<<

Journal of Integrative 2013, 55 (11): 1069–1079

Research Article Phylogeny and Classification of sensu lato ()

Shuo Shi1,2, Jinlu Li1,3, Jiahui Sun1,2, Jing Yu1,2 and Shiliang Zhou1* 1State Key Laboratory of Systematic and Evolutionary Botany, Institute of Botany, the Chinese Academy of Sciences, Beijing 100093, 2University of Chinese Academy of Sciences, Beijing 100043, China 3College of Landscape Architecture, Northeast Forestry University, Harbin 150040, China Correspondence author Tel: þ86 10 6283 6503; Fax: þ86 10 6259 0843; E‐mail: [email protected] Available online on 14 August 2013 at www.jipb.net and www.wileyonlinelibrary.com/journal/jipb doi: 10.1111/jipb.12095

Abstract

The classification of the economically important Prunus L. sensu lato (s.l.) is controversial due to the high levels of convergent or the parallel evolution of morphological characters. In the present study, phylogenetic analyses of fifteen main segregates of Prunus s.l. represented by eighty‐four were conducted with maximum parsimony and Bayesian approaches using twelve chloroplast regions (atpB‐ rbcL, matK, ndhF, psbA‐trnH, rbcL, rpL16, rpoC1, rps16, trnS‐G, trnL, trnL‐F and ycf1) and three nuclear genes (ITS, s6pdh and SbeI) to explore their infrageneric relationships. The results of these analyses were used to develop a new, phylogeny‐based classification of Prunus s.l. Our phylogenetic reconstructions resolved three main clades of Prunus s.l. with strong supports. We adopted a broad‐sensed genus, Prunus, and recognised three subgenera corresponding to the three main clades: subgenus Padus, subgenus Cerasus and subgenus Prunus. Seven sections of subgenus Prunus were recognised. The dwarf , which were previously assigned to subgenus Cerasus, were included in this subgenus Prunus. One new section name, Prunus L. subgenus Prunus section Persicae (T. T. Yü & L. T. Lu) S. L. Zhou and one new species name, Prunus tianshanica (Pojarkov) S. Shi, were proposed.

Keywords: ; phylogeny; Prunoideae; Prunus; .

Shi S, Li J, Sun J, Yu J, Zhou S (2013) Phylogeny and classification of Prunus sensu lato (Rosaceae). J. Integr. Plant Biol. 55(11), 1069–1079.

Introduction P. americana Marshall), (P. armeniaca L.), (Prunus communis (L.) Huds.) and cherries (e.g., P. avium (L.) Prunus L. sensu lato (s.l.) consists of more than 200 species that L., P. pseudocerasus Lindl., P. serrulata Lindl.). are mainly found in the temperate regions of the northern Prunus s.l. together with other two small genera, hemisphere and in subtropical and tropical regions Rchb. and Princepia Royle, were traditionally classified into the (Rehder 1940; Yü et al. 1986). In a narrow sense, Prunus s.l. Amygdaloideae (Prunoideae) (Robertson 1974; Goldblatt 1976; can be sorted into ten groups corresponding to ten genera, which Zhang 1992; Morgan et al. 1994; Ghora and Panigrahi 1995; are, alphabetically, Amygdalus L., Armeniaca Scop., Cerasus Takhtajan 1997; Zhou et al. 1999; Lee and Wen 2001). The Mill., Emplectocladus Torr., Laurocerasus Duhamel, Maddenia members of the Amygdaloideae are also known as drupaceous Hook. f. & Thomson, Padus Mill., Persica Mill., Prunus L. sensu because they all bear . However, recent stricto (s.s.) and Pygeum Gaertn. (Chin et al. 2010). Some molecular evidence suggested that the traditional Amygdaloi- members of Prunus s.l. are common or ornamental species deae should be divided into two non‐sister tribes under the newly of great economic importance, such as (P. persica (L.) circumscribed C. Agardh, in which Prunus s.l. Batsch), (e.g., P. domestica L., P. cerasifera Ehrh., belongs to the monotypic tribe Amygdaleae, while Oemleria,

© 2013 Institute of Botany, Chinese Academy of Sciences 1070 Journal of Integrative Plant Biology Vol. 55 No. 11 2013

Princepia and a non‐drupaceous‐genus, Lindl., comprise the tribe Osmaronieae Rydb (Morgan et al. 1994; Potter et al. 2002; Potter et al. 2007). The circumscription of Prunus s.l. has varied significantly Prunocerasus Armeniaca Microcerasus Emplectocladus Amygdalus Persicae among taxonomists over the past 300 years (Table 1). Although Prunus six genera, Prunus, Armeniaca, Persica, Cerasus, Amygdalus sect. sect. sect. sect. sect. sect. sect. and Laurocerasus, had been recognized in pre‐Linnaean times This study (Tournefort 1700), Linnaeus only accepted Amygdalus and Padus Prunus Prunus Prunus Cerasus Padus Prunus Prunus Prunus in his work (Linnaeus 1753). Prunus Padus However, he divided Prunus into four genera, Padus (replaced Subg. Subg. Subg. Subg. Subg. Subg. Subg. Subg. Subg. Subg. Tournefort’s Laurocerasus), Armeniaca, Cerasus and Prunus,in Subg. Subg. his later work (Linnaeus 1764). After Linnaeus’ revision, some scholars chose to use micro‐genera and divided Prunus s.l. into five to eight genera (Miller 1754; Hutchinson 1964; Yü et al. Padus – Prunus Pygeum Cerasus 1986). There were a number of differences between the genera Cerasus Maddenia Armeniaca Amygdalus Amygdalus they proposed. For example, Laurocerasus and Padus were not Laurocerasus Gen. Gen. Gen. Gen. Gen. Gen. Gen. Gen. distinguished initially (Tournefort 1700; Linnaeus 1753; Miller Gen. 1754) but were later treated as two different genera (Shishkin Gen. and Yuzeqchuk 1971; Yü et al. 1986), and these two genera were sometimes merged with the genus Cerasus (Jussieu 1789; de Candolle 1825). Amygdalus and Persica were either treated Padus Cerasus as separate genera (Miller 1754;deCandolle 1825; Shishkin and Cerasus Amygdalus Amygdalus Prunophora Prunophora Prunophora Yuzeqchuk 1971) or combined into one genus, Amygdalus Laurocerasus subg. subg. subg. (Jussieu 1789; Yü et al. 1986). Hutchinson (1964) included – subg. subg. Maddenia subg. subg. subg. Tournefort’s genera Armeniaca and Cerasus in Prunus. subg. Prunus Additionally, Microcerasus (Spach) Webb (Webb and Gen. Prunus Prunus Prunus Prunus Prunus Prunus Prunus

Berthelot 1836) and Emplectocladus Torr. (Torrey 1851) were Prunus Gen. Gen. proposed but rarely accepted. Several small genera were Gen. Gen. Gen. Gen. Gen. Gen. adopted in regional flora (Shishkin and Yuzeqchuk 1971; Yü Gen. et al. 1986; Yasuhiro 2001; Fedorov 2002; Lu et al. 2003). Some taxonomists preferred to use a broad sense of genus containing all of the species of Prunus s.l. For example, Gray

(1856) and Bentham and Hooker (1865) both proposed the Armeniaca Amygdalus Cerasoides Laurocerasus Laurocerasus inclusive genus Prunus and divided it into three and seven Amygdalopsis sect. sections, respectively. Focke (1894) and Koehne (1893) also sect. Pygeum sect. sect. sect. Maddenia cations sect. sect. sect. recognized the large genus Prunus but divided it into seven fi Gen.

’ Prunus subgenera (which did not correspond with Bentham & Hooker s Gen. Prunus Prunus Prunus Prunus Prunus Prunus seven sections). A single genus of Prunus s.l. was suggested in Prunus Gen. recent papers (McVaugh 1951; Kalkman 1965; Ghora and Gen. Gen. Gen. Gen. Gen. Gen. Panigrahi 1995; Chin et al. 2010). The distinction of Pygeum and Gen.

Maddenia from Prunus s.l. has also been challenged based on . in some classi morphology (Kalkman 1965) and molecular evidence (Chin et al. 2010; Shi et al. 2013). These treatments make the ––– – – – –– – Prunus Prunus Prunus Prunus Prunus

circumscription of Prunus s.l. even broader. Prunus s.l Amygdalus Amygdalus The diverse inconsistency in the concept of the genus

Prunus s.l. is largely due to a lack of knowledge regarding Linnaeus (1753) Bentham and Hooker (1865) Rehder (1940) Yü et al. (1986) character convergence (Bortiri et al. 2006), reproductive isolation and phylogenetic relationships. Therefore, none of the treatments were convincing prior to the availability of molecular data. The

fi laurels Gen. early classi cations were heavily based on the morphology of ‐ the fruits, and narrow‐sensed genera currently remain in use Cherries Gen. ApricotsBird cherries Gen. Gen. Emplectocladus Maddenia Peaches Gen. Dwarf cherries AlmondsAmerican plums Gen. Table 1. Subdivisions of Natural groups Plums Gen. (Shishkin and Yuzeqchuk 1971; Lu et al. 2003). Recent Pygeum Phylogeny and Classification of Prunus 1071 molecular work revealed considerable homoplasy of the fruit Results morphology and supported the broad‐sensed genus (Bortiri et al. 2001; Lee and Wen 2001; Potter et al. 2007). Advances in The variability of the examined DNA regions and their maximum biology of Prunus s.l. have also furthered our understanding of parsimony scores are shown in Table 2. The longest DNA relationships within the group as follows. (1) Eurasian and region was ndhF, and the shortest was psbA‐trnH. The trnL‐F American plums (Prunus s.s. and Prunocerasus) are separate and ycf1 were more variable (p > 0.010), and atpB‐rbcL and and monophyletic groups (Shaw and Small 2004; Reales rpoC1 are more conservative than other genes. et al. 2010). (2) Dwarf cherries are reproductively more Prior to the concatenation of all gene partitions into one for compatible and genetically closer to plums and peaches than phylogenetic analyses, pairwise partition homogeneity tests to true cherries (Kataoka et al. 1988; Shimada et al. 2001). And (PHTs) were conducted and the p‐values (Table S1) indicated no (3) Maddenia has not diverged significantly from bird cherries significant conflict (p > 0.05) between the gene partitions. The (Padus) and is likely a member of Prunus s.l.(Chin et al. 2010; Shi concatenated data matrix of the 84 taxa included 13,203 et al. 2013). However, the problems related to resolving the nucleotides. There were 959 parsimony‐informative characters. phylogeny of Prunus s.l. are far from being solved because data MP searches yielded 1,180 most‐parsimonious with a from limited groups are not sufficient for the whole genus. consistency index (CI) of 0.804, a retention index (RI) of 0.846 The main goals of this study are to resolve the genetic and a tree length of 3,517. Consensus trees from the MP and BI relationships among the natural groups of species within Prunus s. analyses showed identical topologies (Figure 1). l. and provide a formal taxonomic treatment of the subdivisions of Prunus s.l. was shown to be monophyletic with strong Prunus s.l. under a phylogenetic framework. We sampled most of bootstrap support (BS) and posterior probability (PP) (BS such groups, including Maddenia and Pygeum,toensurehigh ¼ 100%, PP ¼ 1, Figure 1B) and is treated to be a single genus representations of every group within Prunus s.l. We developed a (see below). There were three major, monophyletic clades single copy nuclear marker of high resolution, SbeI, and integrated resolved under Prunus s.l. with good support (BS 99%, it with other two nuclear regions and twelve chloroplast regions to PP ¼ 1), which are treated to be three subgenera. Clade I reconstruct the systematic relationships of all groups of Prunus s.l., (BS ¼ 99%, PP ¼ 1, subgenus Padus) consists of species and a refined classification of Prunus s.l. based on the belonging to groups Laurocerasus, Maddenia, Padus and phylogenetic relationships of its subdivisions is presented. Pygeum. Clade II (BS ¼ 100%, PP ¼ 1, subgenus Cerasus)

Table 2. Variability of twelve chloroplast regions and three nuclear genes in Prunus s.l. and their maximum parsimony tree scores NLaScSNH p k Vs Is L CI RI Chloroplast genes atpB‐rbcL 26 760 679 23 16 0.004 2.917 24 7 26 0.962 0.963 matK 114 817 644 90 52 0.0085 5.467 70 40 88 0.852 0.967 ndhF 39 1913 1800 92 32 0.0102 18.351 98 59 128 0.812 0.946 psbA‐trnH 66 370 167 17 17 0.0092 1.541 36 12 40 0.925 0.961 rbcL 111 1043 487 53 31 0.0096 4.689 113 46 149 0.826 0.944 rpL16 31 771 656 32 16 0.0073 4.763 47 19 56 0.946 0.955 rpoC1 13 487 487 6 5 0.004 1.359 6 2 6 1 1 rps16 27 946 663 20 10 0.0068 4.473 46 18 47 1 1 trnS‐G 82 1644 575 53 42 0.0105 6.025 162 94 212 0.821 0.955 trnL 101 514 344 31 31 0.0093 3.189 42 24 50 0.92 0.984 trnL‐F 113 524 296 67 41 0.023 6.779 115 53 141 0.915 0.984 ycf1 30 888 853 72 21 0.0174 14.878 73 53 86 0.93 0.971 Nuclear genes ITS 91 671 294 76 45 0.012 3.646 256 147 618 0.547 0.81 s6pdh 12 876 791 25 7 0.0094 7.47 34 15 40 0.9 0.889 SbeI 29 728 718 45 18 0.01 7.458 48 26 58 0.845 0.898 N, number of species; La, aligned length; Sc, sites considered; S, number of polymorphic sites, excluding sites with missing data; NH, number of haplotypes; p, nucleotide diversity per site after exclusion of sites with missing data; k, average number of nucleotide differences; Vs, variable site; Is, parsimony‐information site; L, the tree length; CI, consistency index; and RI, retention index. 1072 Journal of Integrative Plant Biology Vol. 55 No. 11 2013

Figure 1. Phylogenetic relationships of Prunus s.l. based on a concatenated dataset of chloroplast atpB‐rbcL, matK, ndhF, psbA‐trnH, rbcL, rpL16, rpoC1, rps16, trnS‐G, trnL, trnL‐F and ycf1, and nuclear ITS, s6pdh and SbeI.

(A) Bayesian phylogram with branch length. (B) 50% majority‐rule Bayesian consensus tree with bootstrap supports. Values beside the branches are the bootstrap percentages for the branches in the maximum parsimony trees and Bayesian posterior probabilities. “‐” indicates the branch collapse in the parsimony tree. Phylogeny and Classification of Prunus 1073

includes the true cherry species and two species that were the genetic divergence among subordinate classification groups. previously treated as bird cherries. Clade III (BS ¼ 100%, Separation of Clade I from Clade II þ Clade III is obvious and PP ¼ 1, subgenus Prunus) comprises almonds, apricots, dwarf necessary because they have diverged significantly according cherries, peaches, and plums. The clade has diverged into to Figure 1. Clade I consists of tetraploid groups, such as seven monophyletic subclades which correspond to seven Laurocerasus, Maddenia, Padus and Pygeum which are of sections: section Amygdalus, section Armeniaca, section paleopolypoid origins, while Clade II and Clade III consists of Emplectocladus, section Microcerasus, section Persicae, sec- diploid and neopolyploid species. The basic inflorescence tion Prunocerasus, and section Prunus. Section Emplectocladus of species in Clade I is long , but the flowers of the has diverged the earliest. Section Amygdalus and section species in Clades II & III are solitary, clustered or in corymbs. Persicae are closer each other than to any other sections. Distinction of Clade II and Clade III is also considered appropriate when comparing the branch lengths of the two clades with those within each clade (Figure 1A). Clade III consists of almonds, Discussion apricots, dwarf cherries, peaches and plums which have flowers solitary or clustered, while Clade II consists of cherries which Circumscription of the genus Prunus flowers are in corymbs. The three clades correspond to three subgenera, i.e. subgenus Padus, subgenus Cerasus,and ‐ ‐ Whether a broad sensed Prunus genus or several narrow sensed subgenus Prunus. Such a subdivision of Prunus s.l. is completely genera should be used is a controversial topic and still under different from any of the existing classifications (Table 1). dispute. In recent decades, more scholars have preferred to use Somewhat similar treatments can be found in classifications of Prunus in the broad sense (Rehder 1940; Potter et al. 2007; Chin Koehne (1893) or Focke (1894) who united Laurocerasus and ‐ et al. 2010). We choose to use a broad sensed Prunus genus Padus into subgenus Padus, grouped true cherries into subgenus instead of a few genera according to Figure 1 for the following Cerasus, and Focke classified dwarf cherries into subgenus considerations. Firstly, Prunus s.l.ismonophyletic(Lee and Microcerasus. A significant difference of our proposal from the Wen 2001; Bortiri et al. 2006; Potter et al. 2007), which, similar to early classifications is the grouping of plums, peaches, almonds, fi some of the other reasons provided below, is not a suf cient apricots and dwarf cherries of singular flowers into a one condition for this treatment, but a necessary one. The anatomical subgenus, the true cherries of corymbs into the other subgenus, characters of suggested the monophyly of Prunus s.l. and the racemose groups into the third subgenus. (Zhang 1992) as well. Secondly, the species of Prunus s.l.show We might be questioned why not to give all the monophyletic high consistency in their morphology. All species in Prunus s.l. groups the generic ranks. The branching patterns and the branch bear and nearly every species has glands at the base of a lengths do tell us varying levels of genetic divergences which fl blade or at the top of a (Rehder 1940). The in orescences deserve corresponding taxonomic ranks. Judging from the of Prunus s.l. show continuous variation among different groups, branch lengths, the genetic divergences of the natural groups fl ranging from racemes (Subgenus Padus) to a solitary ower are not so significant when comparing to those of the three (Subgenus Prunus). As noted by Kalkman (1965),sharp clades. Therefore, the natural groups of species are better sorted delimitation among groups of Prunus s.l. is lacking, and it is into sections rather than genera. therefore preferable to adopt a broad sense of genus Prunus,with some subdivisions. Thirdly, the basic chromosome number of Prunus s.l.isx¼ 8, the same number as the well‐diverged Relationships within subgenus Padus – Exchorda, Oemleria and (Goldblatt and Johnson 1979 Subgenus Padus is a combination of four previously recognised 2012) and natural hybridisation occurs between apricots, plums genera: Laurocerasus, Maddenia, Padus and Pygeum. Pygeum and dwarf cherries (Rehder 1940; Kataoka et al. 1988; and Laurocerasus form two strongly supported subclades under Watkins 1995), indicating low genetic divergence among them. subgenus Padus. The monophyly of Maddenia is well supported ‐ Fourthly and the most importantly, if narrow sensed genera of but is nested within Padus, which is a polyphyletic group based on Prunus s.l. were to be adopted, many changes of names would be our results. Although these two taxa are readily distinguishable inevitable, which would bring troubles to users. On the contrary, morphologically, they have not diverged into monophyletic ‐ when a broad sensed genus Prunus is used, there will be very few lineages. The extensive morphological character parallelisms of nomenclatural changes. Maddenia and Padus within the subgenus are most likely due to the multiple origins of these species. All of the species of this subgenus Three subgenera of Prunus whose chromosome numbers have been counted to date are tetraploids (Goldblatt and Johnson 1979–2012), e.g., P. serotina The subdivisions of the genus Prunus were based on the following Ehrh., P. padus L., P. grayana Maxim. and P. buergeriana Miq. The two considerations: (1) the monophyly of the subdivisions, and (2) ancestors of the subgenus have not been documented. Collection 1074 Journal of Integrative Plant Biology Vol. 55 No. 11 2013

of different species and further phylogenetic analysis will be (Gray 1874). The representative species P. fasciculata (Torr.) A. necessary to explain the relationships of subgenus Padus. Gray displays unisexual flowers due to abortion of or Therefore, subdividing this subgenus is premature. pistils (Mason 1913) and bears pubescent fruits. The section shows affinities to the sections Amygdalus and Prunus and we Relationships within subgenus Cerasus treat it as a section of subgenus Prunus. Section Amygdalus (almonds) and section Persica (peaches) The subgenus Cerasus consists only of true cherries. The dwarf are very similar in morphology but monophyletic groups cherries, Microcerasus, do not belong to this subgenus. True (Figure 1; Bortiri et al. 2006). They were initially treated as cherries bear only one winter bud at each axils, while dwarf two genera by Tournefort (1700) but were combined into one cherries bear three winter buds at each leaf axils (a character genus or subgenus by some later researchers (Linnaeus 1753; state observed on peaches and almonds). It had been realised Rehder 1940; Yü et al. 1986; Yasuhiro 2001). The differences that the dwarf cherries of are more similar to plums between the two sections are mainly found in their fruit: the than to true cherries (Catling et al. 1999). It is easier to hybridise mesocarp of almonds is dry and splits when ripe, while the the dwarf cherries with plums, apricots or even peaches than with mesocarp of peaches is flesh and does not split. Section true cherries (Garley 1980; Kataoka et al. 1988). The dwarf Persicae is endemic to China and section Amygdalus has the cherries are more similar to plums in their fruit . No solid richest species diversity in central . evidence has been found so far to support the placement of dwarf Section Prunocerasus, the American plums, is distributed cherries together with true cherries. throughout North America and is very similar to section Prunus, Our molecular data support the membership of P. mahaleb L. the Eurasian plums. The two groups of plums were both found to and P. maackii Rupr. in subgenus Cerasus. These two species form monophyletic groups in this study. The clade including were sometimes treated as members of subgenus Padus these two groups of plums was not strongly supported because of elongated corymbs. has been (BS < 50%, PP ¼ 0.53), indicating that they should be treated reported to be similar to true cherries (Krüssmann 1986) and is a as different sections. Prunus texana D. Dietr., a species was diploid (González Zapatero et al. 1988), as the most species of once treated as a member of Amygdalus because of its subgenus Cerasus, while the examined species of subgenus pubescent fruits (Wight 1913), was again confirmed to be in Padus are all tetraploids (Goldblatt and Johnson 1979–2012). section Prunocerasus besides the earlier studies by Shaw and The similarity of P. maackii to cherries is supported by Small (2004, 2005). morphology (Li and Jiang 1998), and this species has been Apricots and dwarf cherries are grouped into section reported to be able form natural hybrids with Prunus maximo- Armeniaca and section Microcerasus, respectively. Apricots wiczii Rupr., a species of subgenus Cerasus. are endemic to China. The French , Vill. The divergence of species in subgenus Cerasus appears to (¼P. brigantiaca Vill.) is actually a . The dwarf cherries be insignificant, which may be due to quantum speciation or low are no doubt members of subgenus Prunus instead of subgenus resolutions of the markers. Interspecific hybridisations (Ohta Cerasus (Mowrey and Werner 1990; Bortiri et al. 2006). et al. 2007) have complicated the taxonomy of this subgenus. Therefore, we treated the dwarf cherries a separate section here. Since the subgenus was not well represented in this study, more Although Flora of China was aimed to be a revised version of extensive species sampling and more variable genes are Flora Reipublicae Popularis Sinicae (Li 2008), many Chinese expected to be used to resolve phylogenetic relationships within authors preferred to leave such tasks to next generations as the subgenus in the near future. articulated by L. T. Lu (an expert of Rosaceae taxonomy, personal communication). This work is a commitment to our fi Seven sections of subgenus Prunus predecessors and also outlines the classi cation of Prunus in the forthcoming Flora of Pan‐Himalayas. The subgenus Prunus is economically important and the phylogenetic relationships and taxonomic treatments of the Taxonomic treatments of the subdivisions of Prunus L. subgenus are the major concern of this study. The phylogeny of this subgenus indicates that a radiation event occurred that gave Prunus L., Sp. Pl. 1: 473. 1753. (Lectotype: Prunus birth to seven natural groups of species, which can be treated as domestica L.) different sections: section Amygdalus, section Armeniaca, Synonyms. Amygdalus L., Sp. Pl. 1: 472. 1753 section Emplectocladus, section Microcerasus, section Persi- Armeniaca Scop., Meth. Pl. 15. 1754 cae, section Prunocerasus, and section Prunus. Cerasus Mill., Gard. Dict. Abr. (ed. 4) vol. 1. 1754 The American section Emplectocladus diverged the earliest Padus Mill., Gard. Dict. Abr. (ed. 4) vol. 3. 1754 within the subgenus, It was described as a distinct genus Persica Mill., Gard. Dict. Abr. (ed. 4) vol. 3. 1754 (Torrey 1851) and then treated as a section within Duhamel, Traité Arbr. Arbust. 1: 345. 1755 Phylogeny and Classification of Prunus 1075

Pygeum Gaertn., Fruct. Sem. Pl. 1: 218. 1788 Section 2. Prunus L. subgenus Prunus section Amygdalus Prunophora Neck., Elem. Bot. 2:718. 1790 (L.) Benth. & Hook. f., Gen. Pl. 1: 610. 1865 Microcerasus Webb & Berthel., Hist. Nat. Îles Canaries 3(2. Basionym. Amygdalus L., Sp. Pl. 1: 472. 1753 2): 19. 1842 Synonym. Amygdalus L. subgenus Amygdalus section Amyg- Emplectocladus Torr., Proc. Amer. Assoc. Advancem. Sci. dalus T. T. Yü & L. T. Lu, Fl. Reip. Pop. Sin. 38:11, 1986 4: 192. 1851 Maddenia Hook. f. & Thomson, Hooker’s J. Bot. Kew Gard. Section Amygdalus includes the almonds, but Prunus Misc. 6: 381. 1854 texana, which was previously treated as an , does not belong to this section. Subgenus 1. Prunus L. subgenus Padus (Mill.) Peterm., Deutschl. Fl. 159. 1846 Section 3. Prunus L. subgenus Prunus section Armeniaca Basionym. Padus Mill., Gard. Dict. Abr. (ed. 4) vol. 3. (Scop.) Koch, Syn. Fl. Germ. Helv. 1: 205. 1837 1754 Basionym. Armeniaca Scop., Meth. Pl. 15. (Mar.) 1754 Synonym. Prunus L. subgenus Laurocerasus (Duhamel) Synonym. Prunus L. section Armeniaca (Mill.) Benth. et Rehder, Man. Cult. Trees & 478. 1927 Hook. f., Gen. Pl. 1:610, 1865

Most of the species previously classified under Laurocerasus Section 4. Prunus L. subgenus Prunus section Emplecto- Duhamel, Maddenia Hook. f. & Thomson, Padus Mill. and cladus (Torr.) A. Gray, Proc. Amer. Acad. Arts 10: 70. 1874 Pygeum Gaertn. belong to this subgenus. Basionym. Emplectocladus Torr., Proc. Amer. Assoc. Advancem. Sci. 4: 192. 1851 Subgenus 2. Prunus L. subgenus Cerasus (Mill.) A. Gray, Manual (Gray), ed. 2. 112. 1856. p. p. Section 5. Prunus L. subgenus Prunus section Micro- Basionym. Cerasus Mill., Gard. Dict. Abr., ed. 4. 1754. p. p. cerasus (Spach) C. K. Schneid., Ill. Handb. Laubholzk. Synonyms. Prunus L. subgenus Cerasus (Mill.) Focke in 1. 601. 1906 Engler & Prantl, Nat. Pflanzenfam. 3: 54. 1888. p. p. Basionym. Cerasus section Microcerasus Spach, Hist. Vég. Prunus L. subgenus Cerasus (Mill.) Koehne, Sarg. Pl. Wils. I. 423. 1834 1:226. 1912. p. p. Synonym. Cerasus Mill. section Spiraeopsis Koehne, Deutsche Dendr. 306, 1893 Subgenus Cerasus includes the true cherries, P. maackii Rupr. and P. mahaleb L. The dwarf cherries (Microcerasus)do Section 6. Prunus L. subgenus Prunus section Persicae not belong to this subgenus. (T. T. Yü & L. T. Lu) S. L. Zhou, com. nov. Basionym. Amygdalus subgenus Persica section Persicae Subgenus 3. Prunus L. subgenus Prunus T. T. Yü & L. T. Lu in Acta Phytotax. Sin. 23(3):209. 1985. Synonyms. Prunus L. subgenus Prunophora (Neck.) Synonyms. Amygdalus subgenus Persica section Mirae T. Focker in Engler & Prantl, Nat. Pflanzenfam. 3(3): 52. T. Yü & L. T. Lu in Acta Phytotax. Sin. 23(3):209, 1985. 1888, p. p. Prunus L. subgenus Prunus section Persica (L.) S. L. Zhou & Prunus L. subgenus Amygdalus (L.) Focke in Engler & X. Quan in J. Syst. Evol. 49(2):138. 2011 Prantl, Nat. Pflanzenfam. 3(3): 53. 1888 Prunus L. subgenus Microcerasus (Spach) Focke in Engler Section Persicae includes peaches which are endemic to China. & Prantl, Nat. Pflanzenfam. 3(3): 54. 1888 Prunus L. subgenus Emplectocladus (Torr.) Sarg., Silva of Section 7. Prunus L. subgenus Prunus section Prunocer- North America. Boston, 4: 7. 1892 asus Koehne, Deutsche dendrologie 302. Stuttgart: Prunus L. subgenus Armeniaca (Scop.) Nakai, Fl. Sylv. Kor. Verlag von Ferdinand Enke. 1893 5: 38. 1915 Prunus L. subgenus Lithocerasus Ingram, Ornamental Circumscription of the genus Prunus Cherries. 1948 The name of the following taxon should be changed within Section 1. Prunus L. subgenus Prunus section Prunus Prunus according to the International Code of Nomenclature for Synonyms. Prunus L. section Prunus Benth. & Hook. f., Gen. , fungi, and plants. Pl., 1: 610. 1865 Prunus tianshanica (Pojarkov) S. Shi, comb. nov. Basionym. Prunus L. section Prunophora (Neck.) Fiori & Paoletti, Fl. Cerasus tianshanica Pojarkov, Bot. Zhurn. S.S.S.R. 24(3): 242. Italia 12: 557. 1897. p. p. 1939. 1076 Journal of Integrative Plant Biology Vol. 55 No. 11 2013

Materials and Methods regions, atpB‐rbcL, matK, ndhF, psbA‐trnH, rbcL, rpL16, rpoC1, rps16, trnS‐G, trnL, trnL‐F and ycf1, and three nuclear genes, Sampling ITS, s6pdh and SbeI, were to be sequenced in this study or downloaded from GenBank. A total of eighty‐four species were collected: seventy‐seven The sequences of species not available in this study were species representing all ten groups of Prunus s.l., six species of downloaded from GenBank (Table S2). The accession numbers other genera of Rosaceae and one species of Moraceae were of the sequences determined in this study were also listed in chosen as outgroups (Table 3). Sequences of twelve chloroplast Table S2. Voucher specimens of the species sequenced in this

Table 3. A list of samples and their vouchers of Prunus s.l. and outgroups sequenced in this study

Taxon Subgenus and section Source Voucher (PE)

Prunus armeniaca L. Subg. Prunus sect. Armeniaca Xinjiang: Urumchi Botanical Garden S. L. Zhou BOP016969 var. holosericea Batal. Subg. Prunus sect. Armeniaca Liaoning: Xiongyue X. Quan 67 Miq. Subg. Padus Yunnan: Weixi S. Shi et al. SZ3264 Ehrh. Subg. Prunus sect. Prunus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010039 Prunus communis (L.) Hus. Subg. Prunus sect. Amygdalus Xinjiang: Urumchi Botanical Garden S. L. Zhou BOP016972 Prunus conradinae Koehne Subg. Cerasus Guizhou: Guiyang S. Shi et al. Z3022 Prunus conradinae Koehne Subg. Cerasus Hubei: Wuhan Botanical Garden S. Shi et al. Z3087 (Carrière) Franch. Subg. Prunus sect. Persicae Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010343 Prunus davidiana(Carrière) Franch. Subg. Prunus sect. Persicae Beijing Botanical Garden, CAS S. L. Zhou et al.BOP010009 Prunus ferganensis (Kost. & Rjab.) Kovalev & Kost. Subg. Prunus sect. Persicae Henan: Zhengzhou (from Xinjiang) X. Quan 103 Thunb. Subg. Prunus sect. Microcerasus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010170 Maxim. Subg. Padus Hunan: Xining Z. C. Luo BOP002863 Bunge Subg. Prunus sect. Microcerasus Hebei: Weichang S. L. Zhou 090730‐3 Prunus hypoleuca (Koehne) J. Wen Subg. Padus Gansu: Wuning S. L. Zhou BOP002508 Prunus hypoxantha (Koehne) J.Wen Subg. Padus Sichuan: Dujingyan S. L. Zhou 053 Prunus incisoserrata (T.T.Yü & T.C.Ku) J. Wen Subg. Padus Shaanxi: Qinling S. L. Zhou 054 Thunb. Subg. Prunus sect. Microcerasus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010337 Rehd. Subg. Prunus sect. Persicae Gansu: Xihe S. L. Zhou 51–63 (Maxim.) Koehne Subg. Prunus sect. Armeniaca Heilongjiang: Harbin Botanical Garden S. L. Zhou 135 Koehne Subg. Prunus sect. Persicae Tibet: Lasha, Norbu Lingka S. L. Zhou 2001816 (Siebold) Siebold & Zucc. Subg. Prunus sect. Armeniaca Hubei: Wuhan Botanical Garden X. Quan 127 Prunus persica (L.) Batsch Subg. Prunus sect. Persicae Henan: Xinyang S. L. Zhou et al. BOP017388 Prunus phaeosticta (Hance) Maxim. Subg. Padus Hunan: Xinning S. L. Zhou BOP002538 Prunus phaeosticta (Hance) Maxim. Subg. Padus Xinjiang: Urumchi Botanical Garden S. L. Zhou BOP016970 Lindl. Subg. Prunus sect. Prunus Gansu: Yuzhong S. L. Zhou 090915E Rehd. Subg. Cerasus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010500 L. Subg. Prunus sect. Armeniaca Qinghai: Xining Botanical Garden H. L. Sun S0042 Prunus spinulosa Siebold & Zucc. Subg. Padus Zhejiang: Hangzhou Botanical Garden S. L. Zhou 050b var. pendula Tanaka Subg. Cerasus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010112 Prunus tianshanica (Pojarkov) S. Shi Subg. Prunus sect. Microcerasus Xinjiang: Urumchi Botanical Garden S. L. Zhou BOP016971 Thunb. Subg. Prunus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010008 Lindl. Subg. Prunus Gansu: Xihe S. L. Zhou 63–65 Prunus undulata (D. Don) Roem. Subg. Padus Hunan: Xinning Z. C. Luo 121 Prunus ussuriensis Kovalev & Kostina Subg. Prunus sect. Prunus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010329 Prunus yedoensis Matsum. Subg. Cerasus Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010038 Prunus zippeliana Miq. Subg. Padus Guangxi: Ziyuan Z. C. Luo 118 Pygeum topengii Merr. Subg. Padus Guangdong: Zhaoqing S. L. Zhou BOP002820 sylvester Kostel. ex Maxim. Heilongjiang: Tahe S. L. Zhou & J. L. Li BOP016706 (Lindl.) Rehd. Outgroup Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010047 micromalus Makino Outgroup Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010089 Prinsepia sinensis (Oliv.) Oliv. ex Bean Outgroup Beijing Botanical Garden, CAS S. L. Zhou et al. BOP002502 Prinsepia uniflora Batalin Outgroup Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010133 Lindl. Outgroup Beijing Botanical Garden, CAS S. L. Zhou et al. BOP010007 The subgenera and sections the species belong are based on the taxonomic treatment in this study. GenBank accession numbers of the genes are listed in Table S2. Phylogeny and Classification of Prunus 1077

study were housed in the herbarium (PE) of the Institute of (MP) analyses using PAUP 4.0b10 (Swofford 2003), were Botany, Chinese Academy of Sciences. carried out on single gene datasets and the concatenated dataset. In the MP analyses, all characters were unordered DNA extraction, amplification and sequencing and equally weighted, and gaps were treated as missing data. Heuristic searches using 1,000 random sequence‐addition DNA was extracted from fresh or dried using the mCTAB replicates were conducted with tree‐bisection‐reconnection (modified cetyltrimethyl ammonium bromide) approach de- (TBR) branch‐swapping. Bootstrap support (BS) percentages scribed by Li et al. (2013). Approximately 20 mg of dried plant were calculated using 1,000 replicates, with ten trees being tissue was used per extraction. The primer sequences employed saved per replicate. In the BI analyses, the GTR þ I þ G in this study can be found in previous reports: rbcL (Olmstead model was selected by ModelTest 3.7 (Posada and Crandall et al. 1992; Gastony and Rollo 1995), matK (Yu et al. 2011) and 1998). Bayesian inference was run for 15,000,000 generations. ndhF (Wen et al. 2008). Primers for the amplification of ycf1 are Four chains, each starting with a random tree, and trees were to be published. Primers for SbeI were SbeIF: 50‐GCT CCA CGA sampled every 500 generations. Posterior probabilities (PP) ATA TAT GAG GCA CAT G‐30, SbeIR: 50‐TTC CAT GAA ATT were calculated from the majority consensus of all of the TCC TTC ATT GAC CA‐30). Polymerase chain reaction sampled trees when the standard deviation of the split amplification was carried out in a 25 mL volume with the following frequencies (SDSF) permanently fell below 0.01, and the trees reagents: 20 ng DNA template, 200 mM dNTPs, 0.25 mM each sampled during the burn‐in phase were discarded. primer, 1 unit of Taq polymerase and 1 Taq buffer (Takara Biotechnology Co., Dalian, China). The thermal cycling con- ditions were 3 min at 94° C, followed 35 cycles of 30 s at 94° C, Acknowledgements 40 s at 52° C and 1.5 min at 72° C, with a final extension of 10 min at 72° C. The obtained PCR products were purified with This study was supported partly by National Natural Science PEG8000 and sequenced using ABI PRISM® BigDye™ Foundation of China (31270239) and the National High‐tech Terminator Cycle Sequencing Kits v. 3.1 on an ABI 3730xl R&D Program (863 Program, 2012AA021602). DNA Analyzer (Life Technologies, 5,791 Van Allen Way, Carlsbad, 92008), following the manufacturer’s Received 28 Mar. 2013 Accepted 22 Jul. 2013 instructions.

Phylogenetic analyses References

The sequences of all Prunus s.l. species that were available in Bentham G, Hooker JD (1865) Genera Plantarum, Vol. 1. Reeve & Co., GenBank (Table S2) were downloaded. The newly generated London, UK. sequences were edited and assembled using Sequencher v. 4.7 Bortiri E, Heuvel BV, Potter D (2006) Phylogenetic analysis of (Gene codes Corporation, Ann Arbor, Michigan, USA). The morphology in Prunus reveals extensive homoplasy. Plant Syst. resulting sequences were combined with those downloaded Evol. 259, 53–71. from GenBank, aligned using the Clustal 2.0 (Thompson Bortiri E, Oh SH, Jiang JG, Baggett S, Granger A, Weeks C, et al. 1997; Larkin et al. 2007) and manually adjusted using Buckingham M, Potter D, Parfitt DE (2001) Phylogeny and SE‐Al 2.0 (Rambaut 1996). Each dataset was then subjected to systematics of Prunus (Rosaceae) as determined by sequence several rounds of phylogenetic evaluations to select reliable analysis of ITS and the chloroplast trnL‐trnF spacer DNA. Syst. Bot. sequences. Sequences that were misidentified or exhibited large 26, 797–807. amounts of missing data were excluded from the datasets. To Candolle AP de (1825) Catalogus Plantarum, Horti Botanaici Mon- understand the phylogenetic performance of individual gene speliensis. Montpellier, Paris, France. partitions, the variability of the genes was parameterised using Catling PM, McKay‐Kuja SM, Mitrow G (1999) Rank and typification in DnaSP 5.0 (Librado and Rozas 2009) and PAUP 4.0b10 North American dwarf cherries, and a key to the taxa. Taxon 48, 483– (Swofford 2003) based on factors such as nucleotide diversity 488. and the number of polymorphic sites (Table 2). Chin SW, Wen J, Johnson G, Potter D (2010) Merging Maddenia with Prior to combining the dataset of each gene partition, the morphologically diverse Prunus (Rosaceae). Bot. J. Linn. Soc. pairwise partition homogeneity tests (PHTs) were performed 164, 236–245. on all fourteen datasets, and the datasets were finally Fedorov AA (2002) Flora of Russia: The European Part and Bordering concatenated using SquenceMatrix (Vaidya et al. 2011). Regions, Vol. 6. Balkema, Rotterdam. Bayesian inference (BI) using the MrBayes ver. 3.2.1 Focke WO (1894) Rosaceae. In: Engler A, Prantl K, eds. Die Natürlichen (Ronquist and Huelsenbeck 2003) and maximum parsimony Pflanzenfamilien, Vol. 3. Engelmann, Leipzig, Germany. pp. 1–61. 1078 Journal of Integrative Plant Biology Vol. 55 No. 11 2013

Garley B (1980) Ah‐oon‐ye‐ya‐pa, the sand cherry: Its origin, improve- Lu L, Gu C, Li C, Alexander C, Bartholomew B, Brach A, Boufford D, ment and nomenclature. Fruit Varieties J. 34, 13–17. Ikeda H, Ohba H, Robertson K, Spongberg S (2003) Rosaceae. In: Gastony GJ, Rollo DR (1995) Phylogeny and generic circumscriptions Wu Z, Raven P, eds. Flora of China. Science Press, Missouri of cheilanthoid (Pteridaceae: Cheilanthoideae) inferred from Botanical Garden Press, Beijing, St. Louis. pp. 46–434. rbcL nucleotide sequences. Am. J. 85, 341–360. Mason SC (1913) The pubescent‐fruited species of Prunus of the Ghora C, Panigrahi G (1995) The Family Rosaceae in , Vol. 2. Southwestern States. J. Agric. Res. 1, 147–177. Bishen Singh Mahendra Pal Singh, Dehra Dun, India. McVaugh R (1951) A revision of the North American black cherries Goldblatt P (1976) Cytotaxonomic studies in the Tribe Quillajeae ( Ehrh. and relatives). Brittonia 7, 279–315. (Rosaceae). Ann. Mo. Bot. Gard. 63, 200–206. Miller P (1754) The Gardener’s Dictionary. Printed for the author, Goldblatt P, Johnson DE (1979–2012) Index to Plant Chromosome London. Numbers. Missouri Botanical Garden, St. Louis. Morgan DR, Soltis DE, Robertson KR (1994) Systematic and González Zapatero MA, Elena‐Roselló JA, Andrés FN (1988) evolutionary implications of rbcL sequence variation in Rosaceae. Números cromosómicos para la flora Española. Lagascalia 15, Am. J. Bot. 81, 890–903. 112–119. Mowrey BD, Werner DJ (1990) Phylogenetic‐relationships among Gray A (1856) Manual of the Botany of the Northern . 2nd species of Prunus as inferred by isozyme markers. Theor. Appl. ed. George P. Putnam & Co., New York. Genet. 80, 129–133. Gray A (1874) Contributions to the Botany of North America. In: Am. Ohta S, Yamamoto T, Nishitani C, Katsuki T, Iketani H, Omura M Acad. Arts Sci, ed. Proceedings of the American Academy of Arts and (2007) Phylogenetic relationships among Japanese flowering Sciences. 10, Press of John Wilson and Son, Boston. pp. 39–78. cherries (Prunus subgenus Cerasus) based on nucleotide sequen- Hutchinson J (1964) The Genera of Flowering Plants. Clarendon Press, ces of chloroplast DNA. Plant Syst. Evol. 263, 209–225. Oxford, UK. Olmstead RG, Michaels HJ, Scott KM, Palmer JD (1992) Monophyly Jussieu AL de (1789) Genera Plantarum. V. Herissant, Paris, France. of the Asteridae and identification of their major lineages inferred Kalkman C (1965) The old world species of Prunus subgen. from DNA‐sequences of rbcL. Ann. Mo. Bot. Gard. 79, 249– Laurocerasus including those formerly referred to Pygeum. Blumea 265. 13, 1–115. Posada D, Crandall KA (1998) MODELTEST: Testing the model of DNA Kataoka I, Sugiura A, Tomana T (1988) Interspecific hybridization substitution. Bioinformatics 14, 817–818. between Microcerasus and other Prunus spp. J. Jpn. Soc. Hortic. Sci. Potter D, Eriksson T, Evans RC, Oh S, Smedmark JEE, Morgan DR, 56, 398–407. Kerr M, Robertson KR, Arsenault M, Dickinson TA, Campbell CS Koehne BAE (1893) Deutsche Dendrologie. Verlag von Ferdin and (2007) Phylogeny and classification of Rosaceae. Plant Syst. Evol. Enke, Stuttgart, Germany. pp. 302, 305, 310–313. 266, 5–43. Krüssmann G (1986) Manual of Cultivated Broad‐leaved Trees and Potter D, Gao F, Bortiri PE, Oh SH, Baggett S (2002) Phylogenetic Shrubs. Timber Press, Portland. (English translation). relationships in Rosaceae inferred from chloroplast matK and Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, trnL‐trnF nucleotide sequence data. Plant Syst. Evol. 231, 77– McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, Thompson 89. JD, Gibson TJ, Higgins DG (2007) Clustal W and Clustal X version Rambaut A (1996) Se‐Al: Sequence alignment editor. ver. 2.0. Available: 2.0. Bioinformatics 23, 2947–2948. http://tree.bio.ed.ac.uk/software/seal/. Lee S, Wen J (2001) A phylogenetic analysis of Prunus and the Reales A, Sargent DJ, Tobutt KR, Rivera D (2010) Phylogenetics of Amygdaloideae (Rosaceae) using ITS sequences of nuclear Eurasian plums, Prunus L. section Prunus (Rosaceae), according to ribosomal DNA. Am. J. Bot. 88, 150–160. coding and non‐coding chloroplast DNA sequences. Tree Genet. Li CL, Jiang SY (1998) New combinations in Ameniaca Mill. and Genomes 6, 37–45. Cerasus Juss. (Rosaceae). J. Syst. Evol. 36, 367–372. Rehder A (1940) Manual of Cultivated Trees and Shrubs Hardy in North Li DZ (2008) Floristics and plant biogeography in China. J. Integr. Plant America Exclusive of the Subtropical and Warmer temperate Biol. 50, 771–777. Regions. MacMillan, New York. Li J, Wang S, Yu J, Wang L, Zhou S (2013) A modified CTAB protocol Robertson KR (1974) The genera of Rosaceae in the southeastern for plant DNA extraction. Chin. Bull. Bot. 48, 72–78. United States. J. Arnold Arbor. 55, 303–332. Librado P, Rozas J (2009) DnaSP v5: A software for comprehensive Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian analysis of DNA polymorphism data. Bioinformatics 25, 1451– phylogenetic inference under mixed models. Bioinformatics 19, 1452. 1572–1574. Linnaeus C (1753) Species Plantarum. Stockholm, Sweden. Shaw J, Small RL (2004) Addressing the “hardest puzzle in American Linnaeus C (1764) Genera Plantarum: Eorumque characteres naturales pomology”: Phylogeny of Prunus sect. Prunocerasus (Rosaceae) secundum numerum, figuram, situm, et proportionem omnium based on seven noncoding chloroplast DNA regions. Am. J. Bot. 91, fructificationis partium. Impensis direct. Laurentii Salvii, Holmiae. 985–996. Phylogeny and Classification of Prunus 1079

Shaw J, Small RL (2005) Chloroplast DNA phylogeny and phylogeog- Wen J, Berggren ST, Lee CH, Ickert‐Bond S, Yi TS, Yoo KO, Xie L, raphy of the North American plums (Prunus subgenus Prunus section Shaw J, Potter D (2008) Phylogenetic inferences in Prunus Prunocerasus, Rosaceae). Am. J. Bot. 92, 2011–2030. (Rosaceae) using chloroplast ndhF and nuclear ribosomal ITS Shi W, Wen J, Lutz S (2013) Pollen morphology of the Maddenia clade of sequences. J. Syst. Evol. 46, 322–332. Prunus and its taxonomic and phylogenetic implications. J. Syst. Wight WF (1913) North American Species of the Genus Amygdalus, Evol. 51, 164–183. Dudley Memorial Volume. Leland Stanford Junior University Shimada T, Hayama H, Nishimura K, Yamaguchi M, Yoshida M (2001) Publications, University Series, California, USA. The genetic diversities of 4 species of subgenus Lithocerasus (Prunus, Yü TT, Lu LT, Ku TC, Li CL, Chen SX (1986) Rosaceae (3) Prunoideae. Rosaceae) revealed by RAPD analysis. Euphytica 117, 85–90. In: Yü TT, ed. Flora Reipublicae Popularis Sinicae, Tomus 38. Shishkin BK, Yuzeqchuk SV (1971) Rosaceae: , Amygda- Science Press, Beijing. pp. 1–133. loideae. In: Komarov VL, ed. Flora of the U.S.S.R,Vol.10, Smithsonian Yasuhiro E (2001) Rosaceae. In: Iwatsuki K, ed. Flora of Japan. Institution Washington, D.C., USA. pp. 1–506 (English translation). Kodansha Ltd, Tokyo. pp. 127. Swofford DL (2003) PAUP: Phylogenetic Analysis Using Parsimony Yu J, Xue J, Zhou S (2011) New universal matK primers for DNA (and Other Methods). Version 4. Sinauer Associates, Sunderland, barcoding angiosperms. J. Syst. Evol. 49, 176–181. Massachusetts. Zhang SY (1992) Systematic anatomy of the Rosaceae. Blumea Takhtajan AL (1997) Diversity and Classification of Flowering Plants. 37, 81–158. Columbia University Press, New York. Zhou LH, Wei ZX, Wu ZY (1999) Pollen morphology of Prunoideae of Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG China (Rosaceae). Acta Bot. Yunnan. 21, 207–211. (1997) The CLUSTAL_X windows interface: Flexible strategies for (Co‐Editor: Hongya Gu) multiple sequence alignment aided by quality analysis . Nucleic Acids Res. 25, 4876–4882. Torrey J (1851) On some new plants discovered by Col. Fremont, in California. Proc. Am. Assoc. Advance Sci. 4, 190–193. Supporting Information Tournefort JP de. (1700) Institutiones Rei Herbariae. 1st ed. Typograhia Regia, Paris, France. Additional supporting information may be found in the online Vaidya G, Lohman DJ, Meier R (2011) SequenceMatrix: Concatenation version of this article at the publisher’s website: software for the fast assembly of multi‐gene datasets with character set and codon information. Cladistics 27, 171–180. Table S1. P values from pair‐wise tests for partition Watkins R (1995) Cherry, plum, , apricot and almond. In: Smartt J, homogeneity of fourteen data sets. P values smaller than Simmonds N, eds. Evolution of Crop Plants. Longman Scientific and 0.05 are painted yellow. Technical, Burnt Mill. pp. 423–428. Table S2. The sequences of all species belonging to Webb PB, Berthelot S (1836) Histoire Naturelle Des Iles Canaries. 2, 19. drupaceous Rosaceae that were used in this study. Béthune éditeur, Paris. Accessions yellow are new for this study.