Duvall MR, Clegg MT. 1999.
Patterns of genetic diversification within the adh gene
family in the grasses (Poaceae). Molecular Biology and
Evolution 16: 1086–1097.
Ge S, Sang T, Lu B-R, Hong D-Y. 1999. Phylogeny of rice
genomes with emphasis on origins of allotetraploid
species. Proceedings of the National Academy of Sciences
USA 96: 14400–14405.
Grob GB, Gravendeel B, Eurlings MC. 2004. Potential
phylogenetic utility of the nuclear
FLORICAULA/LEAFY second intron: comparison with
three chloroplast DNA regions in Amorphophallus
(Araceae). Molecular Phylogentics and Evolution 30:
13–23.
Gugerli F, Sperisen C, Buchler U, Brunner I, Brodbeck S,
Palmer JD, Qiu YL. 2001. The evolutionary split of
Pinaceae from other conifers: evidence from an intron loss
and a multigene phylogeny. Molecular Phylogenetics and
Evolution 21: 167–175.
Guo Y-L, Ge S. 2005. Molecular phylogeny of Oryzeae
(Poaceae) based on DNA sequences from chloroplast,
ZHAO et al.: Phylogenetic analysis of Paeonia sect. Moutan 571
mitochondrial and nuclear genomes. American Journal of
Botany 92: 1548–1558.
Hillis DM, Huelsenbeck JP. 1995. Assessing molecular
phylogenies. Science 267: 255–256.
Hong D-Y. 1997a. Notes on Paeonia decomposita Hand.-Mazz.
Kew Bulletin 52: 957–963.
Hong D-Y. 1997b. Paeonia (Paeoniaceae) in Xizang (Tibet).
Novon 7: 156–161.
Hong D-Y (洪德元), Pan K-Y (潘开玉). 1999a. Taxonomical
history and revision of Paeonia sect. Moudan
(Paeoniaceae). Acta Phytotaxonomica Sinica (植物分类学
报) 37: 351–368.
Hong D-Y, Pan K-Y. 1999b. A revision of the Paeonia
suffruticosa complex (Paeoniaceae). Nordic Journal of
Botany 19: 289–299.
Hong D-Y, Pan K-Y, Pei Y-L. 1996. The identity of Paeonia
decomposita Hand.-Mazz. Taxon 45: 67–69.
Hong D-Y (洪德元), Zhang Z-Y (张志宪), Zhu X-Y (朱相云).
1988. Studies of the genus Paeonia (1)―report of
karyotypes of some wild species in China. Acta
Phytotaxonomica Sinica (植物分类学报) 26: 33–43.
Huelsenbeck JP, Ronquist FR. 2001. MrBayes: Bayesian
inference of phylogeny. Bioinformatics 17: 754–755.
James TY, Kauff F, Schoch CL, Matheny PB, Hofstetter V, Cox
CJ, Celio G, Gueidan C, Fraker E, Miadlikowska J,
Lumbsch HT, Rauhut A, Reeb V, Arnold AE, Amtoft A,
Stajich JE, Hosaka K, Sung GH, Johnson D, O’Rourke B,
Crockett M, Binder M, Curtis JM, Slot JC, Wang Z,
Wilson AW, Schüssler A, Longcore JE, O’Donnell K,
Mozley-Standridge S, Porter D, Letcher PM, Powell MJ,
Taylor JW, White MM, Griffith GW, Davies DR, Humber
RA, Morton JB, Sugiyama J, Rossman AY, Rogers JD,
Pfister DH, Hewitt D, Hansen K, Hambleton S,
Shoemaker RA, Kohlmeyer J, Volkmann-Kohlmeyer B,
Spotts RA, Serdani M, Crous PW, Hughes KW, Matsuura
K, Langer E, Langer G, Untereiner WA, Lücking R, Büdel
B, Geiser DM, Aptroot A, Diederich P, Schmitt I, Schultz
M, Yahr R, Hibbett DS, Lutzoni F, McLaughlin DJ,
Spatafora JW, Vilgalys R. 2006. Reconstructing the early
evolution of fungi using a six-gene phylogeny. Nature 443:
818–822.
Kluge AG. 1989. A concern for evidence and a phylogenetic
hypothesis of relationships among Epicrates (Boidae,
Serpentes). Systematic Zoology 38: 7–25.
Koch MA, Haubold B, Mitchell-Olds T. 2000. Comparative
evolutionary analysis of chalcone synthase and alcohol
dehydrogenase loci in Arabidopsis, Arabis, and related
genera (Brassicaceae). Molecular Biology and Evolution
17: 1483–1498.
Lan B-Q (蓝保卿), Li J-J (李家珏), Duan Q-X (段全绪). 2002.
An encyclopedia of tree peonies in China (中国牡丹全
书 ). Beijing: Chinese Science and Technology Press.
1–310.
Lewis CE, Doyle JJ. 2001. Phylogenetic utility of the nuclear
gene malate synthase in the palm family (Arecaceae).
Molecular Phylogenetics and Evolution 19: 409–420.
Lin Q-B (林启兵), Zhou Z-Q (周志钦), Zhao X (赵宣), Pan
K-Y (潘开玉), Hong D-Y (洪德元). 2004. Interspecific
relationships among the wild species of Paeonia sect.
Moutan DC. based on DNA sequences of Adh gene
family. Acta Horticulturae Sinica ( 园 艺 学 报 ) 31:
627–632.
Mason-Gamer RJ, Weil CF, Kellogg EA. 1998. Granule-bound
starch synthase: structure, function, and phylogenetic
utility. Molecular Biology and Evolution 15: 1658–1673.
Miyamoto MM, Fitch WM. 1995. Testing species phylogenies
and phylogenetic methods with congruence. Systematic
Biology 44: 64–76.
Nishida I, Tasaka Y, Shiraishi H, Murata N. 1993. The gene
and the RNA for the precursor to the plastid-located
glycerol-3-phosphate acyltransferase of Arabidopsis
thaliana. Plant Molecular Biology 21: 267–277.
Olsen KM, Schaal BA. 1999. Evidence on the origin of
cassava: phylogeography of Manihot esculenta.
Proceedings of the National Academy of Sciences USA
96: 5586–5591.
Pan K-Y (潘开玉). 1979. Paeonioideae. In: Flora Reipublicae
Popularis Sinica. Beijing: Science Press. 27: 37–59.
Pan K-Y (潘开玉). 1995. The analysis of distribution pattern in
the Paeoniaceae and its formation. Acta Phytotaxonomica
Sinica (植物分类学报) 33: 340–349.
Perry DJ, Furnier GR. 1996. Pinus banksiana has at least seven
expressed alcohol dehydrogenase genes in two linked
groups. Proceedings of the National Academy of Sciences
USA 93: 13020–13023.
Posada D, Crandall K A. 1998. MODELTEST: testing the
model of DNA substitution. Bioinformatics 14: 817–818.
Provan J, Powell W, Hollingsworth PM. 2001. Chloroplast
microsatellites: new tools for studies in plant ecology and
evolution. Trends in Ecology and Evolution 16: 142–147.
Sang T. 2002. Utility of low-copy nuclear gene sequences in
plant phylogenetics. Critical Reviews in Biochemistry and
Molecular Biology 37: 121–147.
Sang T, Crawford DJ, Stuessy TF. 1995. Documentation of
reticulate evolution in peonies (Paeonia) using internal
transcribed spacer sequences of nuclear ribosomal DNA:
implications for biogeography and concerted evolution.
Proceedings of the National Academy of Sciences USA
92: 6813–6817.
Sang T, Crawford DJ, Stuessy TF. 1997a. Chloroplast DNA
phylogeny, reticulate evolution, and biogeography of
Paeonia (Paeoniaceae). American Journal of Botany 84:
1120–1136
Sang T, Donoghue MJ, Zhang D. 1997b. Evolution of alcohol
dehydrogenase genes in peonies (Paeonia): phylogenetic
relationships of putative nonhybrid species. Molecular
Biology and Evolution 14: 994–1007.
Sang T, Zhang D. 1999. Reconstructing hybrid speciation using
sequences of low copy nuclear genes: hybrid origins of
five Paeonia specis based on Adh gene phylogenies.
Systematic Botany 24: 148–163.
Sanjur OI, Piperno DR, Andres TC, Wessel-Beaver L. 2002.
Phylogenetic relationships among domesticated and wild
species of Cucurbita (Cucurbitaceae) inferred from a
mitochondrial gene: Implications for crop plant evolution
and areas of origin. Proceedings of the National Academy
of Sciences USA 99: 535–540.
Small RL, Ryburn JA, Cronn RC, Seelanan T, Wendel JF.
1998. The tortoise and the hare: choosing between
noncoding plastome and nuclear Adh sequences for
phylogeny reconstruction in a recently diverged plant
Journal of Systematics and Evolution Vol. 46 No. 4 2008 572
group. American Journal of Botany 85: 1301–1315.
Soltis DE, Hibsch-Jetter C, Soltis PS, Chase MW, Farris JS.
1997. Molecular phylogenetic relationships among
angiosperms: an overview based on rbcL and 18S rDNA
sequences. In: Iwatsuki K, Raven PH eds. Evolution and
diversification of land plants. Tokyo: Springer.
Stern FC. 1946. A study of the genus Paeonia. London: The
Royal Horticultural Society. 1–155.
Swofford DL. 2002. PAUP*: Phylogenetic analysis using
parsimony (*and related methods). Version 4.0b10.
Sunderland, MA: Sinauer Associates.
Tank DC, Sang T. 2001. Phylogenetic utility of the glycerol-
3-phosphate acyltransferase gene: evolution and
implications in Paeonia (Paeoniaceae). Molecular
Phylogenetics and Evolution 19: 421–429.
Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins
DG. 1997. The Clustal X windows interface: flexible
strategies for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Research 25: 4876–4882.
Wendel JF, Doyle JJ. 1998. Phylogenetic incongruence:
window into genomes history and molecular evolution. In:
Soltis DE, Soltis PS, Doyle JJ eds. Molecular systematics
of plants. II. DNA sequencing. Boston: Kluwer Academic
Publishers.
Whitlock BA, Baum DA. 1999. Phylogenetic relationships of
Theobroma and Herrania (Sterculiaceae) based on
sequences of the nuclear gene Vicilin. Systematic Botany
24: 128–138.
Xi Y-Z (席以珍). 1984. The pollen morphology and exine
ultrastructure of Paeonia L. in China. Acta Botanica Sinica
(植物学报) 26: 241–246.
Yu L (于玲), He L-X (何丽霞), Li J-J (李家珏), Cheng F-Y (成
仿云). 1998. Comparative studies on protein zones of wild
tree peony species. Acta Horticulturae Sinica (园艺学报)
25: 99–101.
Zhang D-X, Hewitt GM. 2003. Nuclear DNA analyses in
genetic studies of populations: practice, problems, and
prospects. Molecular Ecology 12: 563–584.
Zhao X (赵宣), Zhou Z-Q (周志钦), Lin Q-B (林启兵), Pan
K-Y (潘开玉), Hong D-Y (洪德元). 2004. Molecular
evidence for the interspecific relationships in Paeonia sect.
Moutan: PCR-RFLP and sequence analysis of glycerol-
3-phosphate acyltransferase (GPAT) gene. Acta
Phytotaxonomica Sinica (植物分类学报) 42: 236–244.
Zhou Z-Q. 2006. Taxonomy, geographic distribution and
ecological habits of tree peonies. Genetic Resources and
Crop Evolution 53: 11–22.
Zhou Z-Q (周志钦), Pan K-Y (潘开玉), Hong D-Y (洪德元).
2003. Phylogenetic analyses of Paeonia section Moutan
(tree peonies, Paeoniaceae) based on morphological data.
Acta Phytotaxonomica Sinica ( 植物分类学报 ) 41:
436–446.
Zou Y-P (邹喻苹), Cai M-L (蔡美琳), Wang Z-P (王子平).
1999. Systemic studies on Paeonia sect. Moudan DC.
based on RAPD analysis. Acta Phytotaxonomica Sinica
(植物分类学报) 37: 220–227.

基于多基因序列和形态性状的牡丹组种间关系
1,2赵 宣 1,2,3周志钦* 2林启冰 2潘开玉 1李名扬
1(西南大学园林园艺学院 重庆 400716)
2(系统与进化植物学国家重点实验室, 中国科学院植物研究所 北京 100093)
3(中国农业科学院柑桔研究所 重庆 400712)

摘要 牡丹被认为是中国的国花, 具有很高的医学、观赏和经济价值。野生牡丹被认为是栽培牡丹的野生祖先, 因此弄清牡
丹组的种间亲缘关系具有重要的理论和实践意义。由于受到信息量的限制, 根据单基因数据或形态数据往往无法对牡丹组的
种间关系得到明确的结果。本研究用12份样品代表野生牡丹组(Paeonia, section Moutan DC., Paeoniaceae) 8个种, 利用包括核
基因(Adh1A、Adh2和GPAT)和叶绿体基因(trnS-trnG和rps16-trnQ)的DNA序列以及形态性状的多套数据来探讨野生牡丹的种
间关系。合并分析得到具高支持率的牡丹组物种间的系统发育关系。结果表明, 芍药属牡丹组8个野生种分为两个亚组, 即肉
质花盘亚组subsect. Delavayanae和革质花盘亚组subsect. Vaginatae。肉质花盘亚组包括滇牡丹P. delavayi和大花黄牡丹P.
ludlowii; 革质花盘亚组包括其余6个种。革质花盘亚组中, 四川牡丹P. decomposita ssp. decomposita和紫斑牡丹P. rockii ssp.
rockii关系密切; 卵叶牡丹P. qiui和矮牡丹P. jishanensis关系密切; 银屏牡丹P. suffruticosa ssp. yinpingmudan与凤丹P. ostii关系
密切, 并且后两个分支为姊妹群。
关键词 叶绿体基因; 形态性状; 核基因; 芍药属牡丹组; 系统发育