reconsideration of the prunus serrulata complex (rosaceae) and...

Botanical Journal of the Linnean Society

, 2007,

154

, 35–54. With 11 figures

© 2007 The Linnean Society of London,

Botanical Journal of the Linnean Society,

2007,

154

, 35–54

35

Blackwell Publishing LtdOxford, UKBOJBotanical Journal of the Linnean Society0024-4074© 2007 The Linnean Society of London? 200715413554Original Article

RECONSIDERATION OF THE P. SERRULATA COMPLEXK.-S. CHANG ET AL.

*Corresponding author. E-mail: [email protected]

Reconsideration of the

Prunus serrulata

complex (Rosaceae) and related taxa in eastern Asia

KAE-SUN CHANG

1

, CHIN-SUNG CHANG

1

*, TAE YOON PARK

2

and MARK S. ROH

3

1

Department of Forest Sciences and The Arboretum, Seoul National University, Seoul 151–921, Korea

2

Graduate School of Education, Yonsei University, Seoul 120–749, Korea

3

Floral and Nursery Plants Research Unit, National Arboretum, Agricultural Research Service, United States Department of Agriculture, Beltsville, MD 20705, USA

Received June 2005; accepted for publication November 2006

The

Prunus serrulata

complex with several infraspecific taxa, including

P. jamasakura

and

P. leveilleana

, is indig-enous to eastern Asia. Although these taxa are very common, the taxonomic confusion surrounding this complex isreflected in the ambiguity shown by the various taxonomic treatments currently used. Patterns of intraspecific vari-ation and recognition of taxa within the

P. serrulata

complex and its related species were investigated in 468 indi-viduals using principal components analysis. Particular emphasis was placed on

P. serrulata

,

P. jamasakura

, and

P. leveilleana

because of their high degree of morphological intergradation and inconsistent circumscription in thevarious systematic treatments. Although morphological variation among different taxa of the

P. serrulata

complexwas continuous for most reproductive and leaf characters, the related taxa (

P. pseudocerasus

,

P. lannesiana

f.

albida

,and

P. sargentii

) were found to be distinct based on the number of flowers per inflorescence and peduncle length

.

Ingeneral, the number of flowers per inflorescence, the flower size, peduncle length, and hair density on the pedicelwere found to be characters of great taxonomic value for most of the related taxa. However, there is not sufficientmorphological evidence for the taxonomic splitting of

P. serrulata

to warrant the designation of varieties, with theexception of the pubescent taxon,

P. serrulata

var.

pubesecens

. © 2007 The Linnean Society of London,

BotanicalJournal of the Linnean Society

, 2007,

154

, 35–54.

ADDITIONAL KEYWORDS:

morphology – principal components analysis – taxonomy.

INTRODUCTION

Prunus

L. is a widespread genus of woody plants of upto 430 species (Huxley, Griffiths & Levy, 1999; Hillier& Coobes, 2002) throughout the northern hemisphere,especially the temperate zone. Bentham & Hooker(1865) proposed seven sections in the genus

Prunus

,but Koehne (1913) recognized four subgenera, mainlybased on leaf, inflorescence, presence/absence of bract,and flower morphology. Rehder (1940) renamed

Amugdali

as

Amygdalus

and added one more subge-nus,

Laurocerasus

, in addition to four subgenera(

Prunophora

,

Amygdalus

,

Cerasus

, and

Padus

)

sensu

Koehne. Recent ITS phylogeny research (Bortiri

et al.

,2001, 2002) generally supports Rehder’s subgenericclassification (Rehder, 1940).

The subgenus

Cerasus

(Adans.) Focke, which ismainly distributed throughout eastern Asia, com-prises

c

. 40 species (Koehne, 1913; Rehder, 1940;Krüssmann, 1986). Although several divisions (as sec-tions) within this subgenus or independent generictreatment (

s.s.

) were proposed by many authors (Koe-hne, 1913; Rehder, 1940; Yü & Li, 1986; Charkevicz,1996; Ohba, 2001; Li & Bartholomew, 2003), thesetreatments were not supported by the recent ITS phy-logeny (Bortiri

et al

., 2001, 2002; Lee & Wen, 2001).Therefore, any subdivisions of the subgenus

Cerasus

or generic treatment are no longer accepted.

Prunus serrulata

Lindl.,

P. jamasakura

(Siebold)Koidz., and

P. leveilleana

Koehne with severalinfraspecific taxa, and which are widely cultivated asornamental trees, are here collectively referred to asthe

P. serrulata

complex.The

P. serrulata

complex is characterized bycorymbs with one to five flowers, double serration on

36

K.-S. CHANG

ET AL

.

© 2007 The Linnean Society of London,

Botanical Journal of the Linnean Society,

2007,

154

, 35–54

the leaf, and glabrous styles (Koehne, 1913; Nakai,1915; Wilson, 1916; Rehder, 1940; Kitamura &Murata, 1979; Lee, 1980; Kuitert & Peterse, 1999;Ohba, 2001). The taxa of the

P. serrulata

complex havereceived varied treatments in the floras of easternAsia (Koehne, 1913; Nakai, 1915; Wilson, 1916; Kita-mura & Murata, 1979; Lee, 1980; Lee, 1996; Kuitert &Peterse, 1999; Ohba, 2001), but the treatments of taxaare still debatable.

Prunus serrulata

was first described by Lindley(1830) with double white flowers as a cultivated form(Wilson, 1916). However, Wilson (1916) describedplants with single flowers as a wild form and called it

P. serrulata

var.

spontanea

(Maxim.) E. H. Wilson.Therefore,

P. serrulata

comprises two varieties: var.

serrulata

, the type variety and var.

spontanea

. Someauthors (Koehne, 1913; Wilson, 1916) considered the

P. serrulata

complex as one polymorphic species withseveral varieties, but many taxonomists (Kitamura &Murata, 1979; Ohwi, 1984) treated these varieties atthe rank of species in China, Korea, and Japan. Forexample, Koehne (1913), Wilson (1916), Lee (1980),and Lee (1996) recognized

P. serrulata

var.

spontanea

,Kitamura & Murata (1979), Ohwi (1984), and Ohba(2001) placed this taxon at the rank of species.

P. jamasakura

and

P. leveilleana

were delimited onlyby blooming 2 weeks later at the same site, comparedwith

P. serrulata

var.

pubescens

(Makino) Nakai (Kita-mura & Murata, 1979; Ishihidemi

et al

., 2000). How-ever, Kuitert & Peterse (1999) stated that most taxa of

Prunus

kept their flowers longer than was generallybelieved. In particular, it may be 2 or 3 weeks betweenthe first flowers and the first shedding of the petals.Kitamura & Murata (1979), Lee (1980), Ohwi (1984),and Ohba (2001) recognized

P. leveilleana

at the spe-cies level, but Wilson (1916) and Kuitert & Peterse(1999) treated it as var.

pubescens

.Apart from these two taxa, numerous infraspecific

treatments of the

P. serrulata

complex have been gen-erated since Lindley. Several varieties of

P. serrulata

inKorea were proposed by Nakai (1915), such as var.

gla-bra

(Makino) Nakai, var.

pubescens

(Makino) Nakai,var.

tomentella

Nakai, var.

verecunda

(Koidz.) Nakai,var.

sontagiae

(Koehne) Nakai, var

. compta

Nakai, andvar.

intermedia

Nakai based just on the degree ofpubescence of the leaf and peduncle. Lee (1966, 1980)recognized six varieties [var.

pubescens

, var.

spontanea

,var.

sontagiae

, var.

verecunda

, var.

densifolia

(Koehne)Uyeki, and var.

quelpaertensis

(Nakai) Uyeki]. InJapan, two varieties and a form [

P. jamasakura

f.

pube-scens

(Makino) Ohwi,

P. jamasakura

var.

chikusiensis

(Koidz.) Ohwi, and

P. jamasakura

var.

superflua

(Koidz.) Ohwi] were recognized (Kitamura & Murata,1979; Ohwi, 1984), whereas one variety [

C. serrulata

var.

pubescens

(Makino) T. T. Yü et C. L. Li] and one cul-tivated Japanese taxon [

C. serrulata

var.

lannesiana

(Carrière) T. T. Yü et C. L. Li] were reported in China(Yü & Li, 1986). A major difficulty can be attributed tothe use of minor differences in leaf morphology and thedegree of pubescence for taxon definition, and conse-quently the differences used to distinguish taxa werefar from consistent. There is a clear need for a study ofthis complex that goes beyond the limited work thathas been done previously as a small part of local florastudies.

Here we present the results of a phenetic analysis ofmorphological characters of the

P. serrulata

complex.We were specifically interested in determiningwhether the variability within the

P. serrulata

com-plex and other related taxa are primarily attributableto morphological discontinuities among the geograph-ically isolated groups and whether a suite of mor-phological characters can consistently distinguishthe

P. serrulata

complex from other related taxa, suchas

P. lannesiana

f.

albida

E. H. Wilson,

P. sargentii

Rehder, and

P. pseudocerasus

Lindl.The primary objective of the first part of this

research was to redefine the phenetic relationshipsamong the different morphological entities of the

P. serrulata

complex as well as related taxa. Particularattention was focused on

P. serrulata

and

P. jamasakura

because of the high degree of morpho-logical intergradation and inconsistent circumscrip-tion across the various systematic treatments.Additionally,

P. leveilleana

was investigated to deter-mine if its morphological differentiation from theother taxa warranted species rank.

MATERIAL AND METHODS

Analysis of the morphological characters was per-formed on material borrowed from the following her-baria: University of Tokyo (TI), Tokyo MetropolitanUniversity (MAK), and the Chinese Academy ofSciences, Beijing (PE). Herbarium specimens wereselected to represent the entire geographical rangeand to encompass the morphological variabilitypresent within each taxon.

Mature leaves from

P. serrulata

var.

spontanea

,

P. serrulata

var.

pubescens, and P. sargentii were col-lected from 23 localities in South Korea from 2002 to2004. Herbarium specimens from the material col-lected are stored at the T. B. Lee Herbarium of theArboretum of Seoul National University (SNUA).Illustrations of the leaf and fruit morphology (Fig. 1)were adapted from Nakai (1916).

The characters selected for analyses included thosemost frequently used in keys and diagnoses. Charac-ters considered useful by previous authors (Koehne,1913; Wilson, 1916; Kitamura & Murata, 1979; Lee,1980; Ohwi, 1984; Yü & Li, 1986; Ohba, 2001) for reli-able identification were also selected (Tables 1 and 2).

RECONSIDERATION OF THE P. SERRULATA COMPLEX 37

© 2007 The Linnean Society of London, Botanical Journal of the Linnean Society, 2007, 154, 35–54

As both fruiting and flowering specimens were notavailable at the same time, flower and fruit with leafcharacters were chosen based on the observation ofseparate specimens. For leaf measurements, a typicalfully mature leaf was selected. The initial data matri-ces were constructed from 17 vegetative and floralcharacters, including three ratios (Table 1), and 18vegetative and fruit characters, including two ratios(Table 2). Measurements were made with a ruler anda binocular micrometer.

Two hundred and ten individuals for flowers withleaves from P. serrulata var. spontanea (78 individu-als), P. jamasakura (15 individuals), P. serrulata var.pubescens (107 individuals), P. leveilleana (ten indi-viduals), and 122 individual mature fruits with leavesfrom P. serrulata var. spontanea (28 individuals),P. jamasakura (13 individuals), P. serrulata var. pube-scens (73 individuals), and P. leveilleana (eight indi-viduals) were measured for the various characters (seeAppendix). Also, for comparisons, this study consid-ered a group of species [P. lannesiana f. albida (sevenfor flowers and two for fruit specimens), P. sargentii(82 for flowers and 39 for fruit specimens), andP. pseudocerassus (six for fruits without flowersspecimens)] that shared similar characters withthe P. serrulata complex and were once treated as

infraspecific taxa of P. serrulata (Chang, 2004).Because of the insufficient number of flowering speci-mens, flower characters of P. pseudocerassus could notbe measured for this multivariate analysis (Fig. 2).

Individuals of P. serrulata examined in Korea andChina showed both the absence and the presence ofhair on the leaf, petiole, and pedicel (plants of var.spontanea and var. pubescens), whereas individualsin Japan that we have identified based on thesame criteria had been described in the literatureas P. jamasakura and P. leveilleana, respectively(Wilson, 1916; Kitamura & Murata, 1979; Lee, 1980;Ohwi, 1984; Yü & Li, 1986; Ohba, 2001).

Also, the criteria used classically in floras to dis-tinguish between P. serrulata var. spontanea andP. sargentii concerned the umbel-like/corymb-likeinflorescence. Individuals in the Izu Island of Japanwith awn-teeth and corymbose inflorescences weretreated as P. lannesiana f. albida by us. P. lannesianaf. albida has simple flowers, whereas those ofP. lannesiana f. lannesiana are double.

Morphological variation within and among taxa wasassessed using univariate statistics (mean, maximum,minimum) and multivariate morphometric analysis(principal components analysis: PCA). The PCAs andunivariate statistics were produced with the SAS pro-

Figure 1. Diagram of Prunus serrulata var. spontanea with an indication of the measurement of the characters of flowerswith leaves (A) and fruits with leaves (B). Keys are given in Tables 1 and 2.

A B

38 K.-S. CHANG ET AL.


gram (SAS Institute Inc., 1988). A correlation matrixwas generated using selected significant characterstogether with univariate statistics and the analysis ofvariance (ANOVA). Also, bivariate scatter diagramswere performed and characters associated with indi-viduals in several operation taxonomic units (OTUs)were plotted.

Kuitert & Peterse (1999) reported that most cher-ries keep their flowers for more than 2 or 3 weeks inthe wild, but Kitamura & Murata (1979) recognizedthe difference between P. leveilleana and P. serrulatavar. pubescens in the flowering period. In fact, theflowering time of P. serrulata var. pubescens exhibitsthe highest degree of variation, even within the samesite. For this study, 41 individuals of P. serrulata var.pubescens in Mt. Gwanak-san of Seoul city from mid-April to early May 2004 were randomly selected toinvestigate the variation in flowering time. We dividedstages of flowering into four scales; the first floweropening, full bloom, the first shedding of petals, andthe last shedding of petals. We recorded the date offlowering for each tree and compared them with eachother.

The length of the peduncle and the pedicel and inflo-rescence types have been used as key characters inclassification by many researchers (Koehne, 1913;

Kitamura & Murata, 1979; Ohwi, 1984; Yü & Li, 1986;Ohba, 2001; Li & Bartholomew, 2003). However, Wil-son (1916) mentioned that the length of the peduncleand inflorescence type may vary with the number offlowers per corymb, the age and vitality of thetree, weather, and soil nutrition. Two individualsof P. serrulata var. spontanea in Suwon Arboretum ofSeoul National University were selected for this studyfrom mid-April to mid-July 2002 to measure thelength of the peduncle at two different stages: the firstflower opening and full bloom.

RESULTS

CHARACTER ANALYSIS OF FLOWERS WITH LEAVES

For flower with leaf characters, the first three PCAaxes accounted for only 55% of the total variance: prin-cipal component (PC) 1 had the highest loadings forthe width of the petal (C), flower diameter (D), thelength of the calyx tube (F), the width of the calyx tube(G), the length of the calyx lobe (H), and the width of

Table 1. Morphological characters for flowers with leavesof the Prunus serrulata complex used in the principalcomponents analysis. Units of measurement are given inparentheses

Code Character

A Number of flowersB Petal length (mm)C Petal width (mm)D Flower diameter (mm)E Length difference between stamen and carpel

(mm)F Length of calyx tube (mm)G Diameter of calyx tube top (mm)H Length of calyx lobe (mm)I Width of calyx lobe (mm)J Peduncle length (mm)K Distance from the end of the peduncle to the

second pedicel (mm)L Length of bract (mm)M Leaf length (mm)N Petiole length (mm)O Stipule length (mm)P Pedicel length (mm)Q Number of hairs on pedicel (number/1 mm)R Ratio of length/width of petalS Ratio of calyx tube length/calyx lobe lengthT Widest portion of the long axis of petal

Table 2. Morphological characters for fruits with leavesof the Prunus serrulata complex used in the principalcomponents analysis. Units of measurement are given inparentheses

Code Character

A Number of fruitsB Maximum leaf length (mm)C Maximum leaf width (mm)D Angle of leaf apex (°)E Angle of leaf base (°)F Length of leaf apex (mm)G Number of serrations at the widest part of leaf

(number/2 cm)H Length of serration at the widest part of leaf

(mm)I Width of serration base at the widest part of

leaf (mm)J Number of hairs on abaxial side of leaf

(number/25 mm2)K Number of hairs on adaxial side of leaf

(number/25 mm2)L Number of lateral veinsM Petiole length (mm)N Number of hairs on petiole (number/1 mm)O Peduncle length (mm)P Length from the end of the peduncle to the

second pedicel (mm)Q Ratio of length/width of leafR Widest portion of the long axis of leafS Pedicel length (mm)T Number of hairs on pedicel (number/1 mm)



the calyx lobe (I); PC 2 had the highest loadings for thelength of the peduncle (J), the length of the leaf (M),and the length of the petiole (N); and PC 3 had thehighest loadings for the number of flowers (A), the dis-tance from the end of the peduncle to the secondpedicel (K), and the number of hairs in the pedicel (Q).

PC 1 vs. PC 2 provided better separation ofP. lannesiana f. albida, whereas PC 1 vs. PC 3revealed clusters of P. sargentii that corresponded tothe previously recognized taxonomic species (Fig. 2). Itwas clear that there was minimal overlap in the clus-ters of individuals of P. lannesiana f. albida and

P. sargentii. The characters (Fig. 3) that contributedmost to the separation were the length of the peduncle(J), the number of flowers (A), the length of the pedicel(P), and flower size (characters B, D, F, G, H, and I).

Univariate statistics (Fig. 4), in addition to theminimum and maximum values for many characters,showed that values overlapped extensively forP. serrulata var. spontanea, P. serrulata var. pubes-cens, P. jamasakura, and P. leveilleana [e.g. the lengthof the peduncle (J), the distance from the end of thepeduncle to the second pedicel (K), and flower diame-ter (D)]. The number of hairs on the petiole (Fig. 4,

Figure 2. Scatter diagrams from the principal components analysis results of flowers for taxa of the Prunus serrulatacomplex and related species. P. serrulata var. spontanea (�); P. jamasakura (= Japanese P. serrulata var. spontanea) (�);P. serrulata var. pubescens (�); P. leveilleana (�); P. sargentii (�); P. lannesiana f. albida (�). A, PC 1 vs. PC 2. B, PC 1vs. PC 3.

-3

-2

-1

0

1

2

3

4

- 3 - 2 - 1 0

PC2

PC

1

1 2 3 4

A

B

-4

-3

-2

-1

0

1

2

3

4

- 3 - 2 - 1 0 1 2 3 4

PC3

PC

1



character Q) was only useful in separating pubescenttaxa (P. serrulata var. pubescens and P. leveilleana)from glabrous taxa (P. serrulata var. spontanea andP. jamasakura). However, much overlap occurred inthe central region of the scatter diagram (Fig. 2)between P. serrulata var. pubsecens in Korea andChina and P. leveilleana in Japan. Also, no strong dis-continuities existed among P. serrulata var. sponta-nea, P. serrulata var. pubescens, P. jamasakura, andP. leveilleana in the diagram (Fig. 2).

CHARACTER ANALYSIS OF FRUIT WITH LEAVES

For fruit with leaf characters, the first three PCA axesalso accounted for 55% of the total variance. The firstaxis was positive, except for the number of fruits (A),the number of serrations at the widest position of theleaf (G), the number of hairs on the petiole (N), thelength from the end of the peduncle to the secondpedicel (P), the number of hairs on the pedicel (T),with major contributions being made by leaf length(B), leaf width (C), the length of the leaf apex (F), thelength of the serration at the middle of the long axis(H), and the width of the serration base at the widestposition of the leaf (I). The second axis primarily

reflected the number of hairs on the petiole (N) andthe number of hairs on the pedicel (T). Only 9.8% ofthe total variance was accounted for by the third axis,with the number of fruits (A) and the length from theend of the peduncle to the second pedicel (P) makingthe greatest contributions.

PC 1 vs. PC 2 and PC 2 vs. PC 3 (Fig. 5) revealedclusters of OTUs that did not correspond to the priorrecognized taxa, except P. pseudocerasus. The OTUsfrom P. pseudocerasus occupied the end of the PC 1axis and the middle of the second principal componentaxis. In a plot of PC 1 vs. PC 2 (Fig. 5), OTUs fromP. jamasakura occupied the left area of the plot andoverlapped with a few OTUs from P. serrulata var.spontanea.

Fruit with leaf characters were determined by theunivariate analysis to be insignificant in separatingmost taxa, except P. pseudocerasus in China. Themean values and ranges are given in Figure 6. Thisspecies can be consistently distinguished from theremainder of the P. serrrulata complex, as well asP. sargentii, in characters of leaf width (C), the num-ber of leaf serrations at the widest position of the leaf(G), peduncle length (O), and the number of lateralveins (L; Fig. 7).

Figure 3. Univariate statistics with the minimum and maximum values for the discriminating characters for flowers ofthe Prunus serrulata complex (S) and P. lannesiana f. albida (L). The characters that contributed most to the separationbetween these two taxa are presented here. A, number of flowers. B, length of calyx lobe (mm). C, peduncle length (mm).D, pedicel length (mm).

A B

0

1

2

3

4

5

L S

Species

0

10

20

30

40

L S

Species

C D

0

2

4

6

8

10

L S

Species

0

10

20

30

40

50

L S

Species



In general, the number of flowers per inflorescence,flower size, peduncle length and hair density on thepedicel can be considered characteristics of taxonomicvalue for most related taxa, except for the P. serrulatacomplex.

DEGREE OF PUBESCENCE WITHIN A POPULATION AND AMONG POPULATIONS

Prunus leveilleana and P. serrulata var. pubescens,which were characterized by pubescence of the petiole,pedicel, and peduncle, remained as problematic taxa.Hence, the frequency distribution for pubescence ofthe petiole, pedicel, and peduncle of individuals withina population and among populations was examined inorder to clarify their relationships.

Korean populations were predominantly pubescent,although they contained both pubescent and glabrousspecimens. On the other hand, our herbariumobservations confirmed that glabrous individuals weremore commonly found in Japan and China (Fig. 8).Evident in a number of pubescent individuals col-lected in Korea, pubescence was not uniformly presentfrom all populations, except Mt. Du-ryun-san, Mt.Bang-tae-san, and Je-ju Island populations (Fig. 8).There was virtually no separation of taxa with respectto the degree of pubescence (Fig. 8). This study showedthat the degree of pubescence was continuous within apopulation as well as across populations. Therefore,the distribution of P. leveilleana and P. serrulata var.pubescens for this character overlapped considerablymore and could not be considered entirely significant.

Figure 4. Univariate statistics with the minimum and maximum values for the discriminating characters for flowers ofthe Prunus serrulata complex including P. sargentii. Many flower-related characters showed that values overlappedextensively for the P. serrulata complex. The length of the peduncle and the length from the end of the peduncle to thesecond pedicel were the only characters to separate P. sargentii from the P. serrulata complex. C, Chinese P. serrulata var.spontanea; K, Korean P. serrulata var. spontanea; J, P. jamasakura; PC, Chinese P. serrulata var. pubescens; PK, KoreanP. serrulata var. pubescens; PJ, P. leveilleana; SAR, P. sargentii. A, peduncle length (mm). B, distance from the end of thepeduncle to the second pedicel (mm). C, number of hairs on the petiole (in 1 mm). D, flower diameter (mm).

A B

0

10

20

30

40

C K J PC PK PJ SAR

Spec ies

0

3

6

9

12

C K J PC PK PJ SAR

Spec ies

C D

0

15

30

45

60

75

C K J PC PK PJ SAR

Spec ies

0

10

20

30

40

50

C K J PC PK PJ SAR

Spec ies



ANALYSIS OF PEDUNCLE LENGTH

According to our study, the average peduncle length ofP. serrulata var. spontanea and P. sargentii was foundto be 7.1 and 2.1 mm, respectively. In fact, it has beenknown that the length of the peduncle is typical anddistinctive in the P. serrulata complex and P. sargentii.

Peduncle lengths at first flowering and after fullblooming within the same individual of P. serrulatavar. spontanea were measured in order to determine

the growth of the peduncle over a given period of time.Although it varied with environmental conditions, thelength of the peduncle varied less than 3.5 mm fromthe first flower bloom to the full bloom for the sameindividual. For this reason, at the first flowering stage,it may be difficult to distinguish P. sargentii fromP. serrulata var. spontanea based only on the lengthof the peduncle. However, this study showed thatthe variability within individuals of P. serrulata var.spontanea with time maintained a certain degree of

Figure 5. Scatter diagrams from the principal components analysis results of fruits with leaves for taxa of the Prunusserrulata complex and related species. P. serrulata var. spontanea (�); P. jamasakura (�); P. serrulata var. pubescens (�);P. leveilleana (�); P. sargentii (�); P. lannesiana f. albida (�); P. pseudocerasus (X). A, PC 1 vs. PC 2. B, PC 1 vs. PC 3.

- 4

- 3

- 2

- 1

0

1

2

3

- 5 - 1 0 1 2 3 4

PC2

PC

1

- 4 - 3 - 2

- 3

- 2

- 1

0

1

2

3

4

- 5 - 4 - 3 - 2 - 1 0 1 2 3 4

PC3

PC

1

B

A



Figure 6. Univariate statistics with the minimum and maximum values for the most discriminating characters for fruitswith leaves of Prunus pseudocerasus and the P. serrulata complex. The length of the peduncle and pedicel characters, aswell as several leaf characters, contributed most to the separation among the three taxa presented here. L, P. lannesianaf. albida; P, P. pseudocerasus; S, the P. serrulata complex (including P. serrulata var. spontanea, P. jamasakura, P. serrulatavar. pubescens, and P. leveilleana). A, maximum leaf length (mm). B, maximum leaf width (mm). C, angle of the leaf base(°). D, length of the leaf apex (mm). E, number of serrations at the widest part of the leaf (number/2 cm). F, number oflateral veins. G, peduncle length (mm). H, pedicel length (mm).

A B

0306090

120150

L P S

Species Species

Species Species

Species species

Species Species

0

30

60

90

120

L P S

C D

0

40

80

120

160

L P S

0

10

20

30

L P S

E F

0

10

20

30

L P S

0

4

8

12

16

L P S

G H

0

10

20

30

L P S0

20

40

60

L P S



Figure 7. Univariate statistics with the minimum and maximum values for characters of fruit with leaves of the Prunusserrulata complex and P. sargentii. Many vegetative characters showed that values overlapped extensively for theP. serrulata complex. The length of the peduncle and the distance from the end of the peduncle to the second pedicel werethe only characters for P. sargentii. C, Chinese P. serrulata var. spontanea; K, Korean P. serrulata var. spontanea; J,P. jamasakura; PC, Chinese P. serrulata var. pubescens; PK, Korean P. serrulata var. pubescens; PJ, P. leveilleana; SAR,P. sargentii. A, maximum leaf length (mm). B, maximum leaf width (mm). C, angle of the leaf apex (°). D, number ofserrations at the widest part of the leaf (number/2 cm). E, number of lateral veins. F, the widest portion of the long axisof the leaf. G, distance from the end of the peduncle to the second pedicel (mm). H, length of the peduncle (mm). I, numberof hairs on the abaxial of the leaf (25 mm2). J, number of hairs on the adaxial of the leaf (25 mm2). K, number of hairs onthe petiole (1 mm). L, number of hairs on the pedicel (1 mm).

A B C

0306090

120150

C K J PC PK PJ SARSpec ies

0

30

60

90

120

C K J PC PK PJ SARSpecies

0

40

80

120

160


D E F

0369

121518

C K J PC PK PJ SARSpc ies

0369

1215


0

1

2

3

4


G H I

0

4

8

12

16


0

10

20

30


05

10152025


J K L

050

100150200250


020406080

100


0

15

30

45

60




morphological integrity (at least > 5 mm in pedunclelength), remaining somewhat distinct fromP. sargentii (Fig. 9).

TIME OF FLOWERING

With respect to flowering time, there may be 2 or3 weeks between the first bloom and the first sheddingof petals (Kuitert & Peterse, 1999). However, thisstudy based on an observation of 41 individuals ofP. serrulata var. pubescence showed that most individ-uals were short lived, only c. 10–11 days existingbetween the first flower opening and the first petalshed, although the total length of the flowering seasonwithin the population was 20 days. A few individuals(e.g. individuals 7 and 30 in Figure 10) within the

population came to bloom a week or two later, com-pared with most individuals.

DISCUSSION

The results of the PCA indicated that all related taxaof the P. serrulata complex, such as P. pseudocerasus(Fig. 5), P. lannesiana f. albida (Figs 2, 5), andP. sargentii (Figs 2, 5) were morphologically distinctwithout a major morphological continuum. On theother hand, morphological variations for most charac-ters were continuous with a high degree of overlapacross taxa within the P. serrulata complex.

RELATED TAXA OF THE P. SERRULATA COMPLEX

Prunus lannesiana f. albida, which can be separatedfrom P. lannesiana f. lannesiana in having simpleflowers, is characterized by a corymbose inflorescencewith glabrous leaf, petiole and pedicel, and awn-teethof the leaf. Several authors (Koidzumi, 1911; Makino,1912; Koehne, 1913; Wilson, 1916; Kuitert & Peterse,1999) placed this Japanese endemic at the varietalor form rank under P. serrulata, as P. serrulata var.serrulata f. albida Makino, P. serrulata var. albidasubvar. speciosa (Koidz.) Makino, or P. serrulatavar. speciosa (Koidz.) Koehne. Ohba (2001) treatedP. lannesiana f. albida even at the rank of species,Cerasus speciosa (Koidz.) H. Ohba. This study showedthat P. lannesiana f. albida differed from other relatedtaxa in always having three to four flowers per inflo-rescence, large flowers with serrated petals, and veryacute and awn-teeth shape of leaf serration (Wilson,

Figure 8. Variation in hair density on pedicels of the Prunus serrulata complex within a population as well as amongpopulations. BH, Mt. Buk-han-san; KA, Mt. Gwan-ak-san; KK, Mt. Gwang-gyo-san; KR, Mt. Key-ryong-san; JR, Mt. Ji-ri-san; DR, Mt. Du-ryun-san; JJ, Je-ju-do; BT, Mt. Bang-tae-san; SB, Mt. So-baek-san; DT, Mt. Du-ta-san; CH, China; JP,Japan. Symbols for hair density (1 mm2): 0 (�); 1–10 ( ); 11–20 ( ); 21–30 ( ); 31–40 ( ); 41–50 ( ); 51–55 ( ).

0%

20%

40%

60%

80%

100%

BH KA KK KR JR DR JJ BT SB DT CH JP

Locality

Per

cent

age

of fr

eque

ncy

Figure 9. Frequency distribution of peduncle length ofPrunus serrulata var. spontanea between the first flowering(�) and the first shedding of the petals (�).



1916; Kitamura & Murata, 1979; Ohwi, 1984). Thistaxon is known to be restricted to a small geographicalarea of Honshu: Shizuoka, Yamanashi, Kanagawa pre-fectures and the Izu peninsula. Current results sup-port the separation of P. lannesiana f. albida fromother taxa of the P. serrulata complex as a distinctivetaxon.

Prunus pseudocerasus in central and eastern Chinaexhibited morphological similarity to P. serrulata andwas recognized as being close to P. serrulata in China,Korea, and Japan by Maximowicz (1883) and Makino(1908). Our morphometric analysis showed thatstrong discontinuities existed between the P. serrulatacomplex and P. pseudocerasus. Comparisons of theunivariate, conceptually related means and range,pointed out the differences between P. pseudocerasusand other taxa with respect to the number of flowersper inflorescence [three flowers vs. two (three) flow-ers], umbel type (relatively short pedicel and pedun-cle), small leaves, short apex of the leaf, pubescence ofthe leaf vein, petiole and pedicel, and black (vs. red)drupes (Koehne, 1913; Wilson, 1916; Yü & Li, 1986).

Prunus sargentii is generally characterized by twoor three flowers in an umbel inflorescence with a shortpeduncle and glabrous leaf, petiole, and pedicel(Rehder, 1908). Wilson (1916) placed P. sargentii atvarietal rank of P. serrulata, namely P. serrulata var.sachalienensis (F. Schmidt) E. H. Wilson (Koidzumi,1911), rather than at species rank. A detailed investi-gation of this taxon in both Korea and Japan revealeda comparable degree of continuous variation inleaf characters with P. serrulata var. spontanea andP. lannesiana f. albida, but due to an umbel inflores-cence with a short peduncle, a relatively large flowersize, and sticky buds and young twig (Kuitert &Peterse, 1999), it could be supported as an indepen-dent species, as Wilson suggested (Wilson, 1916;Chang, Choi & Chang, 2004).

In general, the geographical distribution should betaken into consideration along with an analysis of sim-ilarities and differences in morphological charactersas a basis for evaluating relationships among taxa(Radford et al., 1974). On the basis of observation ofherbarium specimens, as well as the literature,

Figure 10. Distribution of the flowering period of 41 individual trees of Prunus serrulata var. pubescens from April toMay 2004.



P. sargentii seems to be restricted to northern parts ofKorea, with a few exceptions, northern Japan, and fareastern Russia, whereas the P. serrulata complex wasdistributed from central and eastern China to south-ern Japan through central and southern Korea(Kurata, 1971; Kitamura & Murata, 1979; Ohwi,1984; Kuitert & Peterse, 1999).

One of the key characters used to distinguishbetween P. sargentii and the P. serrulata complex isthe difference in flower size (Kitamura & Murata,1979; Ohwi, 1984). Indeed, this study demonstratedthat between P. sargentii and the P. serrulata complexfrom some populations in eastern Korea, flower sizeemerged as the most distinct (34–48 mm vs. 30–37 mm). However, inspection of the flower size ofindividuals from southern populations (e.g. Mt. Duk-you-san) of P. sargentii (36–37 mm) revealed thatthere was extensive overlap with the P. serrulata com-plex. Therefore, the use of flower size to discriminatebetween P. sargentii and the P. serrulata complex wasnot reliable.

Prunus verecunda (Koidz.) Koehne has received var-ied treatments in recent studies. Koidzumi (1911)originally described it as P. jamasakura var. vere-cunda Koidz., but later Koehne placed this taxon atthe rank of species (Koehne, 1912; Nakai, 1930).Nakai (1916) and Lee (1980) regarded this taxon as avariety of P. serrulata, namely P. serrulata var. vere-cunda. Although Kitamura & Murata (1979) andOhba (2001) treated it as a synonym of P. leveilleana,Wilson (1916) and Kuitert & Peterse (1999) did notdistinguish between P. serrulata var. pubescens andP. verecunda due to the presence of a pubescent petioleand pedicel, listing only P. serrulata var. pubescens.

Prunus verecunda is characterized by umbellateinflorescences with a very short peduncle (< 4 mmlong) and one to two (to three) flowers per inflores-cence. Except for the pubescence, plants ofP. leveilleana and P. serrulata var. pubescens weremarkedly different from P. verecunda due to very longpeduncles and corymbs (Koehne, 1912). In addition,the distribution of P. verecunda is limited to northernJapan, which is correlated with that of P. sargentii,not that of the P. serrulata complex. The inspection of74 specimens from eastern Korea revealed that 14individuals were consistent with the characters ofP. verecunda by the presence of a short pubescentpeduncle, also differing from the P. serrulata complexin inflorescence type. Accordingly, our systematicinterpretation of P. verecunda is not in agreementwith the previous taxonomists (Koidzumi, 1911; Koe-hne, 1912; Nakai, 1916, 1930; Wilson, 1916; Kitamura& Murata, 1979; Lee, 1980; Kuitert & Peterse, 1999;Ohba, 2001), who regarded this taxon as a variety ofP. serrulata. It appears that plants from northernJapan and eastern Korea fall well within the bounds

of P. verecunda. Therefore, we reduce the status of spe-cies P. verecunda to a varietal status of P. sargentii, i.e.P. sargentii var. verecunda (Koidz.) C.S. Chang (Changet al., 2004).

In addition, P. takesimensis Nakai, which isrestricted to the island of Ulleong-do, east of Korea,is recognized as being related to P. sargentii.P. takesimensis is characterized by having relativelysmall flowers (21–32 mm in diameter; in general 28–46 mm in diameter in the P. sargentii complex) andmany flowers per inflorescence (two to five flowers; ingeneral one to three flowers). Evidence obtained frommorphological analyses indicated that the entity ofP. takesimensis formed a cohesive group somewhatdistinct from P. sargentii. Therefore, this taxon shouldbe recognized as an independent and endemic taxon toKorea (Chang et al., 2004).

THE P. SERRULATA COMPLEX

Prunus jamasakura was placed at the rank of speciesin Japan by several authors (Kitamura & Murata,1979; Ohwi, 1984), whereas this taxon was considereda variety of P. serrulata, i.e. namely, P. serrulata var.spontanea in Korea and China (Koehne, 1913; Wilson,1916; Lee, 1980). As exemplified by the current PCAanalysis, P. jamasakura and P. serrulata are some-what distinct taxa, distinguished by leaf characteris-tics (oval vs. elliptic shape), but univariate statistics(Fig. 7), in addition to the minimum and maximumvalues for leaf characters (a greater number ofserrated teeth and leaf veins) from the plants ofP. serrulata var. spontanea (Chinese and Korean indi-viduals) and P. jamasakura investigated in this studyshowed that values overlapped extensively. Therefore,a weak separation between P. serrulata var. spontaneaand P. jamasakura was confirmed by this study.

According to a representative study of the floweringtime and period of the P. serrulata complex, the flow-ering period of an individual flower was c. 10 days. Theshort and explosive spring in eastern Asia generallyhas only 3 weeks between the average early and theaverage late cherry blossoms (Kuitert & Peterse,1999). Two weeks later, flowers on individuals ofP. serrulata var. pubescens at any specific site (Kita-mura & Murata, 1979) may not be uncommon andoccur in wild populations (Fig. 10). This characteristicdifference between P. serrulata var. pubescens andP. leveilleana does not justify the recognition ofP. leveilleana at the rank of species.

Recent studies of Japanese Prunus in flora of Japanrevised by Ohba (2001) caused further taxonomic con-fusion. P. leveilleana was recognized under the genusCerasus [cited as Cerasus leveilleana (Koehne)H. Ohba]. Specifically, Ohba (2001) characterizedP. serrulata var. pubescens and P. verecunda as forms



of P. leveilleana, and merged the two names toC. leveilleana (Koehne) H. Ohba (= P. serrulata var.pubescens), although he recognized P. verecunda atthe species rank in his previous paper (Ohba, 1992). Itimplied that he never distinguished the P. serrulatacomplex from P. sargentii and P. verecunda, which wasusually characterized by flowers in an umbel inflores-cence with a short peduncle and pedicel. In addition,Ohba (2001) recognized two major taxa,C. jamasakura (Siebold ex Koidz.) H. Ohba andC. leveilleana (Koehne) H. Ohba as the P. serrulatacomplex, but he treated P. jamasakura var. pubescens(Makino) Nakai as a synonym for both P. leveilleanaand P. jamasakura (= P. serrulata var. spontanea).Therefore, his concept of P. leveilleana, as well asP. serrulata var. pubescence, was very confusing, andthese treatments should not be accepted.

Furthermore, there has been a great deal of confu-sion regarding C. sargentii var. akimotoi H. Ohba &Mas (Ohba, 2001). As described, this taxon is charac-terized by a long peduncle (3.5 mm) and triangularpetals, which seems to be associated with theP. serrulata complex, rather than P. sargentii. The geo-graphical range of C. sargentii var. akimotoi through-out southern Kyushu is similar to that of theP. serrulata complex. Therefore, Ohba (2001) misap-plied this name to a variety of P. sargentii, and to thebest of our knowledge, this taxon should be considereda variety of the P. serrulata complex with respect togeography and inflorescence.

CONCLUSIONS

In summary, P. pseudocerasus, P. lannesiana f. albida,and P. sargentii are distinct taxa based on our resultsfrom the morphometric analyses. The leaf, flower, andfruit data may reflect a lack of divergence amonginfraspecific taxa of P. serrulata. The only cohesivevegetative morphology is the degree of pubescence.Therefore, it is not correct to split P. serrulata into anyvarieties based on morphology, except for the pubes-cent taxon, P. serrulata var. pubsecens. We concludethat the many infraspecific taxa of P. serrulata shouldbe recognized as a single species with two varieties,var. serrulata and var. pubescens.

TAXONOMY

1. PRUNUS SERRULATA LINDL., TRANS. HORT. SOC. LOND. 7: 238–239, 1830

Type: No date without collector (Holotype: A!, seen asa photograph).

1A. VAR. SERRULATA F. SERRULATA

Cerasus serrulata (Lindl.) G.Don, Hort. Brit. (Laudon)480, 1830.

Cerasus serratifolia Lindl. ex Carrière, Rev. Hort. 389–390, 1877. (Holotype: A!, seen as a photograph).Prunus serrulata f. mucronata Koehne, Mitt. Deutsch.Dendrol. Ges. 18: 170, 1909. (Type: probably not col-lected from Arboretum Spath by Koehne.)

Representative specimens examined: None.

Distribution: Known only in cultivation.

1B. VAR. SERRULATA F. SPONTANEA (MAXIM.) C.S. CHANG, COMB. ET STAT NOV.

Prunus pseudocerasus var. spontanea Maxim., Bull.Acad. Imp. Sci. Saint. Pétersbourg 29: 102, 1883. Type:Not identified, probably at LE.Prunus serrulata var. spontanea (Maxim.) E. H.Wilson, Cherries Japan: 28–30, 1916.Prunus donarium var. spontanea (Maxim.) Makino,Illus. Fl. Nippon: 438, 1940.

Cerasus montana Siebold ex Miq., Ann. Mus. Bot. Lug-duno-Batavi 2: 90, 1865. Type: Not identified, probablyat K.Prunus pseudocerasus var. jamasakura subvar. glabraMakino, Bot. Mag. (Tokyo) 22: 93–98, 1908. Type: Notfound at MAK.Prunus jamasakura var. elegans f. glabra (Makino)Koidz., Bot. Mag. (Tokyo) 25: 185, 1911.Prunus donarium ssp. elegans f. glabra (Makino)Koidz., J. Coll. Sci. Imp. Univ. Tokyo 34, 2: 266, 1913.Prunus serrulata var. glabra (Makino) Nakai, Bot.Mag. (Tokyo) 29: 139, 1915.Prunus mutabilis f. glabra (Makino) Nakai, Iconogr.Pl. Asiae Orient. 4–4: 434, 1941.Prunus jamasakura Siebold ex Koidz., Bot. Mag.(Tokyo) 25: 184–185, 1911. Type: Not found at L.Prunus serrulata var. ungerii Sprenger, Gartenwelt 10:89, 1905. Type: Not identified.Prunus serratifolia var. nageri Sprenger, Bull. Soc.Tosc. Ortic. 31: 178–179, 1906. Type: Not identified.Prunus mutabilis Miyoshi, J. Coll. Sci. Imp. Univ.Tokyo 34, 1: 35, 1916. Type: Not found at TI.Prunus densifolia Koehne, Repert. Spec. Nov RegniVeg. 12: 135, 1913. Type: Korea, Quelpaert (= Prov.Jeju-do), May 1911, Taquet 5594 (Holotype at A!, seenas a photograph).Prunus serrulata var. densifolia (Koehne) Uyeki,Woody Pl. 54, 1940.Prunus serrulata f. densifolia (Koehne) W. T. Lee, Lin-eamenta Florae Koreae 501, 1996.

Representative specimens examined: JAPAN: Pref.Chiba: Chiba city, Toge-cho, 50–100 m alt. Im, H.T. &T. Kawahara 10099 (TI); Pref. Ehime: Kitauwa-gun,Yoshida-machi, Chinaga, J. Murata 150421 (TI); Pref.



Fukushima: Taira (Iwaki), s.leg. (TI); Pref. Hyogo:Akuo city, along the road between Shinden andHosaka-tooge, 10–100 m alt. N. Fukuoka 12050 (TI);Pref. Kanagawa: Ashigarashimo-gun, Hakone (Sag-ami), G. Koidzumi s.n. (TI); Pref. Kanagawa: Hakone,Sokokura, T. Sawada s.n. (TI); Pref. Kochi: Befu, atthe western foot of Mt. Ishidate, 400–500 m alt., H.Ohashi s.n. (TI); Pref. Kochi: Onogahara to Rokutyo,Shimagawa, Yuzuhara-mura, Tokaoka-gun, 900 malt., G. Murata & T. Shimizu s.n. (MAK−63190); Pref.Kyoto: Kitakuwada-gun, Miyama-cho, Ashio KyotoUniversity Forest, Uchisugi Dani, 590 m alt. T. Fujiiet al. 14807 (TI); Pref. Mie: Sima-gun, Ago-cho,Kokofu, Wabe, 0–30 m alt., S. Noshiro et al. (fieldnumber 943) (TWT-15534, TI); Pref. Nara: Yamabe-gun, Yamazoe-mura, Kanno-yama, 400–618 m alt., G.Murata 44519 (TI); Pref. Nara: Gose-city, Ishikawa, enroute from the summit Mt. Katsuragi-san to Ishikawa,200–900 m alt., N. Kurosaki 10636 (MAK); Pref.Okayama: Kawakami-mura, Mt. Maruyama, T. Sug-awara s.n. (MAK−253658); Pref. Tokyo: Nagafuchi, c.150–250 m alt., T. Kawahara & Im, H.T. 915 (TI); Pref.Tokyo: Musa-shi, G. Koidzumi s.n. (TI); Pref. Tokyo:Hachioji-shi, Nagafusa, spontaneous in Asakawa Exp.Forest, on sunny roadside on secondary wood, H.Taoda 3651 (MAK); Pref. Tokyo: Hachioji-shi,Nagafusa, spontaneous in Asakawa Exp. Forest, onsunny roadside on secondary wood, H. Taoda 3743(MAK); Pref. Tokyo: Y. Momiyama s.n. (MAK-235789);Pref. Tokyo: Y. Momiyama s.n. (MAK-268277); Pref.Wakayama: vicinity of Tonabe-city, c. 50 m alt. F.Konta 4935 (TI).

CHINA: Prov. Anhui (PE−1397335); Prov. Guling:Kangsi, R.C. Ching 4798 (PE); Prov. Heilongjiang:(PE−0852644); Prov. Hubei: (PE−1247852); Prov.Hubei: Y-hsien, K.M. Liou 2023 (PE); Prov. Jiangxi:(PE−818107); Prov. Shandong: Li-chuan, Shui-sha-pa,4,300ft, alt., C.T. Hwa 358 (PE); Prov. Zhejiang: EastTienmu-Shan, W.C. Cheng 2281 (PE); (PE−362675);(PE−679714); (PE−912891); (PE−926015); (PE−915245); (PE−920847).

KOREA: Prov. Chungcheongnam-do: Gong-ju city;Mt. Kye-ryong-san from entrance to Nam-mae-top,JKS123 (SNUA); JKS388 (SNUA); Prov. Gangwon-do:Go-seong-gun; Gan-seong-eup, from Mt. Chil-chul-bong to Mt. Hyang-lo-bong, T.B. Lee s.n.(SNUA−26932); Prov. Gangwon-do: To-seong-myeon,Sin-pyeong-ri, from the gate of Hwa-am-sa to summitthrough Cheong-gan-cheon valley, U. Kang 0419(SNUA); Prov. Gangwon-do: In-je-gun; Mt. Bang-tae-san, JKS340 (SNUA); Prov. Gangwon-do: Jin-dong-ri,Mt. Jeom-bong-san, from Jin-dong valley to the top ofMt. Jum-bong-san, C.S. Chang et al. JB016 (SNUA−26871); Prov. Gangwon-do: Sok-cho city; Mt. Sol-ak-san, S.G. March et al. 210 (SNUA), Prov. Gyeonggi-do:An-san city; Dae-bu-do (Island), C. S. Chang et al.KGD061 (SNUA); Prov. Gyeonggi-do: An-yang city;Mt. Gwan-ak-san T.B. Lee & E.S. Kim s.n. (SNUA−26860); Prov. Gyeonggi-do: Mt. Gwan-ak-san from TheArboretum to Seoul National University through Mu-Neo-mi-go-gae, JKS233 (SNUA); Prov. Gyeonggi-do:Ha-nam city; Chang-woo-dong, Mt. Geom-dan-san,Jeon10922 (SNUA); Prov. Gyeonggi-do: Hwa-seongcity; Buk-yang-dong, Mt. Mu-bong-san, c. 0.5–1.5 km

KEY TO THE PRUNUS SERRULATA COMPLEX AND RELATED TAXA

1. Fruits red, pedicel 15–21 mm; leaf apex (7–)10–12(−13) mm (see Figure 2); calyx tube as long as calyx lobes................................................................................................................................................................................ P. pseudocerasus

1. Fruits dark reddish-purple to almost black; pedicel (11–)21–30(−58) mm; leaf apex (6–)14–25(−29) mm, calyx tube 1.5times as long as lobes2. Flowers in corymbs, usually with a common peduncle or with a long peduncle [(0–)3–27(−34) mm]

3. Flowers in corymbs of (two)three to four, flower 41–51(55) mm in diameter, calyx lobes serrulate with minuteglandular-tipped teeth, peduncle (5–)6–16(−20) mm, leaf awn-teeth c. 3.8–4 mm4. Flowers double .......................................................................................................... P. lannesiana f. lannesiana4. Flowers single.................................................................................................................... P. lannesiana f. albida

3. Flowers in corymbs of two to three, flowers (22–)30–37(−45) mm in diameter, calyx lobes entire or nearly so,peduncles (0–)3–27(−34) mm, leaf awn-teeth c. (0.5–)2–3(−4) mm5. Flowers double ................................................................................... 1a. P. serrulata var. serrulata f. serrulata5. Flowers single

6. Petiole, pedicel, and peduncle glabrous...................................... 1b. P. serrulata var. serrulata f. spontanea6. Petiole, pedicel, and peduncle pubescent ......................................................... 2. P. serrulata var. pubescens

2. Flowers in umbels, usually without a common peduncle or with a very short peduncle [(0–)1–2.5(−5) mm]7. Flowers (one)two to three, (28–)34–41(−46) mm in diameter

8. Petiole, pedicel, and peduncle glabrous...................................................................P. sargentii var. sargentii8. Petiole, pedicel, and peduncle pubescent ..............................................................P. sargentii var. verecunda

7. Flowers (two)three to four(five), flowers (21–)25–29(−32) mm in diameter ..............................P. takesimensis



from the Bong-rim-sa (Temple), LHI0094 (SNUA);Prov. Gyeonggi-do: Gwa-cheon city; Mt. Cheong-kye-san, T.B. Lee et al. s.n. (SNUA26849); Prov. Gyeonggi-do: Su-won city; Mt. Gwang-gyo-san, side of Su-gi-ri,JKS081 (SNUA); Prov. Gyeonggi-do: Mt. Gwang-gyo-san, from Gyeong-gi University to Wit-son-gol throughsummit, JKS369-5 (SNUA); Prov. Gyeonggi-do: Mt.Gwang-gyo-san, from Gyeong-gi University to Ji-ji-dae-go-gae, U. Kang 0622 (SNUA); Prov. Gyeonggi-do:Mt. Gwang-gyo-san, T.B. Lee s.n. (SNUA−26861); Prov.Gyeonggi-do: Yang-pyeong-gun; Mt. Chu-eup-san,from village of San-su-yu, JKS1182 (SNUA); Prov.Jeju-do: Je-ju city; Mt. Hal-la-san, T.B. Lee s.n.(SNUA−26915); Prov. Jeju-do: Mt. Hal-la-san, Y-Val-ley, JKS107 (SNUA); Prov. Jeju-do: Seo-gui-po city,T.B. Lee & J.Y. Park s.n. (SNUA−26848); Prov. Jeol-lanam-do: Dam-yang-gun; Su-buk-Myeon, from Dae-bang-reservoir (alt. 120 m) to Byeong-pung-san (alt.822 m), H.T. Im 011792 (SNUA); Prov. Jeollanam-do:Gu-rye-gun; Mt. Ji-ri-san, from Seong-sam-jae to Tem-ple Cheon-wang-sa, JKS214 (SNUA); Prov. Jeollanam-do: Mt. Ji-ri-san, from Seong-sam-jae to Jik-jeon-ri,JKS903 (SNUA); Prov. Jeollanam-do: Gwang-jin-gun;Mt. Wol-chul-san, from camping ground to Gu-geong-chi, H.T. Im & S. Y. Park 037708 (SNUA); Prov. Jeol-lanam-do: Hae-nam-gun; Mt. Du-ryun-san; from tem-ple Dae-heong-sa to summit, JKS1193 (SNUA); Prov.Jeollanam-do: Wan-do-gun; Peak Sang-hwang-bong,JKS1030 (SNUA); Seoul: Mt. Buk-han-san, Y. H. Gongs.n. (SNUA−26865); Seoul: Mt. Buk-han-san, from Gu-gi-dong to Bi-bong, JKS239 (SNUA); Seoul, JKS398(SNUA).

Distribution: North-eastern and central China(Heilongjiang, Hubei, Shandong, Zhejiang, Jiangxi,Anhui, Guling, Jiangsu, Hunan, Guizhou), southernand central Japan (Kyushu to southern Honshu), andKorea.

Taxonomic note: Prunus serrulata was described byLindley (1830) to include double white flowers as agarden cultivated form (Wilson, 1916). Wilson (1916)described single flowers as a wild form, P. serrulatavar. spontanea. This double vs. single flower characteris not considered to be sufficiently differentiatedbetween P. serrulata var. serrulata and P. serrulatavar. spontanea to warrant the designation of the rankof variety. Instead, a plausible suggestion is that var.spontanea be recognized as a form. Therefore, a newcombination is published here.

There has been a little confusion in the literatureregarding the correct author for P. jamasakura.‘Jamasakura’ was used by von Siebold (1830), not as abasionym, but as a common name. Koidzumi (1911)was the first taxonomist to use this name for thetaxon. This study confirms only a weak separation

between P. jamasakura in Japan and P. serrulata var.spontanea in Korea and China. This study also showsthat P. jamasakura and P. serrulata var. spontaneashould be treated as synonyms of P. serrulata var. ser-rulata f. spontanea.

We have not found a type specimen of P. pseudocer-asus var. jamasakura subvar. glabra Makino at MAK.This name was probably based on cultivated plants ingardens without preserving a specimen.

Wilson (1916) and Rehder (1949) consideredP. tenuiflora Koehne, which occurred in central China(western Hubei = Hupeh), conspecific with P. serrulatavar. spontanea. These two taxa are clearly separableon the morphological observations of peduncle length,inflorescence type, and petal form of nine syntypesdeposited at A. It seems that P. tenuiflora is probablyrelated to P. conradinae Koehne, rather thanP. serrulata var. spontanea.

2. PRUNUS SERRULATA VAR. PUBESCENS (MAKINO) NAKAI, BOT. MAG. (TOKYO) 29: 140, 1915

Prunus pseudocerasus var. jamasakura subvar. pube-scens Makino, Bot. Mag. (Tokyo) 22: 98–99, 1908.Types: Prov. Yamato (= Pref. Nara), April 1895, S.Mastuda (syntype: probably at MAK, not found); Prov.Musashi (= Pref. Tokyo), Tokyo, cult. from Hokkaido,April and June 1908, T. Makino (syntype: probably atMAK, not found).Prunus jamasakura var. elegans subvar. pubescens(Makino) Koidz., Bot. Mag. (Tokyo) 25: 185, 1911.Prunus jamasakura var. pubescens (Makino) Nakai, J.Coll. Sci. Imp. Univ. Tokyo 31: 482, 1911.Prunus tenuiflora var. pubescens (Makino) Koehne,Repert. Spec. Nov Regni Veg. 11: 268, 1912.Prunus donarium ssp. elegans var. pubescens (Makino)Koidz., J. Coll. Sci. Imp. Univ. Tokyo 34: 269, 1913.Prunus serrulata var. spontanea subvar. pubescens(Makino) Makino in Makino & Nemoto, Catal. Jap. Pl.Herb. Tokyo Imp. Mus. 222, 1914.Prunus mutabilis var. pubescens (Makino) Nakai,Iconogr. Pl. Asiae Orient. 4–4: 435, 1941.Prunus jamasakura f. pubescens (Makino) Ohwi, Fl.Jap. 657, 1961.Prunus leveilleana Koehne in Sargent, Pl. Wilson 1:250, 1912. Syntypes: Korea, in the mountains ofMokhpo (= Mokpo of Prov. Jeol-la-nam-do), March1909 (lectotype, here designated A!); T. Taquet 2519;ibid., May 1909, T. Taqeuet 2517 (isolectotype: A!).Prunus serrulata var. tomentella Nakai, Bot. Mag.(Tokyo) 29: 140, 1915. Type: Korea, Gwangneung (=Prov. Gyeonggi-do, Pocheon-gun, Sohol-eup, Gwangne-ung) 7 July 1912, T. Mori 235 (holotype: at TI !, seen asa photograph).Prunus mesadenia Koehne in Sargent, Pl. Wilson 1:250–251, 1912. Type: Swasima, Japan. 2 April 1879, J.Matsumura s.n. (probably at B, not seen).



Prunus quelpaertensis Nakai, Repert. Spec. Nov RegniVeg. 13: 276, 1914. Type: Korea, Quelpaert (= Prov.Jeju-do), May 1911, Taquet 5597 (holotype: at TI!, seenas a photograph).

Representative specimens examined: JAPAN: Pref.Gunma: Tano-gun, Manba-machi, Mt. Nishimikabu,near Nageishitoge, 800 m alt., J. Murata et al. 7215(TI); Pref. Gunma: Kiryu-shi, Mt. Narukamu-yama(west side), 600–979 m alt., J. Murata et al. 11026(TI); Pref. Hyogo: Bettsho-cho, Miki-shi, 50 m alt., G.Murata & N. Nishimura 638 (TI); Pref. Osaka:Hirakata-shi, Nagao, Toge-machi, J. Murata 30399(TI); Pref. Tochiki: Lunperiali, praedio, Rostico,Nasueusi, A. Kimura 5107 (TI); Pref. Tottori: Misasa-machi, Mt. Mitoku, 600 m alt., M. Togashi 10225 (TI);Pref. Yamagata: Hagurosan, Minamidani, T. Yoshino30399 (TI).

CHINA: Prov. Hubei (PE−611112); Prov.Heilongjiang (PE−318006); Prov. Liaoning: (PE−255020); (PE−0596353); Prov. Shandong (PE−638718);Prov. Zhejiang (PE−625358).

KOREA: Prov. Chungcheongnam-do: Bo-eun-gun;Mt. Sok-ri-san, T.B. Lee s.n. (SNUA−26957); Prov.Chungcheongnam-do: Cheonan city; Yeon-gi-gun,Gwang-duk-myeon, Mt. Gwang-deok-san, C. S. Chang2948 (SNUA); Prov. Chungcheongnam-do: Gong-jucity, Mt. Kye-ryong-san, from entrance to Nam-mae-top, JKS114 (SNUA); Prov. Chungcheongnam-do:JKS383 (SNUA); Prov. Chungcheongnam-do: Tae-an-gun, An-myeon-do, T.B. Lee et al. s.n. (SNUA−26968);Prov. Chungcheongnam-do: Ye-san-gun; Hae-mi-eup,Mt. Ga-ya-san, C. S. Chang 3806 (SNUA); Prov. Gang-won-do: Go-seong-gun; Gu-jin-eup, Mt. Gun-bong-san,HR279 (SNUA); Prov. Gangwon-do: In-je-gun; Mt.Bang-tae-san, JKS346 (SNUA); Prov. Gangwon-do:Sok-cho city; Buk-eom-ryeong of Mt. Seol-ak-san, T.B.Lee & M.Y. Cho s.n. (SNUA−26976), Prov. Gangwon-do: Won-ju city; So-cho-myeon, Mt. Chi-ak-san, S. G.March et al. 152 (SNUA); Prov. Gangwon-do: Yang-gu-gun; Dong-myeon, Mt. Dae-am-san, C.S. Chang et al.TA085 (SNUA−42004); Prov. Gyeonggi-do: An-sungcity; Juk-san-myeon, Mt. Chil-hyun-san, DY Choi 244(SNUA); Prov. Gyeonggi-do: An-yang city; Gwa-cheoncity; Mt. Gwan-ak-san, hiking trail from Gwa-cheon toAn-yang city through Pal-bong trail, C. S. Chang 2515(SNUA); Prov. Gyeonggi-do: Gwan-ak-san, G.T. Kim& S.Y. Lee s.n. (SNUA−26842); Prov. Gyeonggi-do:E.H.Choi s.n. (SNUA−26935); Prov. Gyeonggi-do: Mt.Gwan-ak-san, from The Arboretum to Seoul NationalUniversity through Mu-Neo-go-gae, JKS176 (SNUA);Prov. Gyeonggi-do: Mt. Gwanak-san, T.B. Lee s.n.(SNUA−26806); Prov. Gyeonggi-do: Mt. Su-ri-san, T.B.Lee et al. s.n. (SNUA26822); Prov. Gyeonggi-do: T.B.Lee & J.Y. Park s.n. (SNUA−26788); Prov. Gyeonggi-do:Hwa-seong city; Hwa-seong-gun, Seo-sin-myeon, Mt.

Gu-bong-san c. 0.5–2 km from Sin-heung-sa (Temple)to Dang-seong (Fortress), LHI0085 (SNUA−75723);Prov. Gyeonggi-do: Hwa-seong-gun, Seo-sin-myeon,Mt. Gu-bong-san c. 0.5–2 km from Sin-heung-sa (Tem-ple) to Dang-seong (Fortress), LHI0265 (SNUA); Prov.Gyeonggi-do: Su-won city; Gwon-seon-gu, Ho-mae-sil-dong, Mt. Chil-bo-san. c. 0.5–3 km from SeoulNational University, LHI0155 (SNUA); Prov. Gyeo-nggi-do: Mt. Gwang-gyo-san, side of Su-gi-ri, JKS082(SNUA); Prov. Gyeonggi-do: Mt. Gwang-gyo-san, fromGyeong-gi University to Wit-son-gol through summit,JKS368 (SNUA); Prov. Gyeonggi-do: Mt. Gwang-gyo-san, Wit-gol to Su-ji-eup through Heyng-je-bong,JKS524 (SNUA); Prov. Gyeonggi-do: Yang-pyeong-gun; Mt. Yong-moon, CBL0106 (SNUA); Prov. Gyeo-nggi-do: Yang-pyeong-eup, Mt. Chu-eup-san, JKS870(SNUA); Prov. Gyeonggi-do: JKS 829 (SNUA); Prov.Gyeonggi-do: Yang-pyeong-myeon, Mt. Sam-gak-san,JKS1155 (SNUA); Prov. Gyeonggi-do: Yang-pyeong-gun, Mt. Chu-eup-san, from village of San-su-yu,JKS1179 (SNUA); Prov. Gyeongsangbuk-do: Cheong-song-gun; Sang-eup-ri, Mt. Ju-wang-san, C. S. Chang3135 (SNUA); Prov. Gyeongsangbuk-do: Mun-Gyeongcity, Mt. Un-dal-san, C. S. Chang et al. UD015(SNUA); Prov. Gyeongsangbuk-do: Yeong-yang-gun;Mt. Il-wol-san, C. S. Chang 3788 (SNUA); Prov. Gyeo-ngsangnam-do: San-cheong-gun, San-cheong-eup,Nae-ri, Mt. Woong-seok-bong c. 3.0–4 km from theJi-gok-sah, Sky Park 0038 (SNUA); Prov. Gyeongsang-nam-do: Si-cheon-myeon, Mt. Ji-ri-san, from Moon-chang-dae to Ro-tae-ri, Sky Park 0069 (SNUA); Prov.Jeollabuk-do: Jeong-eup city, Mt. Nae-jang-san, H.T.Lee & D.M. Im s.n. (SNUA−26496); Prov. Jeollanam-do: Dam-yang-gun, T.B. Lee s.n. (SNUA−26962); Prov.Jeollanam-do: Gu-rye-gun, No-go-dan of Mt. Ji-ri-san,C. S. Chang 3746 (SNUA); Prov. Jeollanam-do: Mt. Ji-ri-san, from Seong-sam-jae to temple Cheon-wang-sa,JKS210 (SNUA); Prov. Jeollanam-do: Mt. Ji-ri-san,from Seong-sam-jae to temple Cheon-wang-sa,JKS377-1 (SNUA); Prov. Jeollanam-do: Hae-nam-gun,Mt. Du-ryun-san, from temple Dae-heong-sa to sum-mit, JKS1186 (SNUA); Prov. Jeollanam-do: Wan-do-gun, Mt. Sang-hwang-bong, JKS1031 (SNUA); Prov.Jeju-do: Seo-gui-po city, T.B. Lee & J.Y. Park s.n.(SNUA−26934), Seoul: Mt. Buk-han-san, from Gu-gi-dong to Bi-bong, JKS238 (SNUA); Seoul: Mt. Gwan-ak-san, from Shin-rim-dong to The Arboretumthrough Mu-neo-mi-go-gae, JKS980 (SNUA).

Distribution: North-east and north China (Shanxi,Shaanxi, Hebei, Zhejiang, Shandong), central andsouthern Japan (Honshu to Kyushu), and Korea(Fig. 11).

Taxonomic note: Prunus serrulata var. pubescens rep-resents a single widespread and highly polymorphic



taxon that comprises a diverse degree of pubescence inpetioles, peduncles, and pedicels (Wilson, 1916; Kita-mura & Murata, 1979; Ohwi, 1984; Kuitert & Peterse,1999; Ohba, 2001; Li & Bartholomew, 2003). Thisstudy treated P. leveilleana as a synonym of thistaxon. According to Kuitert & Peterse (1999),P. verecunda was interpreted as P. serrulata var.pubescence, an infraspecific taxon of P. serrulata.However, contrary to this treatment, we regardP. verecunda as a variety of P. sargentii due to theumbellate inflorescence (Chang et al., 2004).

This variety is distributed from Japan to Chinathrough Korea. It is known that the centre of distri-bution of this taxon is the Korean peninsula (Kuitert& Peterse, 1999). However, in accordance with the newcombination of P. leveilleana, P. serrulata var. pubes-cens also occurs from northern Honshu to southernChina.

ACKNOWLEDGEMENTS

Our thanks to Dr H. N. Qin at PE, Dr Akiko Shimizuat TI and Henry Kesner at A, and the directors at TI,MAK, and PE for their herbarium loans. Appreciationis also expressed to Drs J. I. Jeon, H. Kim, Byoung-HeeChoi, and Mr Ho Choi for their help. We also thank MsKaren Park for editorial suggestions. This researchwas supported by a grant (no. 052-041-026) from theCore Environmental Technology Development Projectfor the Next Generation funded by the Ministry of theEnvironment of the Korean Government.

REFERENCES

Bentham G, Hooker JD. 1865. Genera Plantarum ad Exem-plaria Implrimis in Herbariis Kewensibus Servata Definita.London: Lovell Reeve & Co.

Bortiri E, Oh SH, Gao FY, Potter D. 2002. The phylogeneticutility of nucleotide sequences of sorbitol 6-phosphatedehydrogenase in Prunus (Rosaceae). American Journal ofBotany 89: 1697–1708.

Bortiri E, Oh SH, Jiang JG, Baggett S, Granger A, WeeksC, Buckingham M, Potter D, Parfitt DE. 2001. Phylogenyand systematics of Prunus (Rosaceae) as determined bysequence analysis of ITS and the chloroplast trnL-trnFspacer DNA. Systematic Botany 26: 797–807.

Chang KS. 2004. A systematic reconsideration of Prunus ser-rulata complex in eastern Asia by morphological analysis.MS thesis, Seoul National University (in Korean).

Chang CS, Choi H, Chang KS. 2004. Reconsideration of Pru-nus sargentii complex in Korea: with respect to P. sargentiiand P. takesimensis. Korean Journal of Plant Taxonomy 34:221–244 (in Korean).

Charkevicz SS. 1996. Plantae Vasculares Orientis ExtremiSovietici, 8. St Petersburg: Nauka (in Russian).

Chung TH. 1943. Illustrated woody flora of Korea. Seoul: Cho-sen Natural History Museum Society (in Japanese).

Hillier J, Coobes A. 2002. The Hillier manual of trees andshrubs. Devon: David & Charles.

Huxley A, Griffiths M, Levy M. 1999. The new Royal Horti-cultural Society dictionary of gardening, Vol. 3. New York:Grove’s Dictionaries.

Ishihidemi H, Ota K, Kigawa S, Sakio H, Nakagawa S,Yosiyama H. 2000. Woody plants of Japan, Vol. 1. Tokyo:Yama-kei Publishers Co. (in Japanese).

Kitamura S, Murata G. 1979. Colored illustrations of woodyplants of Japan, Vol. 2. Osaka: Horikusha Publications (inJapanese).

Koehne E. 1912. Neue Japaniche Arten und Formen vonPrunus subgen. Cerasus. Repertorium Specierum NovarumRegni Vegetabilis 11: 271 (in German).

Koehne E. 1913. Prunus L. In: Sargent CR, ed. Plantae Wil-sonianae, Vol. 2. Portland: Dioscorides Press, 196–282.

Koidzumi G. 1911. Notes on Japanese Rosaceae. BotanicalMagazine of Tokyo 25: 184–188.

Krüssmann G. 1986. Cultivated broad-leaved trees andshrubs, Vol. 3. Portland: Timber Press.

Figure 11. Geographical distribution of Prunus serrulatavar. pubescens in South Korea based on specimens (�). Thedistribution in North Korea (�) is adapted from the previ-ous study of Chung (1943).



Kuitert W, Peterse A. 1999. Japanese flowering cherries.Portland: Timber Press.

Kurata S. 1971. Illustrated important forest trees of Japan,Vol. 3. Tokyo: Chikyû Shuppan (in Japanese).

Lee TB. 1966. Illustrated woody plants of Korea. Seoul: ForestExperiment Station (in Korean).

Lee TB. 1980. Illustrated flora of Korea. Seoul: Hyangmunsa(in Korean).

Lee WT. 1996. Lineamenta Florae Korea. Seoul: AcamedySeojeok (in Korean).

Lee ST, Wen J. 2001. A phylogenetic analysis of Prunus andthe Amygdaloidae (Rosaceae) using ITS sequences of nuclearribosomal DNA. American Journal of Botany 88: 150–160.

Li CL, Bartholomew B. 2003. Genus Cerasus. In: Wu ZY,Raven PH, Hong DY, eds. Flora of China, Vol 9. St. Louis:Missouri Botanical Gardens Press, 404–420.

Lindley J. 1830. Report upon new or rare plants. Transactionsof the Horticultural Society of London 7: 238–239.

Makino T. 1908. Observations on the flora of Japan. BotanicalMagazine of Tokyo 22: 93–102.

Makino T. 1912. Observations on the flora of Japan. BotanicalMagazine of Tokyo 23: 114–122.

Maximowicz CJ. 1883. Diagnoses plantarum novarum asi-aticarum. Bulletin de l’Academie Imperiale des Sciences deSaint-Pétersbourg 29: 74–112.

Nakai T. 1915. Plantae novae Coreanae et Japonicae? Reper-torium Specierum Novarum Regni Vegetabilis 13: 276–277.

Nakai T. 1916. Flora Sylvatica Koreana, Vol. 5. Keijyo(= Seoul): Forest Experiment Station, Government Generalof Chosen, 3–47 (in Japanese).

Nakai T. 1930. Plantae Japonicae and Koreanae. BotanicalMagazine of Tokyo 44: 7–40.

Ohba H. 1992. Japanese cherry tress under the genus Cera-sus. Journal of Japanese Botany 67: 276–281.

Ohba H. 2001. Genus Cerasus. In: Iwatsuki K, Boufford DE,Ohba H, eds. Flora of Japan. Tokyo: Kodasha, 128–144.

Ohwi J. 1984. Flora of Japan. Washington, DC: SmithonianInstitution.

Radford AE, Dickinson WC, Massey JR, Bell CR. 1974.Vascular plant systematics. New York: Harper & Row.

Rehder A. 1908. Einige neue oder kritische Gehölze. Mittei-lungen der Deutschen Dendrologischen Gesellschaft. Berlin17: 1–160.

Rehder A. 1940. Manual of cultivated trees and shrubs, 2ndedn. New York: Macmillan.

Rehder A. 1949. Bibliography of cultivated trees and shrubshardy in the cooler temperate regions of the northern hemi-sphere. Jamaica Plain, MA: Arnold Arboretum of HarvardUniversity.

SAS Institute Inc. 1988. SAS/STAT guide for personal com-puters, version 6. Cary, NC: SAS Institute.

von Siebold PF. 1830. Synopsis Plantarum OeconomicarumUniversi Regni Japonici. Verhandelingen van het Batavia-asch Genootschap van Kunsten en Wetenschapen, Batavia 12:1–75 (in German).

Wilson EH. 1916. The cherries of Japan. Cambridge: Cam-bridge University Press.

Yü TT, Li CL. 1986. Flora Reipublicae Popularis Sinicae, Vol.38. Beijing: Beijing Science Press (in Chinese).

APPENDIX

Origin and accession numbers for specimens used forthe morphological analysis. All voucher specimens aredeposited at The Arboretum, Seoul National Univer-sity (SNUA) or as otherwise indicated.

P. SERRULATA VAR. SERRULATA F. SPONTANEA

FlowersChina: PE 1247852, W.C.Cheng 2281 (PE), PE 362672,PE 84973.

Korea: JKS 080, JKS 081, JKS 083, JKS 083–1, JKS085, JKS 087, JKS 090, JKS 096, JKS 100, JKS 101,JKS 103, JKS 105, JKS 106, JKS 107, JKS 108, JKS109, JKS 110, JKS 111, JKS 112, JKS 119, JKS 123,JKS 124, JKS 133, JKS 138, JKS 233, JKS 214–1, JKS214–2, JKS 214–3, JKS 214–4, JKS 214–5, JKS 217–1, JKS 217–2, JK S217-3, JKS 217–4, JKS 218–1, JKS218–2, JKS 218–3, JKS 222, JKS 239, JKS 251, JKS266–1, JKS 266–2, JKS 266–3, JKS 266–4, JKS 266–5, JKS 273, JKS 252, JKS 266, JKS 270, JK 271, JKS340, JKS 341, JKS 342, JKS 368–2, JKS 369–5, JKS759, JKS 762, JKS 763, JKS 764, JKS 765, JKS 766,JKS 776, JKS 779, JKS 903, JKS 904, JKS 905, JKS906, JKS 913–1, JKS 913–2, JKS 913–3, JKS 1182,JKS 1407, JKS 1413, JKS 1416.

FruitsChina: PE 0852644, K.M. Liou 2023 (PE), PE 818107,PE 1397335, R.C. Ching 4798 (PE), PE 679714, PE912891, PE 926015, PE 915245, PE 920847.

Korea: JKS 377–5, JKS 378–4, JKS 379–1, JKS381–1, JKS 381–2, JKS 382, JKS 388, JKS 391, JKS393, JKS 398, JKS 402, JKS 985, JKS 986, JKS1030,JKS1031, JKS1032, JKS 1182, JKS 1193.

P. SERRULATA VAR. SERRULATA F. SPONTANEA (= AS P. JAMASAKURA)

FlowersJapan: T. Kawahara & Im, H.T. 915 (TI), unknown(April 1909, TI), G. Koidzumi s.n. (April 1909, TI), G.Koidzumi s.n. (April 1911, TI), G. Koidzumi s.n. (1910,TI), T. Sawada (22 April 1927, TI), F. Konta 4935 (TI),N. Fukuoka 12050 (TI), J. Murata 150421 (TI), N.Kurosaki 10636 (MAK), MAK 3273, MAK 105174,MAK 235789, H. Tadota 3651 (MAK), H. Tadota 3743(MAK).

FruitsJapan: T. Naito s.n. (June 1932, TI), S. Yoshiara s.n. (26May 1931, TI), H. Ohashi s.n. (TI 660095), Im, H.T andT. Kawahara 10099 (TI), T. Fusii et al. (TI TWT-14807), S. Noshiro et al. 943 (TI), J. Murata 44519(TI), MAK179163, MAK 225557, MAK 253658, MAK268277, MAK 268278, MAK 63190.



P. SERRULATA VAR. PUBESCENS

FlowersChina: PE 611112.

Korea: JKS 082, JKS 084, JKS 086, JKS 088, JKS089, JKS 091, JKS 092, JKS 093, JKS 094, JKS 095,JKS 097, JKS 098, JKS 099, JKS 101, JKS 102, JKS114, JKS 115, JKS 116, JKS 117, JKS 118, JKS 119,JKS 120, JKS 121, JKS122, JKS 125, JKS 126, JKS127, JKS 128, JKS 129, JKS 130, JKS 131, JKS 132,JKS 134, JKS 135, JKS 136, JKS 137, JKS 175, JKS176, JKS 177, JKS 178, JKS 179, JKS 180, JKS 181,JKS 182, JKS 183, JKS 184, JKS 185, JKS 210, JKS211, JKS 212, JKS 213, JKS 215, JKS 216, JKS219,JKS 220, JKS 221, JKS221-1, JKS 222, JKS 222–1,JKS 223, JKS 224, JKS 225, JKS 226, JKS 227, JKS228, JKS 229, JKS 230, JKS 231, JKS 232, JKS 234,JKS 235, JKS 236, JKS 237, JKS 238, JKS 240, JKS241, JKS 242, JKS 243, JKS 244, JKS 245, JKS 246,JKS 247, JKS 248, JKS 249, JKS 250, JKS 253, JKS254, JKS 255, JKS 267, JKS 268, JKS 269, JKS 272,JKS 346, JKS 347, JKS 368, JKS 368–1, JKS 369–3,JKS 369–4, JKS 772, JKS 773, JKS 774, JKS 775, JKS778, JKS 1406, JKS 1408, JKS 1410.

FruitsChina: PE 638718, PE 255020, PE 318006, PE0596353, PE 625358.

Korea: JKS 346, JKS 347, JKS 368, JKS 368–1, JKS377–1, JKS 377–2, JKS 377–3, JKS 377–4, JKS 377–6, JKS 377–7, JKS 378–1, JKS 378–2, JKS 378–3, JKS378–5, JKS 379–2, JKS 379–3, JKS 379–4, JKS 379–5, JKS 380–1, JKS 380–2, JKS 383, JKS 384, JKS 385,JKS 386, JKS 387, JKS 389, JKS 390, JKS 392, JKS394, JKS 395, JKS 396, JKS 397, JKS 399, JKS 400,JKS 400–1, JKS 400–2, JKS 401, JKS 401–1, JKS 403,JKS 403–1, JKS 403–2, JKS 404, JKS 404–1, JKS404–2, JKS 524, JKS 527, JKS 528, JKS 980, JKS 981,JKS 982, JKS 983, JKS 984, JKS 987, JKS 991, JKS992, JKS 993, JKS 993–1, JKS 1170, JKS 1179, JKS1180, JKS 1181, JKS 1183, JKS 1185, JKS 1186, JKS1187, JKS 1190, JKS 1191, JKS 1192.

P. SERRULATA VAR. PUBESCENS (= AS P. LEVEILLEANA)FlowersJapan: MAK 227238, MAK 227788, MAK 216321, J.Murata & H. Nishimura 638 (TI), J. Murata et al.11026 (TI), J. Murata 17 (22 April 1986, TI), T. Yoshinos.n. (8 May 1979, TI), A. Kimura 5107 (TI), M. Togashi10225 (TI), J. Murata et al. 7215 (TI).

FruitsJapan: T. Nuils s.n. (16 June 1910, TI), T. Nakai s.n. (12June 1913, TI), MAK 227242, MAK 225222, MAK216354, J. Murata 30399 (TI), H. Hara 4282 (TI), M.Takasi 58 (TI).

reconsideration of the prunus serrulata complex (rosaceae) and...

Documents