Plant Breeding Reviews: Volume 19 Edited by Jules Janick© 2000 John Wiley & Sons. ISBN: 978-0-471-38787-9

192 K. YONEMORI, A. SUGIURA, AND M. YAMADA

E. Breeding Objectives Other Than Astringency1. Fruit Cracking2. Fruit Ripening Time3. Fruit Weight4. Soluble Solids Content5. Parthenocarpy6. Sex Expression

F. Future NeedsVII. NEW METHODOLOGIES FOR PERSIMMON BREEDING

A. Ploidy Manipulation through Tissue Culture TechniqueB. Genetic TransformationC. Marker-assisted Breeding

VIII. SUMMARYLITERATURE CITED

I. INTRODUCTION

The oriental persimmon (Japanese persimmon or kaki), Diospyros kaki,is believed to have originated in China and it has been an important foodsource in China, Korea, and Japan from prehistoric times. The name per­simmon was given to the American species, D. virginiana, by the Algon­quin Indians of Virginia. However, today, the name is used genericallyin the United States for both species. In some countries, the oriental per­simmon is called by its species name, kaki.

Kaki fruit is very attractive, ranging in color from yellow, orange, todeep red when mature. Even the leaves turn red in autumn season, mak­ing the species an attractive ornamental. In 1998, the global productionof persimmon totaled 2,071,523 tonnes, 66.4% from China, 14.5% fromJapan, and 12.6% from Korea (FAG, 1999). Persimmon has limited pop­ularity elsewhere. However, Italy, Israel, and Brazil are now producingsubstantial amounts, and were responsible for 6.3% of the total globalproduction in 1998. These countries have developed their own cultivarssuch as 'Kaki Tipo' in Italy, 'Triumph' in Israel, and 'Lama Forte' inBrazil (Sugiura and Subhadrabandhu, 1996). Recently, Australia andNew Zealand have started to produce persimmon mainly for export,and the United States is also producing persimmon on a small scale.

As most cultivars of persimmon were established a long time ago, newcultivars, especially non-astringent cultivars with good fruit quality,have been awaited worldwide. In this review, we will discuss persim­mon genetics and breeding with a special reference to astringency, anddescribe the botanical background of persimmon, the current situation

6. PERSIMMON GENETICS AND BREEDING 193

of persimmon breeding, and new techniques that have potential for per­simmon breeding in the future.

II. BOTANY

A. Scientific Name of Persimmon

The genus Diospyros 1. consists of approximately 400 species, withmost distributed in the tropics of Asia, Africa, and Central-South Amer­ica (de Winter 1963; Wagenitz 1964; Whitmore 1978; Cronquist 1981; Ng1986). Only a few species, including Japanese persimmon, are native tothe temperate zone.

The scientific name of the Japanese persimmon is Diospyros kaki, butthe authority is questionable. Some researchers assign Carolus Linnaeus(1.), whereas others regard Linnaeus's son (L.£.), or Carl Peter Thunberg(Thunb.) as the authority for the binomial. In the Index Kewensis (Jack­son 1893), three "Kaki" are listed: D. Kaki Blanco, D. Kaki L.f., and D.Kaki Thunb., but D. Kaki Blanco is listed as a synonym of D. discolor,and D. Kaki Thunb. as a synonym of D. Kaempferi. On this basis, D. KakiL.f. would seem to be the most appropriate scientific name for Japanesepersimmon. However, Dr. Koidzumi's extensive survey of specimens atThunberg's herbarium in Uppsala University revealed that D. kakiThunb. in the herbarium is the true Japanese persimmon (Koidzumi1925). According to Hiern (1873), Thunberg used the name D. kaki inhis Flora ]aponica published in 1784, and Linnaeus's son used D. kakiin his Supplementum Plantarum published in 1781. However, Thunbergfirst used D. kaki in the Nova Acta Regiae Societatis ScientiarumUpsaliensis, vol. 3, p. 208, issued in 1780 (Koidzumi 1925). Thus, theauthority for persimmon should be Thunberg and D. kaki Thunb. is thecorrect binomial.

B. Persimmon and Its Relatives

Several researchers have reported that the wild-type D. kaki exists in theforests of China (Wilson 1929; Kikuchi 1948; Grubov 1967). The firstdescription of persimmon appeared in China a few centuries before thecurrent era and a document on persimmon culture appeared initially inthe 5th or 6th century in China (Kikuchi 1948).

In addition to D. kaki, D.lotus and D. oleifera (Plate 6.1) also have beencultivated as fruit crop in China. Diospyros lotus has been consumed as

194 K. YONEMORI, A. SUGIURA, AND M. YAMADA

a dried fruit as well as a fresh, over-ripened fruit, whereas D. oleifera wasused mainly for obtaining persimmon oil (tannins). Diospyros rhombi­folia (Plate 6.1), of Chinese origin, also is known as an ornamental; itbears tiny attractive-colored fruit on a dwarf tree. Another importantspecies as a fruit crop is D. virginiana (Plate 6.1), which is native to theeastern United States (Darrow 1975). These species are quite importantas horticultural crops among the Diospyros species of temperate origin.

There are other species native to the tropics and subtropics that pro­duce edible fruit (Table 6.1). For example, D. digyna (syn. D. ebenaster),called black sapote, is cultivated in the Virgin Islands. This speciesis thought to be native to Central America, especially Mexico andGuatemala, and was brought to the Philippines by the Spanish coloniz­ers (Ng 1991). Diospyros discolor (syn. D. blancoi) , indigenous to thePhilippines, also produces edible fruit of a good quality (Coronel 1991).Diospyros decandra (Plate 6.1), which produces fragrant fruit, is dis­tributed in Thailand and Indo-China, and is sold at the local market(Phengklai 1978). Utsunomiya et al. (1998) reported on the distributionof tropical and subtropical Diospyros species and their ethnobotanicalusage in Thailand.

C. Cytogenetics of Persimmon and Its Relatives

The chromosome number of D. kaki is 2n =90 (Namikawa and Higashi1928; Zhuang et al. 1990); that of D. lotus and D. oleifera is 2n = 30(Zhuang et al. 1990). Diospyros virginiana is reported to have two kar­yotypes with 2n = 60 and 90 (Baldwin and Culp 1941). The reportedchromosome numbers of some wild species of Diospyros are 2n = 30 (Ng1978), except for 2n = 60 for D. rhombifolia (Zhuang et al. 1990) and2n = 90 for D. ebenum (Ng 1978). Therefore, the basic chromosomenumber of the genus Diospyros is thought to be 15, and D. kaki is hexa­ploid (2n = 6x = 90). Although Japanese persimmon is hexaploid, it isfairly fertile. Chromosome pairing in metaphase I of meiosis wasobserved mostly in the form of bivalents, and few chromosonal aber­rations such as bridges and laggards also were observed in anaphase I(Zhuang 1990). These results suggest that Japanese persimmon wouldbe allopolyploid.

Recently, Zhuang et al. (1990) reported that nonaploid (2n =9x =135)cultivars exist among D. kaki. The nonaploid cultivars of D. kaki, i.e.'Hiratanenashi' and 'Tonewase', can produce seedless fruit due to theirhigh pathenocarpic ability. The genome sizes of D. kaki and some otherspecies have also been estimated by Tamura et al. (1998) by using flowcytometry. According to their report, D. kaki has a DNA content of

Tab

le6.

1.S

ome

use

ful

Dio

spyr

ossp

ecie

sw

ith

thei

rd

istr

ibu

tio

nan

dch

rom

oso

me

num

ber.

Use

Tre

eis

cult

ivat

edas

com

mer

cial

frui

tcr

op.

Fru

itis

eate

nfr

esh

ordr

ied.

Fru

itis

uti

lize

dto

ob

tain

tann

ins.

Tre

eis

use

das

orna

men

tal.

Fru

itis

edib

le,

and

mad

ein

toca

kes

wit

hbr

an.

edib

le.

Fru

itis

Fru

itw

ith

soft

and

choc

olat

e-co

lore

dis

edib

le,

and

frag

rant

.

Fru

itis

con

sum

edfr

esh

wit

hso

me

oran

geor

lem

on

juic

ead

ded

.

Fru

it,

wit

hd

ark

crea

mfl

esh,

ised

ible

.

Tre

eis

aso

urce

ofw

oo

dof

hig

hq

ual

ity

and

bear

sed

ible

frui

t.

Fru

itis

edib

lean

dtr

eeis

use

das

aro

otst

ock

for

per

sim

mo

n.

Fru

ith

asb

een

reco

rded

tob

eed

ible

.

Chr

omos

ome

Spe

cies

Dis

trib

utio

nn

um

ber

(2n)

Tem

per

ate

D.

kaki

Thu

nb.

Chi

na,

Kor

ea,J

apan

,an

din

tro

du

ced

90,1

35to

man

yco

untr

ies

D.

lotu

sL.

Eas

t,C

entr

al,

and

Wes

tern

Asi

a30

D.

olei

fera

Che

ngC

hin

a30

D.

rhom

bifo

lia

Hem

sl.

Ch

ina

60

D.

virg

inia

na

L.E

aste

rnN

orth

Am

eric

a6

0,9

0

Tro

pic

alan

dsu

btr

op

ical

D.

deca

ndra

Lou

r.S

ou

thC

hina

,V

ietn

am,

Indi

a,30

Mya

nmar

,L

aos,

Th

aila

nd

D.

digy

naJa

cq.

Cen

tral

Am

eric

a,in

tro

du

ced

into

un

kn

ow

nP

hil

ipp

ines

and

Ind

on

esia

D.

disc

olor

Wil

ld.

Ph

ilip

pin

es,

Mic

rone

sia

30

D.

eben

um

Koe

nig

exR

etz.

Indi

a,S

riL

anka

90

D.

glan

dulo

saL

ace

Indi

a,M

yanm

ar,

Lao

s,N

ort

her

n30

Th

aila

nd

D.

mo

nta

na

Rox

b.In

dia,

Mya

nmar

,L

aos,

Th

aila

nd

,30

Cam

bodi

a,V

ietn

am,

Chi

na,

Mal

ayP

enin

sula

,In

done

sia,

,....D

.te

xana

Sch

eele

Mex

ico

30CD (J

1

196 K. YONEMORI, A. SUGIURA, AND M. YAMADA

approximately 5.00-5.21 pg/2C in hexaploid cultivars (2n = 6x= 90) and7.51-8.12 pg/2C in nonaploid cultivars (2n = 9x = 135). The chromo­some numbers of some species are listed in Table 6.1 with their originand use.

D. Phylogenetic Relationships of Persimmon to Its Relatives

There is no evidence as to the progenitor of D. kaki and the place of itsorigin. However, Ng (1978) proposed a hypothesis that D. kaki wasderived from D. roxburghii (syn. D. glandulosa), based on morphologi­cal similarities and geographical distribution. Because the chromosomenumber in D. roxburghii/D. glandulosa is 2n = 30 (Somego 1978),genome duplication must have occurred during the development ofD. kaki if Ng's hypothesis is correct. However, our recent research onthe phylogenetic relationship of D. kaki to D. glandulosa and othertemperate zone species, using special regions (rbcL-ORF106 and trnT­trnF) of cpDNA (Yonemori et al. 1998), indicated that D. kaki is notdirectly linked to D. glandulosa, and that D. kaki seems to share com­mon ancestry with D. lotus and D. virginiana (Fig. 6.1). Diospyrosglandulosa is more closely related to D. oleifera than D. kaki, D. lotusand D. virginiana. The close relationship of D. virginiana to D. kaki isof great interest, since D. virginiana has the same karyotype as D. kaki.Diospyros virginiana may share direct ancestry with D. kaki, althoughthey are geographically isolated at the present time. It is possible thata common ancestor of D. virginiana and D. kaki originated inAsia, migrated to North America, and developed into the modern D.virginiana.

III. HORTICULTURAL CLASSIFICATION OF PERSIMMON

A young, developing persimmon (D. kaki) fruit is highly astringent dueto soluble tannins in the vacuoles of tannin cells. However, some culti­vars lose astringency naturally on the tree as fruits develop, whereas oth­ers retain astringency until maturity. Hence, persimmons are classifiedinto two types, astringent and non-astringent, based on the presence orabsence of astringency in the fruit at harvest. However, each type is fur­ther classified into two sub-types, variant- and constant-type, depend­ing on the relationship between presence of seeds and flesh color. Asfirst described by Burne (1913, 1914), flesh color of variant-type is influ­enced by pollination. The flesh of variant-type becomes dark when it hasseeds as a result of pollination, whereas the flesh color of constant-type

<> •C9~

...-.: C9

~. O~·

<>~

/).~

tS'

."

io~

'l"g/

I1/8

.~~

D.ka

"LiJ

/),,

/

D.8

lJ.

D•

fOp:

fjran

..p

yrrh

oca

rpa.---~

D.sa

ndw

icen

siS

D.di

SCol

orD.

cast

anea

D.

wal

lich

ii

~

t!D

.rh

om

bif

olia

.~~

~6

o~

Q't:::

:f

D.

ma/

abar

ica

var.

siam

ensi

sD

.ar

eo/a

ta

l).v~~.

<>7eg~t

.q~

'i!1

~C'~~

D.

dec

and

ra

D.

con

fert

iflo

ra

I-'

CD "1

Fig.

6.1.

Den

drog

ram

for

five

tem

pera

tezo

neD

iosp

yros

spec

ies

and

19tr

opic

alan

dsu

btro

pica

lzo

ne

Dio

spyr

ossp

ecie

s.A

rm-l

engt

hre

p­re

sent

sth

ege

neti

cdi

stan

ce.

Sou

rce:

Yon

emor

iet

al.

1998

.

198 K. YONEMORI, A. SUGIURA, AND M. YAMADA

is not influenced by the presence of the seeds due to pollination. Thus,we refer to variant-type as "pollination variant" and constant-type as"pollination constant." Horticulturally, so far, persimmons are classifiedinto the following four types: (1) pollination-constant non-astringent(PCNA), (2) pollination-variant non-astringent (PVNA), (3) pollination­variant astringent (PVA), and (4) pollination-constant astringent (PCA)(Table 6.2) (Kajiura 1946).

Among these types, the PCNA and PVNA types lose astringency nat­urally during fruit growth and become edible at maturity. However, ifpollination is insufficient and the fruit does not contain seeds, thePVNA-type fruits do not lose astringency, whereas the PCNA-type fruitloses astringency even when pollination is insufficient and the fruitsets parthenocarpically. In the PVNA type, the loss of astringency of thefruit, therefore, depends on the number of seeds produced in the fruit.The PCNA and PVNA types have different flesh colors. The fleshbecomes dark in the PVNA type when the fruit has seeds and losesastringency, whereas flesh color does not change in the PCNA type evenwhen the fruit has lost its astringency (Plate 6.2).

Both the PVA and PCA types have astringent fruit at maturity and areedible only after the astringency has been removed. However, these twotypes have different flesh colors. In the PVA type, a small portion sur­rounding the seed becomes brown, whereas the flesh color of the PCA­type fruit is not influenced by the presence of seeds (Plate 6.2).

Sugiura et al. (1979) and Sugiura and Tomana (1983) found that pro­duction of ethanol and acetaldehyde by the seeds is associated with theloss of astringency, except in the PCNA-type. The seeds of PVNA-typefruit produce a large amount of ethanol and acetaldehyde during the mid­dle stages of fruit development. These volatile compounds, especiallyacetaldehyde, cause coagulation of tannins in the large tannin cells inthe flesh, which results in the complete loss of astringency. After com­pletion of tannin coagulation, tannin cells become brown by furtheroxidative reaction (Sugiura et al. 1985), which causes a dark color of theflesh in PVNA-type fruit (Plate 6.3). The seeds ofPVA-type fruit also pro­duce these volatile compounds during the fruit development, but in lim­ited amounts, so that the coagulation of tannins is restricted around theseeds and the astringency remains in the rest of the flesh. Dark color ofthe flesh caused by tannin coagulation is also restricted around theseeds in PYA-type fruit (Plate 6.3). The seeds of PCA-type fruit producealmost no ethanol and acetaldehyde during their development, hence theydo not lose astringency naturally on the tree. Because the ability ofseeds to generate these volatile compounds is high in the PVNA-type,low in the PCA-type, and intermediate in the PYA-type fruit, fruits of

Tab

le6

.2.

Hor

ticu

ltur

alcl

assi

fica

tion

of

per

sim

mo

ncu

ltiv

ars

by

astr

inge

ncy

and

fles

hco

lor

of

frui

t.

See

def

fect

PC

(Pol

lina

tion

cons

tant

)

PV

(Pol

lina

tion

vari

ant)

NA

(Non

-ast

ring

ent)

PC

NA

Non

-ast

ring

ent

atm

atu

rity

wh

eth

erse

eded

orno

t.F

lesh

colo

run

affe

cted

by

seed

atm

atur

ity.

Tan

nin

sar

en

ot

coag

ulat

edb

yet

han

ol

trea

tmen

tat

imm

atu

rest

ages

wh

enfr

uit

isst

ill

astr

inge

nt.

(VIG

z)

PV

NA

Non

-ast

ring

ent

atm

atu

rity

on

lyif

seed

ed.

Fle

shtu

rns

bro

wn

atm

atu

rity

ifse

eded

.T

ann

ins

are

coag

ulat

edb

yet

han

ol

trea

tmen

tev

enat

imm

atu

rest

ages

.(V

nG)

A(A

stri

ngen

t)

PC

AA

stri

ng

ent

atm

atu

rity

un

less

trea

ted.

Fle

shco

lor

unaf

fect

edb

yse

edat

mat

urit

y.T

ann

ins

are

coag

ulat

edb

yet

hano

ltr

eatm

ent

even

atim

mat

ure

stag

es.

(VnG

Y)

PY

AA

stri

ngen

tat

mat

uri

tyu

nle

sstr

eate

d.B

row

nfl

esh

colo

ro

nly

aro

un

dse

edat

mat

urit

y.T

ann

ins

are

coag

ulat

edb

yet

hano

ltr

eatm

ent

even

atim

mat

ure

stag

es.

(VnG

)

>-""

(,0

(,0

ZVIG

;V

ola

tile

-in

dep

end

entg

roup

.

YV

DG

;V

ola

tile

-dep

end

ent

grou

p.

200 K. YONEMORI, A. SUGIURA, AND M. YAMADA

the PCA- and PYA-types need to be treated with ethanol or CO2 toremove astringency after harvest.

The seeds of most cultivars in the PCNA-type fruit do not producethese volatile compounds, but some in other cultivars, i.e. 'Fuyu' type,produce a relatively large amount of these compounds. However, PCNA­type fruit loses astringency on the tree without producing seed, indi­cating that ethanol and/or acetaldehyde are not responsible for the lossof astringency in PCNA-type fruit. The mechanism for the loss of astrin­gency in the PCNA-type fruit is different from that of the other threetypes. The tannins ofPCNA-type fruit do not coagulate by ethanol treat­ment at an immature stage on the tree, but those of PCA-, PVNA-, andPYA-type fruits do readily. According to these findings, Sugiura (1984)proposed a new classification of persimmon in which the cultivars aregrouped into the volatile-independent group (VIG) and the volatile­dependent group (VDG), corresponding to the PCNA type and thenon-PCNA types (the PCA-, PVNA-, and PVA-types), respectively(Table 6.2).

The difference between VIG (PCNA-type) and VDG (non-PCNA-type)is qualitative. The chemical characteristics of tannins and develop­mental pattern of tannin cells differ greatly between the VIG and VDGgroups (Sugiura et al. 1979; Yonemori et al. 1983; Yonemori and Mat­sushima 1984,1985, 1987a,b). The remarkable difference in qualitativecharacteristics between PCNA-type and non-PCNA-type fruits is theability to accumulate tannins during fruit growth. PCNA-type fruits stopto accumulate tannins into tannin cells at early stages of fruit growth,while the other three (PVNA, PVA, and PCA) types accumulate tanninsgreatly until the middle stage of fruit development. This difference inthe ability to accumulate tannin is responsible for the trait of natural lossof astringency in PCNA-type fruit. The natural astringency-loss inPCNA-type fruit is mainly caused by dilution of tannins with fruitenlargement (Yonemori and Matsushima 1985). As an aid to under­standing the correlation of these classifications, we summarized it inTable 6.2.

IV. PERSIMMON CULTIVAR

A. Cultivar Development in Asia in Relation to Astringency

Currently more than 950 cultivars of persimmon exist from the sub­tropical to temperate regions of China (Wang et al. 1997), but all of themare of PCA type except for one PCNA cultivar, 'Luo Tian Tian Shi'

6. PERSIMMON GENETICS AND BREEDING 201

(Wang 1982). This PCNA cultivar was found growing in Luo Tian pre­fecture in Hubei province. Although Yamada et al. (1993a) identified thiscultivar as PCNA in type, it is still unknown whether its mechanism forlosing astringency on the tree is the same as that in PCNA-type cultivarsof Japanese origin.

Aside from this exceptional existence of a PCNA-type cultivar inChina, almost all non-astringent type cultivars were developed in Japan.Historical records show that 'Zenjimaru', known to be the oldest PVNA­type cultivar, was found in 1214, and that 'Gosho', assumed to be thefirst PCNA-type cultivar, was first documented in the 17th century(Kikuchi 1948). The most popular PCNA-type cultivars, 'Fuyu' and'Jiro', appeared in the 19th century. According to a nationwide surveyon persimmon cultivars in Japan (Agricultural Research Station 1912),there were only six PCNA-type cultivars in contrast to 401 PVNA-typecultivars among more than 1,000 cultivars collected from all over Japan.This means that in addition to its more recent appearance, the PCNAtype probably has very narrow genetic variability. Forty PCNA cultivars,including bud-sports, which may cover almost all PCNA-type cultivarscurrently existing in Japan, are now preserved at the National Instituteof the Fruit Tree Science (NIFTS) in Akitsu, Hiroshima. Based on fourisozyme systems, Sugiura et al. (1990) classified the PCNA cultivars into17 groups (Table 6.3). Some groups consist of bud-sports of the same ori­gin and some others consist of morphologically quite similar cultivarswith different names. The place of their origin is limited to the centralpart of Japan, in contrast to the PVNA-type cultivars that originated allover the country (Ikeda et al. 1985; Yamada 1993). AFLP analysis hasalso confirmed low genetic diversity among PCNA-type cultivars (Kan­zaki et al. 2000a). The PVNA-type cultivars show a wide range of geneticvariation in fruit shape, fruit ripening time, fruit weight, and solublesolids content in fruit, and in that sense they resemble PCA-type culti­vars (Yamada et al. 1994a).

Cho and Cho (1965) collected and identified 186 Korean cultivarsand reported that all were of the astringent type with wide variations infruit shape. Recently, 'Daean Dangam', a non-astringent cultivar wasreported in Korea by Kim et al. (1988). However, 'Daean Dangam' mayhave been introduced from Japan because its characteristics are identi­cal with 'Mushiroda-gosho', a PCNA-type cultivar of Japanese origin.Although two PCNA-, one PVNA-, and two PVA-type cultivars havebeen reported in Korea among 110 native cultivars by Kim and Ko(1997), the remaining 105 native cultivars are the PCA-type. In Korea,as in China, non-astringent cultivars have not been developed duringpersimmon cultivation.

202 K. YONEMORI, A. SUGIURA, AND M. YAMADA

Table 6.3. Grouping of 40 cultivated clones of PCNA by isozyme patterns of fourenzymes. Source: Sugiura et al. 1990.

Banding pattern of isozyme

Group Cultivar GPF PGMY MDHx GDHw

1 Fuyu, Matsumotowase-Fuyu, Aichiwase- A A A AFuyu, Bud sport of Fuyu, Uenishiwase,Sunami, Benisakigake

2 Jiro, Maekawa-Jiro, Ichikikei-Jiro, A B A BWakasugikei-Jiro, Yaizuwase-Jiro,Obana-Jiro

3 Okugosho, Koudagosho B C B C

4 Fukurogosho, Hazegosho, Zen-no-suke, C D A AMisatogosho

5 Mushirodagosho, Mammoth, Isahaya D E B D

6 Mikado, Fukugosho B H A A

7 Ohgosho, Izushi-ohgosho D B E

8 Gosho, Kaibaragosho, Izushi-kogosho, E F B EGosho (Nara clone)

9 Tenjin-gosho J A E

10 Hanagosho F G B E

11 Yamatogosho D E B E

12 Fujiwaragosho G C A G

13 Yoshimotogosho F K A A

14 Midai G E A H

15 Tokudagosho H C A B

16 Ikutomi D A B G

17 Mizugosho, Gosho (Gifu clone) D L B F

zNine banding patterns (A to 1).

YTwelve banding patterns (A to L).

XTwo banding patterns (A, B).

WEight banding patterns (A to H).

6. PERSIMMON GENETICS AND BREEDING

B. Major Cultivars in the Main Persimmon-producingCountries in the World

203

Persimmon is cultivated mainly in Asian countries such as China, Japan,and Korea, and production from these three countries accounts for morethan 900/0 of global production in 1998 (FAO 1999). In China, there aremany local cultivars, but almost all are the astringent types. Thus, theyare consumed as over-ripened or dried fruit, or as fresh, firm fruit afterartificial treatments for astringency removal (Wang 1987). Several com­mercial cultivars are listed in Table 6.4. Non-astringent cultivars intro­duced from Japan are cultivated on a small scale.

Japan produces several selected cultivars of persimmon for intensivecultivation, although they used to cultivate many local cultivars. Per­simmon cultivars have been selected over time for commercial produc­tion. These selected cultivars in Japan are listed in Table 6.4. Earlyripening cultivars such as 'Nishimura-wase' and 'Tonewase' are culti­vated partly in heated plastic houses for earlier harvest. Persimmon cul­tivation in Korea is also relatively intensive and consists mainly of onecultivar, 'Fuyu', which accounts for almost 70% of total planting areaof persimmon in Korea. However, they also cultivate local cultivars(Table 6.4), as well as non-astringent cultivars introduced from Japan,such as 'Jiro' and 'Nishimura-wase'.

Other countries where persimmon is produced in relatively highamounts are Italy, Brazil, and Israel, and these countries have their owncultivars as well as introduced cultivars such as 'Fuyu' and 'Jiro'. Forexample, in Italy, almost 90% is 'Kaki Tipo' (PVNA type), followed by'Vainiglia' and 'Mercatelli' (both PVNA type) (E. Bellini and E. Giordani,pers. commun.). Israel also has its own cultivar, 'Triumph' (PVA type),which is sold under the name of"Sharon Fruit", and is planted on 95%of the total area devoted to persimmon in Israel (A. Blumenfeld, pers.commun.). Most of the production is consumed domestically as fresh orcold stored fruits, but they also export it to Europe. The export of"Sharon Fruit" to Singapore is increasing. The origin of 'Triumph' isunknown but it was assumed to be introduced from Italy. In Brazil,there were about 800,000 persimmon trees in commercial plantings inthe 1980s according to an estimate by the Institute of Agricultural Econ­omy, mainly of two cultivars of local origin, 'Taubate' (PCA type) and'Rama Forte' (PVA type). In these countries, however, PCNA cultivars

NT

able

6.4.

Per

sim

mo

ncu

ltiv

ars

inC

hina

,Ja

pan,

and

Kor

ea.

0 H::o

Ast

ring

ent

Cou

ntry

Are

aN

ame

of

cult

ivar

type

Cha

ract

eris

tics

Chi

naF

ujia

nA

nguy

oush

iP

CA

Big

frui

t,ju

icy

Gua

ngdo

ngD

ahon

gshi

PC

AJu

icy,

hig

hea

ting

qual

ity

Zhe

jian

gG

aoji

aofa

ngsh

iP

CA

Big

frui

t,di

seas

ere

sist

ant

Sha

anxi

Hoj

inP

CA

Sm

all

frui

t,re

dsk

inco

lor,

hig

hsu

gar

con

ten

tS

han

nx

iJi

xinh

uang

PC

AM

ediu

m-s

ized

frui

t,ju

icy,

hig

hsu

gar

cont

ent,

cold

resi

stan

tS

hanx

iJu

mis

hiP

CA

Sm

all

frui

t,hi

ghsu

gar

con

ten

tH

ubei

Luo

tian

tian

shi

PC

NA

Sm

all

frui

tH

ebei

Mo

pan

shi

PC

AB

igfr

uit,

cold

resi

stan

tH

enan

Qiy

ueza

oP

CA

Ear

lyri

peni

ng,

med

ium

-siz

edfr

uit

Zhe

jian

gT

ongp

ensh

iP

CA

Med

ium

-siz

edfr

uit

Sha

nxi

Xia

oyin

iux

insh

iP

CA

See

dles

s,m

ediu

m-s

ized

frui

t,h

igh

prod

ucti

vity

Jian

gxi

Yu

du

hes

hi

PC

AB

igfr

uit,

juic

y,hi

ghsu

gar

con

ten

t

Japa

nA

izu

mis

hir

azu

PY

AL

ate

ripe

ning

,ea

syto

rem

ove

astr

inge

ncy

by

etha

nol

Ata

goP

CA

Lat

eri

peni

ng,

high

prod

ucti

vity

,lo

ngke

epin

gqu

alit

yF

uy

uP

CN

AB

igfr

uit,

juic

y,la

teri

peni

ngH

achi

yaP

YA

Big

frui

t,m

id-r

ipen

ing,

for

dri

edp

ersi

mm

on

Hir

atan

enas

hiP

YA

Med

ium

-siz

edfr

uit,

juic

y,so

ftfl

esh,

hig

hp

arth

eno

carp

icab

ilit

y,se

edle

ss,

easy

tore

mov

eas

trin

genc

yb

yet

han

ol

Izu

PC

NA

Ear

lyri

peni

ng,

wea

ktr

eevi

gor,

low

Jiro

PC

NA

Lat

eri

peni

ng,

big

frui

tM

aeka

wa-

Jiro

PC

NA

Mid

-rip

enin

g,a

bud-

spor

tof

Jiro

Mat

sum

otow

ase-

PC

NA

Mid

-rip

enin

g,a

bud-

spor

tof

Fu

yu

Fu

yu

Sai

joP

CA

Sm

all

frui

t,ju

icy,

high

suga

rco

nten

t,sh

ortk

eepi

ngqu

alit

yT

onew

ase

PY

AE

arly

ripe

ning

,m

ediu

m-s

ized

frui

t,a

bu

d-s

po

rto

fH

irat

anen

ashi

Kor

eaC

hoho

nshi

PCA

Ear

lyri

peni

ngcu

ltiv

ar,

bear

ing

man

ypi

stil

late

flow

ers

Cho

ndob

ansh

iP

YA

Mid

-rip

enin

gcu

ltiv

ar,

vigo

rous

,ea

syto

init

iate

flow

erin

gb

ud

sD

aean

Dan

gam

PC

NA

Obl

ate

insh

ape,

attr

acti

veco

lor

wit

hli

ght-

yell

owsk

in

6. PERSIMMON GENETICS AND BREEDING 205

are in demand, and 'Fuyu', 'Jiro', and their bud sports for early ripeninghave been introduced for cultivation.

New persimmon growing countries include New Zealand and Aus­tralia. In New Zealand, 80% of the cultivation area is devoted to 'Fuyu'and 15% to 'Matsumotowase-Fuyu' (A. D. Mowat, pers. commun.).'Fuyu' is also the leading cultivar in Australia (R. J. Collins, pers. com­mun.). Thus, the demand for PCNA cultivars is worldwide.

V. IMPORTANCE OF CULTIVAR IDENTIFICATION

Persimmon cultivars are now grown in a number of non-Asian countriesincluding Israel, the European Community, the United States, NewZealand, Australia, Brazil, and Chile. However, there is some confusionas to cultivar identification, partly because persimmon, especially thenon-astringent type cultivars, are new to these countries. We cite anexample as a source of such confusion. There is a persimmon germplasmcollection at the University of California, one at the Wolfskill Experi­mental Orchard at Winters, and another at the South Coast Field Stationnear Los Angeles. The original scions were collected from the ChicoPlant Introduction Station of USDA in California. However, mistakes inthe cultivar name indicated on the label were found by isozyme analy­sis (Parfitt et al. 1991). The most serious mistake was made in 'Fuyu' inthe germplasm collection of the University of California, Davis at theWolfskill Experimental Orchard. There were two trees labeled 'Fuyu' inthe collection, but neither is 'Fuyu'. According to isozyme analysis, oneis 'Jiro' and the other is probably 'Okugosho', because the latter producesstaminate flowers sporadically, whereas 'Fuyu' normally produces onlypistillate flowers. The monoecious, staminate-sporadic 'Okugosho' wasunknowingly renamed 'Cal-Fuyu' by Ryugo (Ryugo et al. 1988). Previ­ously scions of 'Cal-Fuyu' were exported to Italy, New Zealand, andIsrael as 'Fuyu' before the name was corrected by isozyme analysis.This mix-up will cause confusion because 'Cal-Fuyu' is being used as amale parent in breeding programs.

There are many "Fuyus" on the commercial market in the UnitedStates. As 'Jiro' was labeled 'Fuyu' in the Wolfskill ExperimentalOrchard, 'Jiro' is also propagated as 'Fuyu' in California. Nurserymenmade mistakes in the names of their scion wood for propagation. In Cal­ifornia, nearly all PCNA-type are marketed as 'Fuyu' even though theNorthern California Persimmon Association was founded in 1924 tooversee the persimmon industry (Schroeder 1987).

The cultivars introduced to other countries have a similar name

206 K. YONEMORI, A. SUGIURA, AND M. YAMADA

problem that needs to be verified. Even in Japan, there is confusionregarding cultivar nomenclature. Confirmation of the cultivar's truename seems to be an important issue for the persimmon industry andfor persimmon breeding.

VI. PERSIMMON BREEDING

A. Background

The PCNA-type fruit is the most desirable persimmon for fresh con­sumption. The PCNA-type persimmon can be eaten while firm like anapple without any postharvest treatment. The PCNA-type fruit has a sig­nificant advantage, even though the natural loss of astringency occursonly when grown in warm regions. Fruits of PVNA cultivars also loseastringency naturally on the tree, but it occurs only if seeded. Hence, thepalatability is unpredictable and the quality of the fruit is usually notgood.

The PVA and PCA fruits lose astringency when over-ripe, and becomeedible without further treatment. However, the over-ripe fruits are verysoft and cannot be transported easily, so they must be treated with car­bon dioxide ot ethanol to be consumed as a firm fruit (Kitagawa and Glu­cina 1984). In some cultivars, however, the astringency is not easilyremoved by these treatments. Moreover, these treatments cause fruitdamage and shorten the shelflife. Thus, PCNA-type fruit is the most suit­able for the fresh market.

The finding of 'Fuyu' and 'Jiro', the current leading PCNA-typecultivars, fostered the commercial persimmon production greatly inthe early years of the 20th century in Japan. However, both are late­ripening cultivars. Therefore, commercial production of early ripeningPCNA-type cultivars is desired. In addition to early ripening, othertraits such as high eating quality, large fruit, longer keeping qualitywithout fruit cracking, high productivity, and high tolerance to diseasesand pests are critical breeding objectives. 'Fuyu' and 'Jiro' are prone tocrack at the calyx and stylar ends, respectively, as are most other nativePCNA-type cultivars. Early-ripening bud sports of 'Fuyu' ('Matsumo­towase-Fuyu') and 'Jiro' ('Maekawa-Jiro') were found in a farmer'sorchard. Their popularity has increased during the last five decades, but'Matsumotowase-Fuyu' has the same cracking tendency as 'Fuyu'. Thecurrent breeding programs focus on overcoming these defects of PCNAcultivars.

6. PERSIMMON GENETICS AND BREEDING 207

B. Present Situation in Persimmon Breeding

1. Asian Countries. In Japan, a persimmon breeding program that wasstarted in 1938 at the Horticultural Experimental Station (presently theDepartment ofCitriculture of the National Institute of Fruit Tree Science)at Okitsu was transferred to its branch at Akitsu in 1968. The mainbreeding objective since the beginning of the project has been to developand release commercially acceptable, attractive PCNA-type cultivars. Sofar, eight PeNA-type cultivars have been released: 'Suruga' (late ripen­ing, less juicy) in 1959 (Iikubo et al. 1961); 'Izu' (early ripening, shortkeeping quality, low productivity) in 1970 (Hirose et al. 1971); 'Shin­shuu' (early to mid-season ripening, high sugar content, susceptible toskin damage) in 1990 (Yamane et al. 1991b); 'Youhou' (a mid-seasonripening, highly parthenocarpic type with firm flesh) in 1990 (Yamaneet al. 1991a); 'Tanrei' and 'Kinshuu' (bright red leaf color in autumn, forornamental use) in 1993 (Yamane et al. 1998b); 'Taishuu' (a large, mid­season ripening with a juicy, soft flesh) in 1994 (Yamane et al. 1995); and'Yubeni' (late ripening, attractive red fruit color) in 1997 (Yamane et al.1998a).

In Korea, 233 local cultivars were collected at the branch of Experi­mental Station at Kim-hae during 1959-1969, and 74 superior cultivarswere selected for persimmon cultivation after identifying the name of188 cultivars among these local cultivars. In Korea, interest in persim­mon cultivation is increasing and two experimental stations for per­simmon have been established, the one for non-astringent persimnlOnestablished in 1994 and the other for astringent persimmon establishedin 1995. In addition, a breeding program for obtaining new PCNA cul­tivars was started in 1995 by crosses among PCNA cultivars that wereintroduced from Japan. The breeding objectives in Korea are also toobtain superior PCNA cultivars with good eating qualities, large fruit,and early ripening characteristics (S. K. Kang, pers. commun.).

In China, more than 200 domestic cultivars were collected and eval­uated at the Pomology Institute of Shaanxi Academy of Agricultural Sci­ences in Shaanxi province. Wang (1987) recommended some superiordomestic cultivars as well as non-astringent cultivars introducedfrom Japan for commercial production. There is no report of breedingprograms.

2. Non-Asian Countries. There are few countries that have an interestin breeding persimmon among the non-Asian countries. In Israel, therewas a breeding program to produce a non-astringent cultivar with long

208 K. YONEMORI, A. SUGIURA, AND M. YAMADA

storage capability, but it was terminated due to lack of financial sup­port (A. Blumenfeld, pers. commun.). There was a persimmon breed­ing program in Brazil beginning in 1950 at the Agronomic Institute atCampinas, Sao Paulo, to create alternative cultivars and to extend theharvest period (Rigitano et al. 1984). Releases included 'Fuyuhana'(lAC 15207), PCNA type, from the cross between 'Fuyu' and 'Hanagosho'(PCNA cultivar of Japanese origin) and 'Pomelo' (PCA), 'Rubi' (PCA),and 'Kaoru' (PVA), all early-ripening selections with high productivity(Rigitano et al. 1984), but adoption of these selections was minimal.Propagation materials of persimmon were introduced in Italy at the endof the 19th century and many local cultivars were developed. Local per­simmon cultivars, as well as introduced new cultivars from Asian coun­tries, were maintained at the germplasm collection of the HorticultureDepartment in the University of Florence. A breeding program was ini­tiated in 1971 at the Institute for Tree Crops (presently the Departmentof Horticulture) at the University of Florence to obtain new scion cul­tivars and rootstocks (Bellini et al. 1990). The focus was on obtainingPCNA cultivars having large fruit, rounded or slightly flattened shape,good keeping qualities, and suitability for industrial uses (drying).Since 1971, approximately 5,000 seedlings from 90 crosses were pro­duced and several superior selections were made, of which DOFI­86.II.034 (PCNA type) is the most promising in terms of fruit size,attractive shape and appearance, and early ripening characteristics(Bellini and Giordani 1998).

C. Inheritance of Astringency

The trait of natural loss of astringency in PCNA-type is qualitativelyinherited (Ikeda et al. 1985; Yamada 1993). The PCNA-type is recessiveto the other three non-PCNA types (PVNA, PVA, and PCA). PCNA-typeF1 offspring is obtained only by inter-crossing PCNA genotypes; almostno F1 PCNA offspring result from crossing PCNA with PVNA, PVA, orPCA, or from crosses among non-PCNA (PVNA, PVA, and PCA) typesof native cultivars of Japanese origin (Ikeda et al. 1985). Native cultivarsofnon-PCNA types in Japan seem to have few or no genes heterozygousfor the PCNA trait, probably due to the recent origin of PCNA type andno spread of the gene for PCNA trait to non-PCNA types by naturalcrosses. However, in a breeding program, crosses of PCNA cultivars('Fuyu', 'Jiro', or 'Wakasugikei-Jiro') with non-PCNA F i s [derived fromtwo progenies of a PCNA x non-PCNA cross; i.e. 'Seidoshi' (PCA) x'Fukurogosho' (PCNA) and 'Okugosho' (PCNA) x 'Shogatsu' (PVNA)]yielded PCNA genotypes in progenies at 10-22% (14.5% in average)

6. PERSIMMON GENETICS AND BREEDING 209

(Ikeda et al. 1985), indicating that at least two and most likely threerecessives are involved. This low rate for obtaining PCNA genotype isthe main obstacle in persimmon breeding programs. Consequently,crosses usually have been made among PCNA cultivars and selectionsto obtain PCNA offspring.

D. Variability of Genetic Resources for PCNA

PCNA-type cultivars of Japanese origin show little phenotypic variation.Most of them are late ripening (Yamada 1993; Yamada et al. 1994a) andthe fruit tend to crack at the calyx and/or stylar ends (Yamada et al.1988). When Yamada et al. (1993a) analyzed 27 fruit characters of 16PCNA-type and 18 non-PCNA-type cultivars using the principal com­ponent analysis, cultivars of each type are clearly grouped on the firstand second principal component plane. The PCNA-type cultivars havespecific morphological features such as flat fruit shape, depressed calyx,crinkles at the calyx end, in addition to late fruit ripening and fruitcracking. This narrow diversity is attributed to their recent origin.

Since the breeding project in Japan has been focused on PCNA-typecultivars, progenies have a tendency to crack and ripen late (Yamada etal. 1988; Yamada et al. 1995c). Inbred PCNA offspring shows small fruitsize (Yamada et al. 1994b). In the current breeding programs, an efforthas been made to avoid inbreeding in choosing parents from the currentPCNA cultivars and selections.

E. Breeding Objectives other than Astringency

1. Fruit Cracking. The native PCNA-type cultivars are distinctly differ­ent from the other types of cultivars with a habit of cracking at the calyxand/or stylar ends (Table 6.5) (Yamada et al. 1988). Fruit cracking rarelyoccurs at the stylar end in seedless fruit; it can be controlled by pre­venting pollination for cultivars having a high parthenocarpic ability(Yamada et al. 1991). Both cracking habits are independently and quan­titatively inherited. Cultivars that do not crack are homozygous, whereascultivars that crack are heterozygous (Yamada et al. 1988). Offspringderived from crossings among crack-resistant parents exhibit little or nocracking, whereas offspring from crack-susceptible parents exhibit awide range of cracking (Table 6.6). Consequently, crossing among crack­resistant parents is desirable, but this is not easy because most PCNA­type cultivars are crack-susceptible. Many selections showed crackingat the calyx and/or stylar ends, especially in the early years of the breed­ing program (Yamada et al. 1988).

210 K. YONEMORI, A. SUGIURA, AND M. YAMADA

Table 6.5. Frequency distribution of cultivars with fruit crackings at the calyx andstylar ends. Source: Yamada et al. 1988.

Number of cultivars in each cracking score

Astringency Total number Calyx end cracking scoreZ Stylar end cracking scoreZ

type of cultivars 0 1 2 3 4 0 1 2 3 4

Cultivars native to JapanPCNA 21 5 7 4 3 2 6 8 2 5 0PVNA 35 28 7 0 0 0 30 4 1 0 0PYA 15 12 1 1 0 1 15 0 0 0 0PCA 37 34 2 0 0 1 33 4 0 0 0

Cultivars introducedfrom Korea and China

PCA 14 11 3 0 0 0 11 1 2 0 0

Total 122 90 20 5 3 4 95 17 5 5 0

zCracking score: 0 =none, 1 =minute, 2 =slight, 3 =medium, 4 severe.

The degree of cracking fluctuates greatly from year to year, and thehigher the degree of cracking the greater is the yearly fluctuation(Yamada et al. 1986b, 1987b). Yamada and Sato (1997) reported that thecultivar-year interaction is the largest factor causing this fluctuation.Selection for crack resistance requires long-term evaluation and offspringthat crack severely, even one year, should be discarded (Yamada et al.1987b).

Table 6.6. Segregation of offspring for fruit cracking under calyx. Source: Yamada etal. 1988.

Segregation

No. of Cracking scoreZ

CrossY seedlings 0 0-1 1-2 2-3 3-4

High x High 145 85 (59%) 31 (21 %) 14 (10%) 8 (6%) 7 (5%)

High x Low 469 284 (61 %) 107 (23%) 43 (9%) 15 (3%) 20 (4%)

High x None 155 134 (87%) 8 ( 5%) 7 (5%) 4 (3%) 2 (1%)

LowxLow 307 180 (59%) 79 (26%) 25 (8%) 9 (3%) 14 (5%)

Low x None 380 321 (85%) 42 (11%) 11 (3%) 4 (1%) 2 (1%)

None x None 213 191 (90%) 20 (9%) 1 (1%) 1 (1%) 0(0%)

zCracking score: 0 =none, 4 =severe.

YParents classified as: High cracking score over 1, Low =score of 0-1, None =score ofo.

6. PERSIMMON GENETICS AND BREEDING 211

2. Fruit Ripening Time. Most native PCNA cultivars ripen late in the fall,so that early ripening has been one of the most important breedingobjectives. The inheritance of fruit-ripening time is under quantitativecontrol.

The fruit-ripening time has a high broad-sense heritability: 0.84 whenthe evaluation was made with five fruits on a single tree without yearlyrepetition, and 0.92 when evaluated with five fruits on a single tree forthree years (Yamada et al. 1993b). This study was investigated in a pop­ulation of 19 PCNA cultivars/selections used as parents in the 1970s and1980s at the National Institute of Fruit Tree Science (NIFTS) at Akitsu.Therefore, it is easy to evaluate the genetic difference for fruit-ripeningtime.

Yamada et al. (1995c) analyzed breeding records of fruit ripening timeapart from the concept of narrow-sense heritability, and divided the totalvariance in an offspring population of full-sib families into between­cross (between-family) and within-cross (within-family) genetic vari­ances, and within-cross environmental variance. The genetic differencesamong families in progeny were explained solely by mid-parental value(MP) (Table 6.7).

Table 6.7. Estimates of variance components in the offspring population for three fruittraits. Source: Yamada et al. 1994a, 1995a, 1997.

Estimates

Fruit ripening Fruit Soluble solidsVariance component timeZ weightY contene

Between-family variance 0.94 1.70 0.18Regression'" 1.13 1.65 0.07Residual -0.19 (0) 0.06 0.11

Within-family variance 1.96 10.96 2.54Genetic variance 1.68 8.76 1.76Environmental variance 0.28 2.20 0.78

ZFruit ripening time was rated on a scale of 1 to 8. The evaluation was made using 10 fruitson a single tree without yearly repeating.

YFruit weight was measured in gram, and its log-transformed data were subjected to theanalysis. The evaluation was made using five fruits on a single tree without yearly repeating.

xSoluble solids content was determined with a calibrated refractometer for non-astringentfruit. The evaluation was made using five fruits on a single tree without yearly repeating.

"'''Regression'' indicates the variance component of between-family variance, explainedby the regression of the mean value of offspring in a full-sib family (Mf) on mid-parentalvalue (MP) for fruit ripening time and soluble solids content, and by the multiple regres­sion of Mf on MP and inbreeding coefficient for fruit weight.

212 K. YONEMORI, A. SUGIURA, AND M. YAMADA

The coefficient of regression of mean values in a full-sib family onmid-parental value (MP) was 0.99±0.10 and the genetic variance amongfamilies in offspring could be mostly explained by the regression. Thisindicates that the mean value of offspring in a full-sib family (Mf) canbe accurately predicted from the MP, and is nearly the same as the MP.They calculated the expected proportion of offspring having genotypicvalues exceeding a critical value (EP) (Yamada et al. 1995c), and foundthat the observed distributions of offspring were almost the same as thepredicted ones (Yamada and Yamane 1997). In addition, the within­cross genetic variance was fairly small. Therefore, crosses among nativePCNA-type cultivars, most of which are late ripening, did not produceearly ripening offspring. However, after several cycles of selection over50 years, the genotypes gradually shifted toward early ripening in thebreeding program (Fig. 6.2) (Yamada 1993). The early-ripening PCNA­type cultivar 'Izu' was obtained from 'Fuyu' (late ripening) x Okitsu-1(an early to mid-ripening offspring resulting from 'Oku-gosho' selfing).

3. Fruit Weight. Fruit weight is a quantitative character with a highbroad-sense heritability: 0.85 when the evaluation was made with fivefruits on a single tree for three years (Yamada et al. 1993b). Yamada etal. (1994b) analyzed the breeding records of 39 families obtained fromthe breeding program during 1982-1985. They divided the total varianceof offspring into three variance components as with fruit ripening time(Table 6.7). The between-cross genetic variance was mostly explainedby the multiple regression of the mean value of a family on inbreedingcoefficient (F) and MP. The inbreeding coefficient had a greater influ­ence on Mf than MP. Thus, Mf and EP can be precisely predicted fromMP and F (Yamada et al. 1994b; Yamada and Yamane 1997). Theexpected proportion of offspring with large fruit decreased as MPdecreased or F increased. The degree of the change in fruit size was influ­enced more by the change in F than that in MP. If MP was 200 g, theprobability for obtaining offspring of the fruit heavier than 200 g was esti­mated to be 34% when the F value was 0 and 12% when the F value was0.25. Thus, fruit weight is reduced greatly by inbreeding.

Moreover, even if F 0, the Mf was expected to be smaller than MP;for example, Mf was 195 g when MP was 250 g (Yamada et al. 1994b).The F value was calculated as zero for crosses between native cultivars,but their ancestry was unknown. Therefore, even crosses with F = 0might not exclude inbreeding if some native PCNA-type cultivars areclosely related to each other. Combined with the distinctive characteris­tics ofPCNA cultivars derived from their recent ancestors, this result sug­gests that some degree of inbreeding was involved in the formation of the

6. PERSIMMON GENETICS AND BREEDING 213

15Crosses among native PCNA eultivars (Okitsu)

10

5

o

15

Crosses in 1968-1981 (Akitsu)

Crosses among native PCNA

10 and their resulting F1 (Okitsu)

o

5

5

o

10

15

15

10

5

o

Crosses in 1982-1989 (Akitsu)

LateSept.

Early Mid-Oct. Oct.

LateOct.

Early Mid­Nov. Nov.

LateNov.

EarlyDec.

Fruit ripening time

Fig. 6.2. Fruit ripening time of cultivars and selections used as parents at Okitsu(1938-1967) and Akitsu (1968-) in Japan. Source: Yamada 1993.

214 K. YONEMORl, A. SUGIURA, AND M. YAMADA

original parental population of native PCNA cultivars (Yamada et al.1994b). Selection over a number of generations accelerates a reductionin fruit weight. Therefore, the breeding program for obtaining early-ripen­ing selections resulted in the reduction of fruit weight after several gen­erations of inbreeding (Fig. 6.3) (Yamada 1993).

4. Soluble Solids Content. Sweetness or sugar content is an importantfactor of fruit quality. Sweetness is often evaluated as soluble solids con­tent (SSC) using a hand refractometer. Increasing sugar content has beenone of the breeding objectives.

SSC is a quantitative character and fluctuates markedly with envi­ronmental conditions (Yamada et al. 1986a, 1993b, 1994c). A broad­sense heritability of SSC is lower than that of fruit ripening time or fruitweight: 0.47 when the evaluation was made with five fruits on a singletree without yearly repeating, and 0.69 with five fruits on a single treefor three years (Yamada et al. 1993b). The components causing yearlyfluctuation are the year effect, cultivar-year interaction, tree-year inter-

300-325

275-300

.- 250-27500~

i 225-250'0~

.t:: 200-225e~

175-200

150-175

125-150

00 0

-- - -00 00 0

• -- -0 000 00

- - - -0 000 0

• --0 0 0 00

0 000 0

0

Late Sept Early Oct. MId-Oct. Late Oct. Early Nov. MId-Nov. Late Nov. Early Dec.

Fruit ripening time

Fig. 6.3. Fruit ripening time and fruit weight in cross-parents at Okitsu and Akitsu inJapan. Source: Yamada 1993.

• PCNA native cultivars used as cross-parents at Okitsu 1938-1967.

o PCNA cultivars and selections derived from PCNA, used as cross-parents at Akitsu1982-1989.

6. PERSIMMON GENETICS AND BREEDING 215

action, the tree effect, and sampling errors within a tree. The year effect,the largest factor among them, can be estimated by the yearly deviation(YD) of the mean performance in control cultivars. Adjustment of theyear effect (subtracting the YD) can reduce the yearly variability of selec­tions. Yamada et al. (1994c) showed that the yearly variability could bereduced to less than a fifth when more than 10 control cultivars wereused.

After adjustment, Yamada et al. (1997) analyzed the breeding recordson SSC during 1982-1985 similarly to fruit-ripening time. The between­family genetic variance was much smaller than the within-family geneticvariance. The genetic variance for MP was also much smaller than thatfor the offspring (Table 6.7).

Therefore, EP was influenced mainly by the within-family geneticvariance (segregation within a family). This means that choosingparental genotypes has only a small effect. The proportion of offspringwith a sse higher than 18% was estimated as 32%, 50%, and 69% inthe groups having anMP of16.5%, 17.5%, and 18.5%, respectively. Thepredicted distribution of offspring was almost the same as that observed(Yamada et al. 1997).

Although the genetic variation of SSC in native PCNA-type cultivarsis slightly smaller than that in the other types, the mean value of SSC inthe PCNA-type cultivars was fairly high (Yamada et al. 1994a). There­fore, restricted crossing within the PCNA-type cultivars in breeding pro­grams may not provide special obstacles in improving SSC, in contrastwith fruit ripening time, fruit weight, and fruit cracking.

5. Parthenocarpy. Parthenocarpy is an important factor controlling pro­ductivity. Physiological fruit drop is a major problem in persimmon cul­tivation, and high parthenocarpic ability stabilizes fruit production.Early fruit drop (after flowering time to July in Japan) occurs in any cul­tivars, whereas late drop (August to harvest time in Japan) occurs onlyin particular cultivars.

Early fruit drop is related to two factors, parthenocarpy and seed for­mation (Kajiura 1941). The higher the parthenocarpic ability, the less theearly fruit drop. The larger the number of seeds formed, the less the earlyfruit drop even in cultivars or selections having a low parthenocarpicability. There are wide genetic variations in these two characters.

Parthenocarpic ability can be estimated from the percentage of fruitdrop after bagging flowers to prevent pollination. The ability markedlyfluctuates from year to year in Japan, but the yearly fluctuation is causedmainly by year effect, probably by the light intensity during June to July,not by cultivar-year interaction (Yamada et al. 1987a). The year effect can

216 K. YONEMORI, A. SUGIURA, AND M. YAMADA

be estimated from the parthenocarpic ability described above. The abil­ity of seed formation, which can be estimated from the seed number inthe fruit pollinated artificially, also varies widely with the cultivar orselection. However, the yearly fluctuation of the ability is small, and theability of seed formation may be assessed easily in the course of selec­tion (Yamada et al. 19S7a). The inheritance of parthenocarpic ability hasyet to be elucidated.

6. Sex Expression. Japanese persimmons exhibit three types of sexexpressions: (1) only pistillate flowers (pistillate-type), (2) both pistillateand staminate flowers (monoecious-type), and (3) hermaphroditic flow­ers in addition to pistillate and staminate flowers (polygamomonoecious­type). The type of sex expression is determined genetically, although thepistillate-type cultivar produces staminate flowers on very rare occa­sions (Kajiura and Blumenfeld 19S9; Yakushiji et al. 1995).

Yonemori et al. (1990) reported on sex conversion by chemical appli­cation. They demonstrated that staminate flowers were significantlyconverted to pistillate flowers by benzylamino purine (BA) beforeand/or during flower initiation. They sprayed BA at 1,000 ppm 6 timesto monoecious-type cultivars between June 2 and July 6 when flower pri­mordia have not initiated and/or just initiated (Yonemori et al. 1993), andinvestigated flower sex in next spring. Branches sprayed with BA had sig­nificantly higher ratio of pistillate flowers than that of control branches.BA applications in April at the same concentration were also effective toconvert staminate to hermaphroditic flowering in the same year.

The monoecious cultivar is interplanted as a pollinizer for cultivarsshowing low parthenocarpic ability. However, production of staminateflowers is not advantageous because it reduces the number of pistillateflowers and crop size. Staminate flowers tend to be produced more on theless vigorous or old trees. Aged trees of 'Nishimurawase' (monoecious)and 'Taishuu' (polygamomonoecious) tend to have fewer pistillate flow­ers. Sex expression is not a breeding target, but pollen is required forhybridization. Superior genotypes do not always bear staminate flowersand that encumbers breeding programs. A staminate-sporadic peNA cul­tivar with good fruit size and eating quality would be an ideal parent.

The genetic characteristic of producing staminate flowers seems to beinherited quantitatively (Oohata et al. 1964). They reported that the per­centage of offspring bearing staminate flowers in crosses between culti­vars with abundant staminate flowers and those with few staminateflowers was 39.3%, whereas that between cultivars with few staminateflowers and those with few staminate flowers was 25.2% and thatbetween cultivars with no staminate flowers and those with few stami­nate flowers was 16.6%.

6. PERSIMMON GENETICS AND BREEDING 217

F. Future Needs

The number of PCNA parents is limited. Hence, breeders tend to usesuperior cultivars repeatedly as parents, because crosses between culti­vars or selections with a few or no defects are expected to yield a highproportion of desirable offspring. A typical example is the early breed­ing program in NIFTS at Akitsu, in which most of progenies producedduring 1982-1989 were derived from 'Fuyu', 'Jiro', 'Oku-gosho', 'Hana­gosho', 'Fukuro-gosho', and 'Tenjin-gosho' and their bud-sports, andone PVNA-type cultivar ('Nishimura-wase'). This breeding program hasresulted in a narrow genetic base and inbreeding depression.

Parental populations have been improved during the last decade byavoiding inbreeding and incorporating other native PCNA cultivars,although these PCNA cultivars have some seriously defective traits.However, it is difficult to make crosses without inbreeding under thecurrent situation. More crosses between PCNA and non-PCNA typesmay be necessary in the future to broaden the genetic base of the breed­ing population, even if the approach impedes cultivar improvement.PCA cultivars of Japanese origin have as wide variations as those of Chi­nese origin in fruit-ripening time and fruit weight (Yamada et al.1995a,b). Therefore, utilizing non-PCNA (PVNA, PYA, and PCA) culti­vars as one parent should lead to a wide diversity of the parental pop­ulation.

To obtain diverse PCNA offspring, F1 offspring of the PCNA- x non­PCNA-type must be crossed to the PCNA cultivar or selection. Usually,this cross yields PCNA offspring at a low rate. If these F1 offspring areintercrossed or selfed, the probability of obtaining PCNA offspring willbe extremely low in the next generation. It may be impossible to designsuch crosses. However, with a molecular marker for the PCNA-type, wecan select the PCNA-type at an early seedling stage, using leaf DNA.Early screening for PCNA by a molecular marker will make greater effi­ciency for persimmon breeding program.

VII. NEW METHODOLOGIES FOR PERSIMMON BREEDING

A. Ploidy Manipulation through Tissue Culture Technique

Among fruit-tree crops, persimmon is one of the most advanced speciesin tissue culture system. Successful examples of micropropagation byshoot-tip culture and regeneration from callus and protoplast have beenreviewed by Tao and Sugiura (1992a,b) and Tamura et al. (1995a). Thetissue-culture technique seems to be effective for persimmon breeding,and new breeding techniques may be developed through tissue culture.

218 K. YONEMORI, A. SUGIURA, AND M. YAMADA

Tamura et al., who manipulated ploidy using tissue culture techniques,found ways of producing dodecaploid persimmon (2n 12x = 180) bycolchicine treatment to protoplasts from 'Jiro' (Tamura et al. 1996) andintraspecific hybrids between 'Jiro' and 'Suruga' by protoplast fusion(Tamura et al. 1995b).

Other interesting techniques for manipulating polyploidy in persim­mon are endosperm culture (Tao et al. 1997b) and pollination with unre­duced giant pollen (Sugiura et al. 2000). Both methods can theoreticallyproduce nonaploid plants (2n = 9x = 135), but the plants regeneratedfrom endosperm culture were either hexaploid (2n =90) or dodecaploid(2n =12x 180) (Tao et al. 1997b). Nonaploid plants were obtained frompollination with unreduced giant pollen followed by embryo rescue(Sugiura et al. 2000). Both techniques may be useful for producing newpersimmon cultivars.

B. Genetic Transformation

Genetic transformation is now being applied to persimmon usingAgrobacterium tumefaciens. The first trial was the introduction of thecrylA(c) gene of Bacillus thuringiensis into leaf discs, using an Agrobac­terium system. Plantlets regenerated from the calli produced from theleaf discs successfully produced an insecticidal protein and resisteddamage from the Oriental moth (Monema flavescens Walker) in vitro(Tao et al. 1997a). Other trials of genetic transformation in persimmonhave been made to confer enhanced tolerance to environmental stressby introducing the gene encoding NADP-dependent sorbitol-6-phos­phate dehydrogenase of apple cDNA (Gao et al. 1997) and the CodA geneof Arthrobacter globiformis encoding choline oxidase, an enzyme cat­alyzing complete oxidation of choline to glycine betaine (Gao et al.1998). These experiments have just been started, so that efficacies ofthese genes against environmental stress have not been determined. Theoutcome of this research may open a new field of technology for breed­ing new persimmon cultivars.

c. Marker-assisted Breeding

A molecular marker closely linked to a desired trait is a powerful tool forthe selection of useful offspring from a breeding population. If the selec­tion has to be done for a special fruit character, marker-assisted breedingis the most effective method. In persimmon, trials to obtain molecularmarkers that are linked to the trait of fruit astringency have been initiated(Kanzaki et al. 2001). As discussed previously, the inheritance of peNA

6. PERSIMMON GENETICS AND BREEDING 219

type is qualitative and the PCNA type is recessive to non-PCNA (PVNA,PVA, and PCA) types. The trait of astringent type of the fruit seems to becontrolled by two or three allele pairs. In order to be PCNA-type, a geno­type must be recessive in all alleles. By contrast, non-PCNA (PVNA,PVA, and PCA) types have at least one dominant gene for the trait of non­PCNA types. Thus, if we can find dominant markers linked to non-PCNAtrait, PCNA type in the breeding progenies can be easily distinguished bythe absence of the non-PCNA-linked marker bands.

Kanzaki et al. (2001) sought molecular markers linked to the dominantnon-PCNA alleles by bulked segregant analysis (BSA), using an AFLPtechnique. They screened 128 primer combinations in AFLP analysisand found a reliable candidate for selecting PCNA type. This marker wasabsent in all 26 PCNA offspring and was present in 13 of the 25 non­PCNA offspring. This result would indicate that the obtained marker islinked to one dominant allele for the trait of non-PCNA type. Further­more, by using this marker as a probe of Southern blotting analysis ofHindIII digests of genomic DNA after DIG-dUTP-Iabeling, they suc­ceeded in distinguishing PCNA type from non-PCNA types in the off­spring of 42 genotypes in breeding population with 100% accuracy(Kanzaki et al. 2001). Native PCNA-type cultivars of Japanese origincould also be distinguished from non-PCNA-type cultivars by the sameSouthern analysis (Kanzaki et al. 2000b). This marker has to be exam­ined for accuracy in a larger breeding population, while research forseeking molecular markers should be expanded for other importanttraits.

VIII. SUMMARY

The Japanese persimmons, especially PCNA-type cultivars such as'Fuyu' and 'Jiro', are now cultivated in many countries, but new PCNA­type cultivars with better eating quality and earlier ripening time areneeded. In Japan, the breeding programs at NIFTS since 1938 havefocused on this goal and eight PCNA-type cultivars have been released.However, to obtain PCNA-type offspring by hybridization, PCNA-typecultivars or its offspring must be utilized as parents. The limited num­ber of parental PCNA-type cultivars leads to inbreeding which, in turn,causes a reduction in fruit size and the loss of vigor. Nevertheless, theparental germplasm pool has been enlarged to overcome these problemsduring the last decade by avoiding inbreeding in the program at NIFTSat Akitsu. The inheritance of some important traits for breeding is beingelucidated by analyzing breeding population at NIFTS at Akitsu.

220 K YONEMORI, A. SUGIURA, AND M. YAMADA

Conventional breeding methods require a long juvenile period forthe selection of PCNA-type offspring. However, the development oftechniques for selecting peNA type at an early seedling stage couldhelp to make the breeding process more efficient. A molecular markerfor selecting PCNA genotypes is promising. Genetic transformation isanother approach to produce promising new PCNA-type cultivars. Thisapproach seems promising for a persimmon improvement program,because regeneration techniques are very advanced in this species andprogress in transgene technology is proceeding rapidly.

LITERATURE CITED

Agricultural Research Station. 1912. Investigation on persimmon cultivars (in Japanese).Bul. Agr. Res. Sta., (extra), 28:1-46.

Baldwin, J. T., and R. Culp. 1941. Polyploidy in Diospyros virginiana L. Am. J. Bot.28:942-944.

Bellini, K, G. Giannelli, and S. Villa. 1990. Oriental persimmon for the obtaining of new"PCNA" cultivars. Abstr. XXIIIth Int. Hort. Congr., Vol. 2. 3049.

Bellini, K, and K Giordani. 1998. Persimmon. p. 675-684. In: G. T. Scarascia Mugnozzaand M. A. Pagnotta (eds.), Italian contribution to plant genetics and breeding. Univ. Tus­cia, XVth Congress of Eucarpia, Viterbo, Italy.

Cho, S. K, and T. H. Cho. 1965. Studies on the local varieties of persimmon in Korea (inKorean with English summary). Res. Rep. RDA 8:147-190.

Coronel, R. K 1991. Diospyros blancoi A. DC. p. 151-152. In: K W. M. Verheiji andR. K Coronel (eds.), PROSEA: Plant Resources of South-East Asia, No.2. Edible fruitsand nuts. Pudoc Wageningen.

Cronquist, A. 1981. An integrated system of classification of flowering plants. ColumbiaUniv. Press, New York. p. 499-501.

Darrow, G. M., 1975. Minor temperate fruits. p. 270-271. In: J. Janick and J. N. Moore (eds.),Advances in fruit breeding. Purdue Univ. Press, West Lafayette, IN.

de Winter, B. 1963. Ebenaceae. p. 54-99. In: R. A. Dyer, L. K Codd, and H. B. Rycroft (eds.),Flora of southern Africa, Vol. 26. Government Printer, Pretoria, South Africa.

FAO. 1999. FAOSTAT database 1998. Web site at http://apps.fao.org/lim500/nph­wrap.pl?Production.Crop.Primary&Domain=SUA.

Gao, M., R. Tao, A. M. Dandekar, and A. Sugiura. 1997. Sorbitol synthesis in transgenicpersimmon with apple cDNA encoding NADP-dependent sorbitol-6-phosphate dehy­drogenase (in Japanese). J. Japan. Soc. Hort. Sci. 66 (Suppl. 1):130-131.

Gao, M., R. Tao, A. Sugiura, A. Sakamoto, and N. Murata. 1998. Transformation ofJapan­ese persimmon with the CodA gene for choline oxidase (in Japanese). J. Japan. Soc. Hort.Sci. 67 (Suppl. 1):55.

Grubov, V. 1. 1967. Family Ebenaceae. In: B. K Shishkin and KG. Bobrov (eds.), Flora ofthe U.S.S.R. Israel Program for Scientific Translations, Jerusalem 18:349-355.

Hiern, W. P. 1873. A monograph of Ebenaceae. Trans. Cambridge Phil. Soc. 12:227-230.Hirose, K., M. Yamamoto, T. Sato, T. Ohata, T. Nishida, 1. Ikeda, 1. Shimura, S. Siba,

M. Yagi, and N. Tominaga. 1971. New Japanese persimmon variety 'Izu' (in Japanesewith English summary). Bul. Hart. Res. Sta. B11:1-17.

Hume, H. 1913. Effect of pollination on the fruit of Diospyros kaki. Proc. Am. Soc. Hort.Sci. 10:88-93.

6. PERSIMMON GENETICS AND BREEDING 221

Hume, H. H. 1914. A kaki classification. J. Hered. 5:400-406.Iikubo, S., T. Sato, and T. Nishida. 1961. New Japanese persimmon variety 'Suruga' (in

Japanese with English summary). Bul. Hort. Sta., Natl. Tokai-Kinki Agr. Expt. Sta.,6:33-37.

Ikeda, I., M. Yamada, A. Kurihara, and T. Nishida. 1985. Inheritance of astringency inJapanese persimmon (in Japanese with English summary). J. Japan. Soc. Hort. Sci.54:39-45.

Jackson, B. D. 1893. Index Kewensis. An enumeration of the genera and species of flow­ering plants. Vol. 1. Clarendon Press, Oxford. p. 767.

Kajiura, I., and A. Blumenfeld. 1989. Diospyros kaki. p. 298-306. In: A. H. Halevy (ed.),CRC handbook of flowering. Vol. VI. CRC Press, Boca Raton, FL.

Kajiura, M. 1941. Studies on physiological fruit drop in persimmon. II. Relationshipsbetween fruit drop and pollination or parthenocarpic ability (in Japanese). J. Japan. Soc.Hort. Sci. 12:247-283.

Kajiura, M. 1946. Persimmon cultivars and their improvement 2. (in Japanese). Breed. Hort.1:175-182.

Kanzaki, S., K. Yonemori, A. Sato, M. Yamada, and A. Sugiura. 2000a. Analysis of thegenetic relationships among pollination-constant and non-astringent (PCNA) cultivarsof persimmon (Diospyros kaki Thunb.) from Japan and China using amplified fragmentlength polymorphism (AFLP). J. Japan. Soc. Hort. Sci. 69: in press.

Kanzaki, S., K. Yonemori, A. Sato, M. Yamada, and A. Sugiura. 2000b. Effectiveness ofRFLP analysis for discriminating PCNA genotype in some persimmon cultivars. J.Japan. Soc. Hort. Sci. 69: in press.

Kanzaki, S., K. Yonemori, A. Sugiura, A. Sato, and M. Yamada. 2001. Identification of mol­ecular markers linked to the trait of natural astingency-loss of Japanese persimmon(Diospyros kaki Thunb.) fruit. J. Amer. Soc. Hort. Sci. 126: in press.

Kikuchi, A. 1948. Pomology-Part I. (in Japanese). Yokendo, Tokyo, Japan, p. 347-400.Kim, T. C., and K. C. Ko. 1997. Classification of persimmon (Diospyros kakiThunb.) cul­

tivars on the base of horticultural traits. Acta Hort. 436:77-83.Kim, Y. S., S. B. Jeong, D. S. Son, K. K. Lee, and U. J. Lee. 1988. A new non-astringent per­

simmon cultivar "Daean Dangam" (in Korean with English summary). Res. Rept. RDA(H) 30:79-82.

Kitagawa, H., and P. G. Glucina. 1984. Persimmon culture in New Zealand. p. 58-60. Sci­ence Information Publishing Centre, DSIR, Wellington, New Zealand.

Koidzumi, G. 1925. Plantae Japonicae Thunbergianae. p. 114. Faculty of Science, KyotoUniv., Japan.

Namikawa, I., and M. Higashi. 1928. On the number of chromosomes in Diospyros kakiL.f. and Diospyros lotus L. Bot. Mag. 42:436-438.

Ng, F. S. P. 1978. Diospyros roxburghii and the origin of Diospyros kaki. Malaysian Forester41:43-50.

Ng, F. S. P. 1986. Ebenaceae. p. 56-94. In: F. S. P. Ng (ed.), Tree flora of Malaya, Vol. 3.Forest Res. Inst., Kepong, Malaysia.

Ng, F. S. P. 1991. Diospyros digyna Jacq. p. 152-154. In: K W. M. Verheiji and R. K Coro­nel (eds.), PROSEA: Plant resources of South-East Asia, 2. Edible fruits and nuts. PudocWageningen.

Oohata, T., I. Ikeda, and T. Nishida. 1964. The inheritance of some characters in Japanesepersimmon (Diospyros kaki). II. The inheritance of bearing habit of staminate flowers(in Japanese with English summary). Bul. Hort. Res. Sta. B3:51-66.

Parfitt, D. K, K. Yonemori, K. Ryugo, and A. Sugiura. 1991. Isozyme identification ofJapan­ese persimmons (Diospyros kaki L.): Comparisons of cultivars in California and Japan.Fruit Var. J. 45:107-113.

222 K. YONEMORI, A. SUGIURA, AND M. YAMADA

Phengklai, C. 1978. Ebenaceae of Thailand. Thai Forest Bul. 11:1-103.Rigitano, 0., M. Ojima, F. A. C. Dall'Orto, A. F. C. Tombolato, W. Barbosa, H. J. Scarani,

and F. P. Martins. 1984. "Fuyuhana": A new non-astringent persimmon cultivar for SaoPaulo (in Spanish). VIIth Brazilian Fruitculture Congr. p. 288-294.

Ryugo, K., C. A. Schroeder, A. Sugiura, and K. Yonemori. 1988. Persimmons for Califar­nia. California Agr. 42 (4):7-9.

Schroeder, C. A. 1987. Some notes on non-astringent persimmon cultivars. J. CaliforniaRare Fruit Growers 19:1-5.

Somego, M. 1978. Chromosome number of Diospyros roxburghii. Malay. Forester 41:51-52.Sugiura, A. 1984. The origin of persimmon and its cultivar differentiation (in Japanese).

p. 30-37. In: T. Ohmura, Y. Yomogihara, H. Ito, T. Kinoshita, andJ. Kadota (eds.), Recentadvances in plant breeding. Vol. 25. Keigaku Pub, Tokyo, Japan.

Sugiura, A., T. Ohkuma, Y. A. Choi, R Tao, and M. Tamura. 2000. Production ofnonaploid(2n = 9x) Japanese persimmons (Diospyros kaki) by pollination with unreduced (2n =6x) pollen and embryo rescue culture. J. Amer. Soc. Hart. Sci. 125: in press.

Sugiura, A., and S. Subhadrabandhu. 1996. Overview of persimmon culture. ChronicaHart. 36 (3):14-15.

Sugiura, A., S. Taira, K. Ryugo, and T. Tomana. 1985. Effect of ethanol treatment on fleshdarkening and polyphenoloxidase activity in Japanese persimmon, cv. Hiratanenashi.J. Japan. Soc. Food Sci. Technol. 32:586-589.

Sugiura, A., and T. Tomana. 1983. Relationships of ethanol production by seeds of dif­ferent types ofJapanese persimmons and their tannin content. HortScience 18:319-321.

Sugiura, A., K. Yonemori, H. Harada, and T. Tomama. 1979. Changes of ethanol and acetalde­hyde contents in Japanese persimmon fruits and their relation to natural deastringency(in Japanese with English summary). Studies from Inst. Hort. Kyoto Univ. 9:41-47.

Sugiura, A., K. Yonemori, T. Tetsumura, R Tao, M. Yamada, and H. Yamane. 1990. Iden­tification of pollination-constant and non-astringent type cultivars of Japanese per­simmon by leaf isozyme analysis (in Japanese). J. Japan. Soc. Hart. Sci. 59 (Suppl.1):44-45.

Tamura, M., R Tao, and A. Sugiura. 1995a. Regeneration of plants from protoplasts ofDiospyros kaki 1. (Japanese persimmon). p. 43-54. In: Y. P. S. Bajaj (ed.), Biotechnol­ogy in agriculture and forestry, vol. 34. Plant Protoplasts and Genetic Engineering 6.Springer-Verlag, New York.

Tamura, M., R Tao, and A. Sugiura. 1995b. Regeneration of somatic hybrids from elec­trofused protoplasts ofJapanese persimmon (Diospyros kaki 1.). Plant Sci. 108:101-107.

Tamura, M., R Tao, and A. Sugiura. 1996. Production of dodecaploid plants of Japanesepersimmon (Diospyros kaki 1.) by colchicine treatment of protoplasts. Plant Cell Rep.15:470-473.

Tamura, M:, R Tao, K. Yonemari, N. Utsunomiya, and A. Sugiura. 1998. Ploidy level andgenome size of several Diospyros species. J. Japan. Soc. Hart. Sci. 67:306-312.

Tao, R, A. M. Dandekar, S. 1. Uratsu, P. V. Vail, and J. S. Tebbets. 1997a. Engineeringgenetic resistance against insects in Japanese persimmon using crylA(cJ gene of Bacil­lus thuringiensis. J. Am. Soc. Hart. Sci. 122:764-771.

Tao, R, K. Ozawa, M. Tamura, and A. Sugiura. 1997b. Dodecaploid plant regenerationfrom endosperm culture of persimmon (Diospyros kaki 1.). Acta Hart. 436:119-128.

Tao, R, and A. Sugiura. 1992a. Micropropagation of Japanese persimmon (Diospyros kaki1.). p. 426-440. In: Y. P. S. Bajaj (ed.), Biotechnology in agriculture and forestry, vol.18. High-Tech and Micropropagation 2. Springer-Verlag, New Yark.

Tao, R, and A. Sugiura. 1992b. Adventitious bud formation from callus cultures ofJapan­ese persimmon. HortScience 27:259-261.

6. PERSIMMON GENETICS AND BREEDING 223

Utsunomiva, N., S. Subhadrabandhu, K. Yonemori, M. Oshida, S. Kanzaki, F. Nakatsubo,and A. ~Sugiura. 1998. Diospyros species in Thailand: Their distribution, fruit mor­phology, and uses. Econ. Bot. 52:343-351.

Wagenitz, G. 1964. Reihe Ebenales (in German). p. 396-403. In: A. (ed.), Syllabusder Pflanzenfamilien. Gebruder Borntraeger, Berlin.

Wang, R 1982. The origin of'Luo Tian Tian Shi' (in Chinese). Chinese Fruit Tree 2:16-19.Wang, R 1987. Persimmon (in Chinese). p. 65. China Prospect Publisher, Beijing.Wang, R, Y. Yang, and G. Li. 1997. Chinese persimmon germplasm resources. Acta Hort.

436:43-50.Whitmore, T. C. 1978. Ebenales. p. 132-134. In: V. H. Heywood (ed.), Flowering plants of

the world. Oxford Univ. Press, London.Wilson, E. H. 1929. China-mother of gardens. Stradford Company, Boston, MA. p. 357.Yakushiji, H., M. Yamada, K. Yonemori, A. Sato, and N. Kimura. 1995. Staminate flower

production on shoots of'Fuyu' and 'Jiro' persimmon. J. Japan. Soc. Hort. Sci. 64:41-46.Yamada, M. 1993. Persimmon breeding in Japan. Japan. Agr. Res. Quart. 27 (1):33-37.Yamada, M., 1. Ikeda, H. Yamane, and T. Hirabayashi. 1988. Inheritance offruit cracking

at the calyx end and stylar end in Japanese persimmon (Diospyros kaki Thunb.) (inJapanese with English summary). J. Japan. Soc. Hort. Sci. 57:8-16.

Yamada, M., M. Kakutani, H. Yamane, and K. Yoshinaga. 1991. Influence of seed forma­tion on the cracking at fruit apex in Japanese persimmon (Diospyros kaki Thunb.) (inJapanese with English summary). Bul. Fruit Tree Res. Sta. 20:1-11.

Yamada, M., A. Kurihara, and T. Sumi. 1987a. Varietal differences in fruit bearing in Japan­ese persimmon (Diospyros kaki Thunb.) and their yearly fluctuations (in Japanese withEnglish summary). J. Japan. Soc. Hort. Sci. 56:293-299.

Yamada, M., and A. Sato. 1997. Varietal differences and environmental variations of calyxseparation in genotypes of pollination constant and non-astringent Japanese persimmon(in Japanese). J. Japan. Soc. Hart. Sci. 66 (Suppl. 1):132-133.

Yamada, M., A. Sato, H. Yakushiji, K. Yoshinaga, H. Yamane, and M. Endo. 1993a. Char­acteristics of 'Luo Tian Tian Shi', a non-astringent cultivar of oriental persim­mon (Diospyros kaki Thunb.) of Chinese origin in relation to non-astringent cultivarsof Japanese origin (in Japanese with English summary). Bul. Fruit Tree Res. Sta.25:19-32.

Yamada, M., R Wang, H. Yamane, A. Sato, and N. Hirakawa. 1995a. Variation in the per­formance of fruit maturing time, fruit weight, and soluble solids content in oriental per­simmon grown at Akitsu, Japan and Meixian, China. J. Japan. Soc. Hort. Sci. 64:221-226.

Yamada, M., R Wang, H. Yamane, A. Sato, and N. Hirakawa. 1995b. Comparisons in thevariations in fruit maturing time, fruit weight, and soluble solids content of oriental per­simmon cultivars of Chinese and Japanese origin. J. Japan. Soc. Hort. Sci. 64:227-233.

Yamada, M., and H. Yamane. 1997. Relationship between the observed and predicted dis­tribution of offspring for fruit ripening time and fruit weight in Japanese persimmon.Scientia Hort. 69:157-167.

Yamada, M., H. Yamane, and T. Hirabayashi. 1986a. Studies on crossbreeding of Japan­ese persimmon (Diospyros kaki Thunb.). IV. A sampling method in determining refrac­tometer index (in Japanese with English summary). Bul. Fruit Tree Res. Sta. E6:11-20.

Yamada, M., H. Yamane, and T. Hirabayashi. 1986b. Studies on crossbreeding of Japan­ese persimmon (Diospyros kaki Thunb.). V. Variation of 'Hetasuki' (Fruit crackingunder the calyx) (in Japanese with English summary). Bul. Fruit Tree Res. Stn. E6:21-30.

Yamada, M., H. Yamane, and T. Hirabayashi. 1987b. Yearly fluctuations of two types offruit cracking in seedling populations of Japanese persimmon (Diospyros kaki Thunb.)J. Japan. Soc. Hort. Sci. 56:287-292.

224 K YONEMORI, A. SUGIURA, AND M. YAMADA

Yamada, M., H. Yamane, A. Sato, N. Hirakawa, and R. Wang. 1994a. Variations in fruitripening time, fruit weight and soluble solids content of oriental persimmon cultivarsnative to Japan. J. Japan. Soc. Hort. Sci. 63:485-491.

Yamada, M., H. Yamane, Y. Takano, and Y. Ukai. 1997. Estimation of the proportion ofoffspring having soluble solids content in fruit exceeding a given critical value. Euphyt­ica 93:119-126.

Yamada, M., H. Yamane, and Y. Ukai. 1994b. Genetic analysis ofJapanese persimmon fruitweight. J. Am. Soc. Hort. Sci. 119:1298-1302.

Yamada, M., H. Yamane, and Y. Ukai. 1994c. Efficiency of use of control genotypes ofreducing yearly fluctuations of quantitative fruit characters in Japanese persimmon. Bul.Fruit Res. Sta. 26:29-37.

Yamada, M., H. Yamane, and Y. Ukai. 1995c. Genetic analysis of fruit ripening time inJapanese persimmon. J. Am. Soc. Hort. Sci. 120:886-890.

Yamada, M., H. Yamane, K Yoshinaga, and Y. Ukai. 1993b. Optimal spatial and temporalmeasurement repetition for selection in Japanese persimmon. HortScience 28:838-841.

Yamane, H., A. Kurihara, K Nagata, M. Yamada, T. Kishi, K Yoshinaga, R. Matsumoto,K Kanato, T. Sumi, T. Hirabayashi, T. Ozawa, K. Hirose, M. Yamamoto, and M. Kaku­tani. 1991a. New Japanese persimmon cultivar 'Youhou' (in Japanese with Englishsummary). Bul. Fruit Tree Res. Sta. 20:49-61.

Yamane, H., A. Kurihara, K Nagata, M. Yamada, T. Kishi, K Yoshinaga, R. Matsumoto,T. Ozawa, T. Sumi, T. Hirabayashi, and M. Kakutani. 1991b. New Japanese persimmoncultivar 'Shinsyuu' (in Japanese with English summary). Bul. Fruit Tree Res. Sta.19:13-27.

Yamane, H., A. Kurihara, M. Yamada, K Nagata, T. Kishi, K. Yoshinaga, N. Hirakawa,A. Sato, R. Matsumoto, H. Iwanami, T. Ozawa, T. Sumi, T. Hirabayashi, M. Kakutani,K. Kanato, and 1. Nakajima. 1998a. 'Yubeni', a new cultivar of Japanese persimmon.J. Japan. Soc. Hort. Sci. 67 (Suppl. 1):51.

Yamane, H., M. Yamada, A. Kurihara, K Yoshinaga, N. Hirakawa, A. Sato, R. Matsumoto,and H. Iwanami. 1995. New Japanese persimmon cultivar 'Taishuu' (in Japanese).J. Japan. Soc. Hort. Sci. 64 (Suppl. 2):184-185.

Yamane, H., M. Yamada, A. Kurihara, K. Yoshinaga, K Nagata, T. Ozawa, T. Sumi,T. Hirabayashi, N. Hirakawa, A. Sato, R. Matsumoto, and M. Kakutani. 1998b. 'Tanrei'and 'Kinshuu', new Japanese persimmon cultivars (in Japanese with English summary).Bul. Fruit Tree Res. Sta. 30.31:15-24.

Yonemori, K., S. Kanzaki, D. E. Parfitt, N. Utsunomiya, S. Subhadrabandhu, and A. Sug­iura. 1998. Phylogenetic relationship of Diospyros kaki (persimmon) to Diospyros spp.(Ebenaceae) of Thailand and four temperate zone Diospyros spp. from an analysis ofRFLP variation in amplified cpDNA. Genome 41:173-182.

Yonemori, K., and J. Matsushima. 1984. Chemical characteristics of tannins from non­astringent and astringent type fruits of Japanese persimmon (Diospyros kaki) with par­ticular reference to ultracentrifugal behavior (in Japanese with English summary). J.Japan. Soc. Hort. Sci. 53:121-126.

Yonemori, K, andJ. Matsushima. 1985. Property of development ofthe tannin cells in non­astringent type fruits of Japanese persimmon (Diospyros kakj) and its relationship to nat­ural deastringency (in Japanese with English summary). J. Japan. Soc. Hort. Sci.54:201-208.

Yonemori, K., and J. Matsushima. 1987a. Morphological characteristics of tannin cells inJapanese persimmon fruit. J. Am. Soc. Hort. Sci. 112:812-817.

Yonemori, K., and J. Matsushima. 1987b. Changes in tannin cell morphology with growthand development ofJapanese persimmon fruit. J. Am. Soc. Hort. Sci. 112:818-821.

6. PERSIMMON GENETICS AND BREEDING 225

Yonemori, K., J. Matsushima, and A. Sugiura. 1983. Differences in tannin of non-astrin­gent and astringent type fruits of Japanese Persimmon (Diospyros kaki Thunb.) (inJapanese with English summary). J. Japan. Soc. Hort. Sci. 52:135-144.

Yonemori, K., A. Sugiura, K. Tanaka, and K. Kamada. 1993. Floral ontogeny and sexdetermination in monoecious-type persimmon. J. Am. Soc. Hort. Sci. 118:293-297.

Yonemori, K., Y. Yomo, and S. Sugiura. 1990. Sexuality in Japanese persimmon. 2. Induc­tion of sex conversion in male flower by cytokinin treatment (in Japanese). J. Japan. Soc.Hort. Sci. 59 (Suppl. 2):230-231.

Zhuang, D. 1990. Cytogenetic studies on Japanese persimmon cultivars. On the chromo­some number of seedless cultivars (in Japanese with English summary). Ph.D. thesis,Kyoto Pref. Univ., Kyoto, Japan.

Zhuang, D., A. Kitajima, M. Ishida, and Y. Sobajima. 1990. Chromosome number of Diospy­ros kaki cultivars (in Japanese with English summary). J. Japan. Soc. Hort. Sci.59:289-297.