Interspecific hybridisation between Hibiscus syriacus,Hibiscus sinosyriacus and Hibiscus paramutabilis

Katrijn Van Laere ÆJohan M. Van Huylenbroeck ÆErik Van Bockstaele

Received: 16 June 2006 / Accepted: 29 November 2006 / Published online: 6 January 2007� Springer Science+Business Media B.V. 2007

Abstract Interspecific hybrids from crosses

between H. syriacus · H. paramutabilis and

H. syriacus · H. sinosyriacus were obtained. In both

cases unilateral incongruity was observed and

reciprocal crosses yielded no fruits. In vitro embryo

rescue, 11 weeks after pollination, increased the

germination capacity of H. syriacus · H. sinosyria-

cus embryos, while this was not the case for

H. syriacus · H. paramutabilis embryos. However,

a lot of the generated H. syriacus · H. sinosyriacus

seedlings were lost due to variegated and total

albinism. In the progeny of H. syriacus ‘Oiseau

Bleu’ · H. paramutabilis about 95% of the seedlings

had an intermediate flower and leaf morphology

compared to both parent plants. Leaves on the adult

F1 hybrids showed a polymorphism. In total four

different leaf types could be observed on the same

plant. This leaf polymorphism also was seen in the

progeny of H. syriacus ‘Melwhite’ · H. sinosyriacus

‘Lilac Queen’. In this progeny about 50% of the

seedlings had an intermediate flower and leaf

morphology compared with the parent plants. The

hybrid nature of the seedlings of both progenies was

also confirmed by AFLP analysis. Despite the low

germination vigour of the pollen of the hybrids, a

small F2 generation was obtained from H. syriacus

‘Oiseau Bleu’ · H. paramutabilis.

Keywords Embryo rescue � Hibiscus syriacus �Hibiscus sinosyriacus � Hibiscus paramutabilis �Interspecific hybrids � Leaf polymorphism

Introduction

Hibiscus belongs to the Malvaceae. It is a poly-

morphic genus of some 250 species of trees,

shrubs and herbs. Geographically Hibiscus is

mainly situated in tropical and subtropical areas,

with some species extending into the temperate

regions of the world i.e. Hibiscus syriacus L.

(althea or rose of Sharon), Hibiscus sinosyriacus

Bailey and Hibiscus paramutabilis Bailey (Bates

1965). These three species are native to China

(Bates 1965) and the similarity in their natural

distribution pattern is an indication of a similar

tolerance to environmental factors. Hibiscus

syriacus is the most popular species. About 40

different cultivars, with varying flower colour and

shape, are commonly in culture in Europe and a

lot more genotypes are present in different

K. Van Laere (&) � J. M. Van Huylenbroeck �E. Van BockstaeleApplied Genetics and Breeding, Institute forAgricultural and Fisheries Research (ILVO) – PlantUnit, Caritasstraat 21, 9090 Melle, Belgiume-mail: katrijn.vanlaere@ilvo.vlaanderen.be

E. Van BockstaeleDepartment of Plant Production, Faculty ofBiosciences Engineering, Ghent University, CoupureLinks 653, 9000 Gent, Belgium

123

Euphytica (2007) 155:271–283

DOI 10.1007/s10681-006-9328-8

collections (Van De Laar 1997). It is a deciduous

shrub with more or less distinctly trilobate leaves.

Flowering is from the end of July till the first part

of October (Kim and Lee 1991). Basic chromo-

some number of Hibiscus syriacus is x = 20 and

most cultivars are tetraploid, 2n = 4x = 80 (Skov-

sted 1941). However, in literature frequently the

term diploid is used instead of tetraploid. Breeding

work resulted also in hexaploid (so called triploid)

and octaploid (so called tetraploid) cultivars,

which in general have larger flowers and a

longer flowering period (Egolf 1971, 1981; Van

Huylenbroeck et al. 2000). Hibiscus sinosyriacus,

2n = 4x = 80 (Skovsted 1941), has broader

leaves compared to Hibiscus syriacus. The leaves

have short triangular lobes and the involucratal

bracts outside the calyx are as long as the calyx or

even longer (Bates 1965). Hibiscus paramutabilis

has big round-shaped leaves and the flowers are

larger compared to H. syriacus (Bates 1965). The

chromosome number is 2n = 82 (Niimoto 1966).

Both, Hibiscus sinosyriacus and Hibiscus para-

mutabilis, are more vigorous in growth compared

to Hibiscus syriacus. The genetic relationships

among these three species and some morpho-

logical characteristics were published before

by Van Huylenbroeck et al. (2000). Also close

relationships with a few other Chinese species,

Hibiscus mutabilis and Hibiscus indicus, are

suggested (Bates 1965).

In the breeding of ornamentals, the role of

interspecific hybridisation in creating more vari-

ation is very important. When using interspecific

hybrids in backcrosses and in new cross combi-

nations hybrid vigour can be captured and traits

that do not occur within a single species can be

combined (Van Tuyl and De Jeu 1997). Inter-

specific and intergeneric hybrids can extend not

only the qualitative traits but also the quantitative

traits of the parent species. The transfer of alien

genes by sexual means requires both the produc-

tion of F1 hybrids and the subsequent fertility of

these hybrids. The difficulty of creating inter-

specific hybrids increases along with the phylo-

genetic distance between the parents (Sharma

1995). Pre- and postfertilisation barriers prevent

the production of F1 hybrid plants (Hogenboom

1973). Nevertheless, the introduction of specific

techniques such as use of mentor pollen and

grafting of the style has opened opportunities to

overcome prefertilisation barriers. Other in vitro

techniques in the form of embryo rescue and

ploidy manipulation can be used to deal with

postfertilisation barriers. All these techniques

have greatly facilitated interspecific hybridisation

in a wide variety of plant species (Eeckhaut et al.

2006).

A lot of breeding work is done in Hibiscus

syriacus. Interspecific hybridisation between

H. syriacus, H. sinosyriacus and H. paramutabilis

was reported (Kyung and Kim 2001a, b; Kyung

et al. 2001a, b). But the best hybridisation

strategy and the inheritance of morphological

traits remains unclear. Embryo rescue in intra-

specific Hibiscus syriacus crosses has already

successfully been performed (Kim et al. 1996a,

b; Paek et al. 1989a, b). Paek et al. (1989c)

reported the establishment of somatic hybridi-

sation of Hibiscus syriacus protoplasts with other

Hibiscus protoplasts, but the regeneration of

the hybridised protoplasts remains impossible.

Also some interspecific breeding work involving

more tropical species of Hibiscus is published.

Tachibana (1958) and Kuwada (1964) made

successful crosses between H. mutabilis and

H. moscheutos, but the obtained F1 seedlings

were sterile. Attempts to create interspecific

hybrids between H. syriacus and H. rosa-sinensis,

especially to introduce new flower colours and

forms, were not successful so far (Yu et al. 1976;

Paek et al. 1989c).

In this study, the objective was to evaluate the

possibilities of interspecific hybridisation within

frost resistant Hibiscus. More specific, the aim

was to introgress increased growth vigour

into Hibiscus syriacus by interspecific hybridi-

sation with Hibiscus paramutabilis and Hibiscus

sinosyriacus. The hybrid character of the obtained

seedlings was analysed by morphological para-

meters and AFLP analysis.

Materials and methods

Plant material

Different cultivars of Hibiscus syriacus [tetraploid

‘Oiseau Bleu’, ‘Melwhite’, ‘Freedom’ and

272 Euphytica (2007) 155:271–283

123

octoploid ‘Purple CV2’, ‘Red Heart CV’ (Shim

et al. 1993)] were used to make reciprocal crosses

with H. paramutabilis and H. sinosyriacus [tetra-

ploid ‘Autumn Surprise’, ‘Lilac Queen’ (Bean

1973) and ‘Melmauve’].

Interspecific crosses

Hand pollination, preceeded by emasculation,

was done from the end of July till the end of

September between 2000 and 2005, under

controlled environmental conditions in an insect

free plastic greenhouse. Pollination was

performed by rubbing anthers with releasing

pollen over a susceptible stigma of a fully

opened flower. Whenever not impeded by

various reasons, also the reciprocal crosses were

performed. An overview of the crosses is given

in Table 1.

One F1 seedling of H. syriacus ‘Red Heart

CV’ · H. paramutabilis and about 250 F1 seedlings

of H. syriacus ‘Oiseau Bleu’ · H. paramutabilis

flowered in 2004 and 2005. An F2 generation was

made of these F1 plants by self-pollination and by

back crossing of the F1 plants to the female parent

H. syriacus.

In vitro ‘embryo rescue’

Ten to eleven weeks after pollination, swollen

fruits were harvested. The seeds were sterilised in

a 10% NaOCl solution with a drop of Teepol and

rinsed three times in autoclaved water. The

embryos could be easily isolated from the ovules

and were transferred to an in vitro germination

medium containing half strength MS medium

(Murashige and Skoog 1962), 30 g l–1 sucrose and

7 g l–1 agar (pH 5.8). The media were autoclaved

(121�C, 30 min). The embryos were cultured in

petri dishes (diameter 5.5 cm, 10 ml medium

per dish, sealed with LDPE foil). After the

formation of the first true leaves, the germinated

embryos were transferred to jars containing

100 ml medium per jar. Cultures were maintained

at 23 ± 2�C under a 16 h photoperiod at

40 lMol m–2 s–1 PAR, supplied by cool white

fluorescent lamps (OSRAM L36W/31). Seedlings

were transferred to fresh medium every 6 weeks.

Three to four months after embryos were placed

on the in vitro medium, the obtained seedlings

were transferred to the greenhouse in 77-well

trays filled with peat mixture (Organic matter

20%, dry matter 25%, 1.5 kg m-3 fertiliser:

12N:14P:24K + trace elements). They were kept

in a fog unit during 3 weeks for acclimatisation.

In vivo sowing

Seeds were harvested when fully matured. They

were sown directly in 77-well trays filled with peat

mixture with the same composition as described

above.

Morphological characteristics

The progenies obtained from crosses made in

2003 were morphologically screened in the field.

Of the F1 seedlings that had intermediate

characteristics compared to the parent plants,

leaf parameters, flower colour and fertility were

determined.

Leaf parameters of these hybrid seedlings and

their parent plants were determined as described

by Van Huylenbroeck et al. (2000). The following

parameters were measured on fully developed

leaves: angle of the leaf basis (a), leaf length over

width (L/B), the relative width of the mid lob

(M/B), relative length of the side lobes (C/L) and

the relative depth of indentation of the lobes

(A/L) (see Fig. 1).

Fig. 1 Schematic overview of the measured Hibiscus leafparameters. The parameters were measured on fullydeveloped leaves: angle of the leaf basis (a), leaf lengthover width (L/B), the relative width of the mid lob (M/B),relative length of the side lobes (C/L) and the relativedepth of indentation of the lobes (A/L)

Euphytica (2007) 155:271–283 273

123

Flower colour of the plants was determined

using the Royal Horticultural Colour Chart (RHS

colours).

The fertility of both parental plants and the

F1 progenies was determined by analysing the

seed formation on the plants and the pollen

germination vigour. Therefore, pollen was

collected and incubated overnight (dark) in

a liquid medium according to Tian and Russell

(1997) with slight modifications (100 g l–1

sucrose, 100 mg l–1 H3BO3, 100 mg l–1 CaCl2,

200 mg l–1 MgSO4 � 7H2O, 100 mg l–1 KH2PO4,

pH 5.4). After incubation, the pollen was trans-

ferred to a glass slide and observed with a light

microscope (LEICA DM IRB). When the pollen

had formed a pollen tube that was twice the

diameter of the pollen, it was considered as

vigorous germinating.

DNA extraction and AFLP

Young leaf material of 27 generated H. syriacus

‘Oiseau Bleu’ · H. paramutabilis seedlings, eight

H. syriacus ‘Melwhite’ · H. sinosyriacus ‘Lilac

Queen’ and their parent plants was harvested

for AFLP analysis. DNA extraction for AFLP

analysis was based on the CTAB method by

Doyle and Doyle (1987). AFLP reactions were

run on an ABI 3130 DNA sequencer using

the commercially available kit for fluorescent

fragment detection (Perkin-Elmer 1995). EcoRI

and MseI were used for DNA digestion. Selective

amplification was done using two fluorescent

labelled EcoRI-MseI primer combinations

with six selective bases: E-AAC + M-CAT and

E-ACA + M-CTG. The number of markers

uniquely present in each of the parent plants

was checked for segregation in the offspring.

Statistical calculations and data analysis

Statistical calculations on the measurements of

the leaf indices were done using SPSS 11.5 for

Windows (one way ANOVA). If significant dif-

ferences occurred, means were separated by the

LSD (P = 0.05) method.

On the AFLP data multivariate analysis was

performed using Simple Matching Similarity

indices in SPSS 11.5 for Windows. A phyloge-

netic tree (dendrogram) was constructed using

clustering with the Unweighted Pair Group

Method with Arithmatic Mean (UPGMA)

(Sneath and Sokal 1973).

Results

In total over the different years, 590 H. syriacus

· H.paramutabilis crosses were made. The crosses

resulted in 263 fruits (44.6% successful pollina-

tions) which contained on average 13.8 seeds

(Table 1). However, in the reciprocal cross with

H. paramutabilis used as female parent, no fruits

were obtained. The success of a cross also was

cultivar dependent. When using H. syriacus ‘Red

Heart CV’ as female parent, only in 5.5% of the

pollinations fruits were obtained.

For the cross H. syriacus · H. sinosyriacus, in

total 824 syriacus flowers were pollinated. Only 27

fruits were obtained (3.3% successful pollinations).

The fruits contained on average 8.7 seeds. Also here

unilateral incongruity was observed. No fruits were

yielded in the H. sinosyriacus · H. syriacus crosses

(Table 1).

In the control crosses of H. syriacus · H. syriacus

17 fruits were obtained (9.7% successful pollinations)

which contained on average 18.6 seeds.

In vitro embryo rescue versus in vivo sowing

A comparison in efficiency between in vitro

‘embryo rescue’ and in vivo sowing was made for

each cross combination (Tables 2, 3). In general

in vivo sowing resulted in more plantlets for

H. syriacus · H. paramutabilis crosses. The

germination rate of the seeds was 42.8% in vitro

versus 80.7% in vivo. However, for the cross

H. syriacus · H. sinosyriacus the in vitro germi-

nation rate of the embryos was 87.4% while the

in vivo germination rate was 52.0%. For

the control cross of H. syriacus · H. syriacus

no difference in germination capacity was

observed between both sowing methods. Germi-

nation of the seeds was 78.2% in vitro and

73.1% in vivo.

A lot of the in vitro obtained seedlings were

lost during acclimatisation due to total and

274 Euphytica (2007) 155:271–283

123

variegated albinism and to growth aberrations. Of

the in vitro obtained H. syriacus · H. paramuta-

bilis seedlings 83.6% could be planted in the field

versus 91.2% of the in vivo seedlings. For the H.

syriacus · H. sinosyriacus seedlings this was

34.5% versus 75.6% respectively. In the control

cross finally 73.3% of the in vitro obtained

seedlings and 88.4% of the in vivo seedlings were

planted on the field (Tables 2, 3).

In general no genotype effect was present for

germination and acclimatisation rates.

Morphological analysis

The seedlings obtained from crosses made in 2003

were morphologically screened in the field. Only in

the progenies of H. syriacus ‘Oiseau Bleu’ ·H. paramutabilis and H. syriacus ‘Melwhite’ ·H. sinosyriacus ‘Lilac Queen’ seedlings with inter-

mediate characteristics were found (Table 4). Of

the F1 progeny from H. syriacus ‘Oiseau Bleu’ ·H. paramutabilis 95.3% grew very vigorous and

seemed to have intermediate characteristics com-

Table 1 Overview of the different interspecific crosses, obtained fruits and seeds between Hibiscus syriacus (H.s.), Hibiscussinosyriacus (H.si.) and Hibiscus paramutabilis during different years

Yearofcross

Parentage No. ofpollinatedflowers

No. offruits

Successfulpollinations(%)

No. ofseeds

No. ofseeds perfruit$ #

2000 H.s. ‘Oiseau Bleu’ H. paramutabilis 41 7 17.0 45 6.4H.s. ‘Red Heart CV’ H. paramutabilis 48 5 10.4 26 5.2H. paramutabilis H.s.’Oiseau Bleu’ 2 –a 0.0 – –H. paramutabilis H.s. ‘Red Heart CV’ 5 – 0.0 – –

2003 H.s. ‘Oiseau Bleu’ H. paramutabilis 62 38 61.2 766 20.12004 H.s. ‘Melwhite’ H. paramutabilis 63 50 79.4 485 9.7

H.s. ‘Oiseau Bleu’ H. paramutabilis 99 79 79.8 877 11.1H.s. ‘Purple CV2’ H. paramutabilis 47 27 57.4 300 11.1H.s. ‘Red Heart CV’ H. paramutabilis 42 – 0.0 – –H. paramutabilis H.s. ‘Melwhite’ 29 – 0.0 – –H. paramutabilis H.s. ‘Oiseau Bleu’ 10 – 0.0 – –H. paramutabilis H.s. ‘Purple CV2’ 8 – 0.0 – –H. paramutabilis H.s. ‘Red Heart CV’ 7 – 0.0 – –

2005 H.s. ‘Freedom’ H. paramutabilis 188 57 30.3 1145 20.12003 H.s. ‘Oiseau Bleu’ H.si. ‘Melmauve’ 80 4 5.0 29 7.3

H.s. ‘Oiseau Bleu’ H.si. ‘Lilac Queen’ 115 2 1.7 24 12H.s. ‘Melwhite’ H.si. ‘Melmauve’ 60 – 0.0 – –H.s. ‘Melwhite’ H.si. ‘Lilac Queen’ 158 12 2.5 82 6.8H.s. ‘Red Heart CV’ H.si. ‘Lilac Queen’ 75 – 0.0 – –H.s. ‘Purple CV2’ H.si. ‘Lilac Queen’ 88 7 8.0 94 13.4

2004 H.s. ‘Melwhite’ H.si. ‘Autumn Surprise’ 11 – 0.0 – –H.s. ‘Purple CV2’ H.si. ‘Autumn Surprise’ 14 2 14.3 8 4H.s. ‘Melwhite’ H.si. ‘Lilac Queen’ 52 – 0.0 – –H.s. ‘Oiseau Bleu’ H.si. ‘Lilac Queen’ 61 – 0.0 – –H.s. ‘Purple CV2’ H.si. ‘Lilac Queen’ 69 – 0.0 – –H.s. ‘Red Heart CV’ H.si. ‘Lilac Queen’ 41 – 0.0 – –H.si. ‘Autumn Surprise’ H.s. ‘Melwhite’ 34 – 0.0 – –H.si. ‘Autumn Surprise’ H.s. ‘Oiseau Bleu’ 37 – 0.0 – –H.si. ‘Autumn Surprise’ H.s. ‘Purple CV2’ 39 – 0.0 – –H.si. ‘Autumn Surprise’ H.s. ‘Red Heart CV’ 56 – 0.0 – –H.si. ‘Lilac Queen’ H.s. ‘Melwhite’ 10 – 0.0 – –H.si. ‘Lilac Queen’ H.s. ‘Oiseau Bleu’ 3 – 0.0 – –H.si. ‘Lilac Queen’ H.s. ‘Purple CV2’ 38 – 0.0 – –H.si. ‘Lilac Queen’ H.s. ‘Red Heart CV’ 47 – 0.0 – –

2003 H.s. ‘Oiseau Bleu’ H.s. ‘Melwhite’ 61 7 11.5 150 21.4H.s. ‘Purple CV2’ H.s. ‘Melwhite’ 88 6 6.8 115 19.1

2004 H. s. ‘Oiseau Bleu’ H. s. ‘Melwhite’ 27 4 14.8 51 12.8

a No fruit set was observed

Euphytica (2007) 155:271–283 275

123

pared to the parent plants. Morphology of the

other 4.7% of the progeny was the same as H.

syriacus and these seedlings were considered to be

the result of a self-pollination of H. syriacus. Of the

F1 progeny from H. syriacus ‘Melwhite’ · H.

sinosyriacus ‘Lilac Queen’ 52.9% of the seedlings

had intermediate characteristics compared to the

parent plants. These seedlings also grow very

vigorous. The other 47.1% of the seedlings looked

morphologically like the mother parent H. syriacus

‘Melwhite’ and were the result of a self-pollination

of H. syriacus.

Leaf morphology of the H. syriacus ‘Oiseau

Bleu’ · H. paramutabilis seedlings (Table 5)

changed since F1 hybrids became older. The

1-year-old hybrids all showed a uniform inter-

mediate leaf type 1 with L/B comparable to

H. paramutabilis and C/L the same as H. syriacus

(Table 5, Fig. 2). The leaf angle and the relative

width of the mid lob (M/B ratio) were higher than

both parent plants. The indentation of the lobes

(A/L) was intermediate. When the hybrid seed-

lings were 2-years-old, two types of leaves could

be distinguished on the same plant. The first type

had a same M/B, A/L and C/L ratio compared

with the leaves of a 1-year-old seedling. They had

a larger L/B ratio and a smaller leaf angle. The

second leaf type was more indented than the first

one as was expressed by a lower A/L and M/B

ratio. Hybrids that were 3-years-old showed four

different types of leaves (Table 5, Fig. 2). Type 1

and type 2 leaves were statistically the same as

type 1 and type 2 leaves of a 2-year-old seedling.

Type 3 leaves were again more indented than

type 2 leaves as shown in the A/L, C/L and M/B

ratios. The 4th leaf type was asymmetric, as it was

indented only at one side (see Fig. 2). The side

that was not indented expressed a same A/L and

C/L ratio as type 1 leaves, while the A/L ratio

of the indented side was lower than for H.

syriacus. Also on the 3-year-old hybrid seedlings

of H. syriacus ‘Melwhite’ · H. sinosyriacus ‘Lilac

Queen’ the same four different types of leaves

could be distinguished on the same plant (Fig. 2).

Type 1 leaves had an L/B ratio, leaf angle and

indentation comparable with H. sinosyriacus, the

M/B ratio was intermediate. Indentation in-

creased in type 2 and type 3 leaves, while the

4th leaf type was indented only at one side.

Compared to both parent plants, the F1 hybrid

seedlings of H. syriacus ‘Oiseau Bleu’ · H. para-

mutabilis had bigger flowers, while the flower colour

was intermediate (Table 6). Flower colours of all

the F1 hybrids belonged, according to the RHS-

charts, to the purple group. Within this group, three

subgroups could be observed: 48.5% of the seed-

lings belonged to RHS purple group 76A, 24.5% to

RHS purple group 76B and 27.0% to RHS purple

group 76D (Table 6). So far, only seven F1 hybrid

seedlings of H. syriacus ‘Melwhite’ · H. sinosyria-

cus ‘Lilac Queen’ flowered. Flowers of these hybrids

were also bigger compared to their parents and

flower colours were intermediate (Table 7).

Molecular analysis

For the AFLP analysis of H. syriacus ‘Oiseau

Bleu’ · H. paramutabilis, leaf material of 27 F1

seedlings was harvested. Among them, two plants

(OISPAR-3 and OISPAR-20) that looked mor-

Table 2 Obtained number of seedlings from different interspecific Hibiscus crosses via in vitro ‘embryo rescue’ 11 weeksafter pollination

Year ofcross

Parentage No.ofseeds

No. ofseedlingsin vitro

Germina-tion ratein vitro (%)

No. ofseedlingstransferredto greenhouse

No. ofseedlingson the field$ #

2000 H.s. ‘Oiseau Bleu’ H. paramutabilis 45 7 15.6 6 3H.s. ‘Red Heart CV’ H. paramutabilis 26 5 19.2 1 1

2005 H.s. ‘Freedom’ H. paramutabilis 144 80 55.6 75 732003 H.s. ‘Oiseau Bleu’ H.si. ‘Melmauve’ 29 28 96.6 10 8

H.s. ‘Oiseau Bleu’ H.si. ‘Lilac Queen’ 24 22 91.6 8 6H.s. ‘Melwhite’ H.si. ‘Lilac Queen’ 12 10 83.3 3 3H.s. ‘Purple CV2’ H.si. ‘Lilac Queen’ 94 79 84.0 33 31

2003 H.s. ‘Purple CV2’ H.s. ‘Melwhite’ 115 90 78.3 69 66

276 Euphytica (2007) 155:271–283

123

phologically like the mother parent H. syriacus

were sampled. The other 25 plants had intermedi-

ate morphological characteristics as described

above. The two primers generated in total 311

marker bands. The two crossing parents appeared

to be very polymorphic. Hibiscus paramutabilis

showed 80 completely unique markers, while

H. syriacus ‘Oiseau Bleu’ could be characterised

by 75 unique markers (Table 8). Of the 27 tested F1

seedlings 25 showed an equal share of markers that

could be traced back to each of the crossing parents

indicating the true hybrid nature of the offspring

(Table 8). In the other two seedlings (OISPAR-3

and OISPAR-20) almost no unique markers of the

male parent inherited, proving that those two

seedlings were self-pollinations of H. syriacus.

Figure 3 gives an overall view of the genetic

relationships between the plants analysed by

AFLP markers. The group of hybrids clearly was

genetically different from both H. syriacus and

Table 3 Obtained number of seedlings from different interspecific Hibiscus crosses after in vivo sowing

Year ofcross

Parentage No. ofseeds

No. ofseedlingsin vivo

Germinationratein vivo (%)

No. ofseedlingson the field$ #

2003 H.s. ‘Oiseau Bleu’ H. paramutabilis 766 660 86.2 6412004 H.s. ‘Melwhite’ H. paramutabilis 485 334 68.9 226

H.s. ‘Oiseau Bleu’ H. paramutabilis 877 705 80.4 592H.s. ‘Purple CV2’ H. paramutabilis 300 298 99.3 298

2005 H.s. ‘Freedom’ H. paramutabilis 1,001 772 77.1 7692003 H.s. ‘Melwhite’ H.si. ‘Lilac Queen’ 70 40 57.1 312004 H.s. ‘Purple CV2’ H.si. ‘Autumn Surprise’ 8 1 12.5 –2003 H.s. ‘Oiseau Bleu’ H.s. ‘Melwhite’ 150 136 90.6 1262004 H.s. ‘Oiseau Bleu’ H.s. ‘Melwhite’ 51 11 21.6 4

Table 4 Generalmorphological analysis ofthe F1 seedlings obtainedafter interspecific crossesin 2003

Parentage No. of plantson the field

No. of plants withintermediatecharacteristics

No. of plantswithH. syriacuscharacteristics

$ #

H. s. ‘Oiseau Bleu’ H. paramutabilis 641 611 30H. s. ‘Oiseau Bleu’ H. si. ‘Melmauve’ 8 0 8H. s. ‘Oiseau Bleu’ H. si. ‘Lilac Queen’ 6 0 6H. s. ‘Melwhite’ H. si. ‘Lilac Queen’ 34 18 16H. s. ‘Purple CV2’ H. si. ‘Lilac Queen’ 31 0 31

Table 5 Leaf indices for H. syriacus ‘Oiseau Bleu’,H. paramutabilis and H. syriacus ‘Oiseau Bleu’ · H. para-mutabilis F1 hybrids (crosses in 2003) for different plant

ages and leaf types (n = 10). (C1, A1: measured at mostindented site ; C2, A2: measured at less indented site)

Leaftype L/B a M/B C1/L A1/L C2/L A2/L

H. syriacus ‘Oiseau Bleu’ 1.41 e z 112 a 0.27 b 0.55 ab 0.40 c 0.56 bc 0.41 bH. paramutabilis 0.88 ab 242 d 0.43 d 0.70 d 0.64 f 0.71 d 0.65 gF1 (1-year-old) Type 1 0.92 bc 274 e 0.50 ef 0.51 a 0.49 de 0.52 ab 0.49 deF1 (2-year-old) Type 1 1.04 d 190 b 0.52 f 0.56 b 0.53 e 0.55 bc 0.53 f

Type 2 0.92 bc 196 bc 0.35 c 0.62 c 0.47 d 0.59 c 0.47 cdF1 (3-year-old) Type 1 0.98 cd 201 bc 0.47 de 0.55 ab 0.53 e 0.56 bc 0.53 ef

Type 2 0.91 b 212 c 0.34 c 0.56 b 0.41 c 0.57 c 0.43 bcType 3 0.82 a 236 d 0.18 a 0.48 a 0.25 a 0.49 a 0.24 aType 4 0.92 bc 192 b 0.38 c 0.61 c 0.32 b 0.58 c 0.53 ef

z Means indicated by the same symbol are not statistically different (LSD 0.05)

Euphytica (2007) 155:271–283 277

123

H. paramutabilis. The hybrids were more related to

each other than to one of the crossing parents. Only

the two hybrids that were considered to be the

result of a self-pollination of H. syriacus, were

indeed clustered close to H. syriacus ‘Oiseau Bleu’.

For the AFLP analysis of H. syriacus

‘Melwhite’ · H. sinosyriacus ‘Lilac Queen’, leaf

material of eight F1 seedlings was harvested

from which two plants (MELLILAC-2 and

MELLILAC-6) looked morphologically like the

mother parent H. syriacus. Also here, in the

intermediate looking seedlings enough unique

markers could be found from both parent plants

(Table 9). This indicated the true hybrid nature of

these plants. The two seedlings MELLILAC-2 and

MELLILAC-6 had no unique markers of the male

parent inherited, as was expected. Clustering of the

seedlings as presented in a dendrogram (Fig. 4)

demonstrated also the hybrid and self-pollinated

origin of the seedlings.

F2 generation of H. syriacus ·H. paramutabilis

Only 1.9% of the pollen of the H. syriacus ‘Oiseau

Bleu’ · H. paramutabilis hybrids germinated in

vitro, while H. syriacus pollen had an in vitro

germination rate of 56.0% and H. paramutabilis

43.0%. Despite this low germination vigour of

the hybrid pollen, a small F2 generation was

obtained from both H. syriacus ‘Oiseau Bleu’ ·H. paramutabilis hybrids and H. syriacus ‘Red

Heart CV’ · H. paramutabilis hybrids. After in

vitro ‘embryo rescue’, three F2 seedlings resulting

from self-crosses of H. syriacus ‘Red Heart

CV’ · H. paramutabilis and four F2 seedlings

of H. syriacus ‘Oiseau Bleu’ · H. paramutabilis

could be acclimatised. Also two F2 plants from

backcrosses of H. syriacus ‘Red Heart CV’ · (H.

syriacus ‘Red Heart CV’ · H. paramutabilis) were

obtained (Table 10).

Fig. 2 Leaf morphology of the four different leaf types ofthe 3-year-old H. syriacus ‘Oiseau Bleu’ · H. paramutabilisand H. syriacus ‘Melwhite’ · H. sinosyriacus ‘Lilac Queen’hybrids compared to the leaf morphology of the parent plants

Table 6 Flower characteristics of H. syriacus ‘Oiseau Bleu’, H. paramutabilis and the H. syriacus ‘Oiseau Bleu’ · H.paramutabilis F1 hybrids obtained after interspecific crosses in 2003

Petala Basal spot on thepetals

Percentage ofseedlings

Diameterflowers (cm)b

Type offlower

H. syriacus ‘Oiseau Bleu’ Violet–purple group, 93B Red–purple group, 71A 10.5 a SingleH. paramutabilis 56D (white) Red group, 45A 14.0 b SingleF1 Purple group, 76A Red–purple group, 60A 48.5 16.0 c Single

Purple group, 76B Red–purple group, 60B 24.5 16.0 c SinglePurple group, 76D Red–purple group, 60B 27.0 16.0 c single

a Colour was determined by using the Royal Horticultural Society Colour Chartb Means indicated by the same symbol are not statistically different (LSD 0.05)

278 Euphytica (2007) 155:271–283

123

Discussion

Our objective was to evaluate the possibilities

of interspecific hybridisation within frost

resistant Hibiscus. Interspecific hybrids between

H. syriacus, H. paramutabilis and H. sinosyria-

cus were obtained. However crossing compati-

bility was unilateral and only successful when

H. syriacus was used as the mother parent. The

success rate of a cross was also cultivar depen-

dent. In our study, H. syriacus ‘Red Heart CV’

was less successful. Differences in compatibili-

ties in crosses among different Hibiscus species

were also reported by Kyung and Kim (2001a,

b). Unilateral incongruity has been described

in the H. mutabilis · H. moscheutos crosses

Table 7 Flower characteristics of H. syriacus ‘Melwhite’, H. sinosyriacus ‘Lilac Queen’ and the H. syriacus ‘Melwhite’ · H.sinosyriacus ‘Lilac Queen’ F1 hybrids obtained after interspecific crosses made in 2003

Petalsa Basal spot on the petals No.ofseedlings

Diameterflowers(cm)b

Type offlower

H. syriacus ‘Melwhite’ Purple group 76A–76D Red–purple group 60A 8 a SingleH. sinosyriacus ‘Lilac Queen’ 56D (White) / 11 b SingleF1 Purple group, 76A Red–purple group, 60A 3/7 15.2 c Single

Purple group, 76B Red–purple group, 60B 3/7 15.2 c SinglePurple group, 76D Red–purple group, 60B 1/7 15.2 c single

a Colour was determined by using the Royal Horticultural Society Colour Chartb Means indicated by the same symbol are not statistically different (LSD 0.05)

Table 8 Distribution ofAFLP––markers that areunique to one of thecrossing parents afterinterspecific H. syriacus‘Oiseau Bleu’ · H.paramutabilis crossesmade in 2003

Plant No. of markers segregating for H.syriacus parent

No. of markers segregating forH. paramutabilis parent

Unique to H. s.parent

In offspringplant

Unique to H. p.parent

In offspringplant

OISPAR-1 75 60 80 57OISPAR-2 75 60 80 60OISPAR-3 75 69 80 3OISPAR-4 75 60 80 61OISPAR-5 75 55 80 63OISPAR-6 75 57 80 63OISPAR-7 75 60 80 57OISPAR-8 75 59 80 61OISPAR-9 75 57 80 54OISPAR-10 75 55 80 54OISPAR-11 75 58 80 55OISPAR-12 75 54 80 64OISPAR-13 75 57 80 55OISPAR-14 75 57 80 64OISPAR-15 75 54 80 58OISPAR-16 75 57 80 63OISPAR-17 75 56 80 61OISPAR-18 75 54 80 65OISPAR-19 75 58 80 64OISPAR-20 75 72 80 8OISPAR-21 75 58 80 63OISPAR-22 75 57 80 61OISPAR-23 75 55 80 59OISPAR-24 75 56 80 59OISPAR-25 75 56 80 50OISPAR-26 75 57 80 55OISPAR-27 75 58 80 63

Euphytica (2007) 155:271–283 279

123

Fig. 3 Dendrogramshowing the geneticrelatedness of 27 H.syriacus ‘OiseauBleu’ · H. paramutabilisseedlings and the parentplants tested by AFLP

Table 9 Distribution ofAFLP––markers that areunique to one of the crossingparents after interspecificH. syriacus‘Melwhite’ · H. sinosyriacus‘Lilac Queen’ crosses madein 2003

Plant No. of markers segregating forH. syriacus parent

No. of markers segregating forH. sinosyriacus parent

Unique toH. s. parent

In offspring plant Unique toH. si. parent

In offspring plant

MELLILAC-1 47 31 73 55MELLILAC-2 47 43 73 1MELLILAC-3 47 30 73 52MELLILAC-4 47 33 73 51MELLILAC-5 47 35 73 55MELLILAC-6 47 40 73 0MELLILAC-7 47 37 73 52MELLILAC-8 47 31 73 53

280 Euphytica (2007) 155:271–283

123

of Kuwada (1964) and Tachibana (1958) and

in many other species, like Lilium (Van Tuyl et al.

1991) and Bromeliaceae (Vervaeke et al. 2001).

In our study in vitro embryo rescue was

compared with direct in vivo sowing of the seeds

in the greenhouse, since it was possible for the

fruits to stay on the plant until they where fully

matured. For the embryo rescue, the embryos

were isolated 10–11 weeks after pollination. They

had developed by then to the torpedo stage or in

some cases even to the cotyledonairy stage and

could easily be isolated from the ovules. So it was

possible to perform embryo culture instead of

ovule culture. Embryo rescue in H. syriacus was

already successfully performed following intra-

specific tetraploid (so called diploid) · octoploid

(so called tetraploid) pollination (Kim et al.

1996a, b). Embryo culture also was applied in

many other ornamental species where pollinated

flowers could remain on the plant for a notable

time as in Alstroemeria (Buitendijk et al. 1992)

and Lilium (Van Tuyl et al. 1991). In our study

both in vitro embryo rescue and in vivo sowing

were successful, since for the two methods F1

seedlings were obtained. Although higher amount

of seedlings from H. syriacus · H. sinosyriacus

crosses could be generated in vitro, a lot of them

were lost during acclimatisation due to albinism

and growth aberrations. This phenomenon of

chlorosis is frequently observed in interspecific

crosses as in Rhododendron (Michishita et al.

2002) and Zantedeschia (Yao and Cohen 2000). A

logical approach towards this problem would be

to perform the reciprocal cross. Our experimental

setup already anticipated to this problem since

reciprocal crosses were performed whenever

possible. However, those were hampered by

unilateral incongruity.

Morphological screening of the F1 seedlings on

the field together with AFLP analysis revealed that

most of the in vitro and in vivo H. syriacus ‘Oiseau

Bleu’ · H. paramutabilis seedlings were true

interspecific hybrids. But of the progeny obtained

from H. syriacus ‘Melwhite’ · H. sinosyriacus ‘Lilac

Queen’ crosses only about 50% of the obtained

seedlings were true hybrids as proven by morpho-

logical and genetic analysis. The other 50% of the

seedlings as well as all acclimatised seedlings from

other H. syriacus · H. sinosyriacus cross

combinations were self-pollinations of H. syriacus.

Fig. 4 Dendrogramshowing the geneticrelatedness of eight H.syriacus ‘Melwhite’ · H.sinosyriacus ‘Lilac Queen’seedlings and the parentplants tested by AFLP

Table 10 Self-crosses and backcrosses of the H. syriacus · H. paramutabilis F1 hybrids and the generated F2 seedlings.(REDPAR: H.s. ‘Red Heart CV’ · H. paramutabilis; OISPAR: H.s. ‘Oiseau Bleu’ · H. paramutabilis)

Cross No. ofpollinations

No. offruits

Success ofpollination (%)

No. ofseeds

Seedsper fruit

No. ofseedlings invitro

No. of seedlings ingreenhouse

REDPAR · REDPAR 274 1 0.36 14 14.0 4 3OISPAR · OISPAR 708 15 2.12 29 1.9 5 4‘Oiseau

Bleu’ · OISPAR337 1 0.30 2 2.0 – –

‘Red HeartCV’ · REDPAR

249 3 1.20 21 7.0 2 2

Euphytica (2007) 155:271–283 281

123

Probably these self-pollinations were the green

vigorous growing seedlings in vitro, while the

aberrant (albino) plantlets which could not be

acclimatised might have been hybrids.

In both H. syriacus · H. paramutabilis and

H. syriacus · H. sinosyriacus, true hybrids had

larger flowers compared to their parent plants and

flower colour was intermediate. On adult plants of

both hybrid progenies four different leaf types could

be observed completely at random. Leaf dimor-

phisms were described before for some other plant

genera. Juvenile plants or parts of plants of Hedera

helix bear alternately arranged, palmately lobed

leaves while the mature leaf is entirely ovate or

rhombic (Metcalfe 2005; Wareing and Frydman

1976). Similar observations of different foliage types

are reported in many species of Populus (Critchfield

1960) and also in Morus Alba (Everett 1960). It is

not clear whether or not this phenomenon on the

Hibiscus F1 seedlings can be explained by the

change from juvenile to adult phase. Changes in leaf

shape have never been seen in seedlings of the

parental species H. syriacus, H. sinosyriacus

or H. paramutabilis. So most probably the leaf

polymorphisms in the hybrids are caused by genetic

modifications after interspecific hybridisation.

The obtained hybrid plants were almost sterile.

According to Heslop-Harrison (1999) errors in

chromosome segregation at meiosis may cause F1

sterility. It is a well-known phenomenon

after interspecific hybridisation. Despite the low

fertility of the H. syriacus · H. paramutabilis

hybrids, a small F2 progeny could be obtained

so far. However, a larger F2 progeny is needed to

study segregation of leaf types and flower colours.

Breeding efforts to introgress new features in

Hibiscus cultivars is continued.

Acknowledgements The authors wish to thank the staffof the biotech lab for their support and skilful assistance.This work is financed by BEST-Select and IWT-Flanders(VIS-CO 020802).

References

Bates DM (1965) Notes on the cultivated Malvaceae. 1.Hibiscus. Baileya 13:57–130

Bean WJ (1973) Trees and shrubs hardy in the British isles(8th edition revised volume II). John Murray,London, UK

Buitendijk JH, Ramanna MS, Jacobsen E (1992) Micro-propagation ability: towards a selection criteria inAlstroemeria breeding. Acta Hortic 325:493–498

Critchfield WB (1960) Leaf dimorphism in Populustrichocarpa. Am J Bot 47:699–711

Doyle J, Doyle J (1987) Genomic plant DNA preparationfrom fresh tissue - CTAB method. Phytochem Bull19:11

Eeckhaut T, Van Laere K, De Riek J, Van HuylenbroeckJ (2006) Overcoming interspecific barriers in orna-mental plant breeding. In: Taxeira da Silva JA (ed)Floriculture, ornamental and plant biotechnology:advances and topical issues, 1st edn. Global ScienceBooks, London, UK. pp 540–551

Egolf DR (1971) Hibicsus syriacus ‘Diana’ a new cultivar(Malvaceae). Baileya 17:75–78

Egolf DR (1981) Meet ‘Helene’ a new Hibiscus syriacus.Am. Nurseryman 154:11–67

Everett TH (1960) New illustrated encyclopedia of gar-dening, vol. 7. Greystone Press

Heslop-Harrison Y (1999) Aspects of the cell biology ofpollination and wide hybridisation. In: Cresti M, CaiG, Moscatelli A (eds) Fertilisation in higher plants(molecular and cytological aspects). Springer, Berlin,Germany, pp 139–144

Hogenboom N (1973) A model for incongruity in intimatepartner relationships. Euphytica 22:219–233

Kim JH, Lee KC (1991) Studies on the flower colourvariation in Hibiscus syriacus L. I. Spectral propertiesof fresh petals and flower colour classification. JKorean Soc Hortic Sci 32:103–111

Kim J, Ham D, Lim H, Lee K (1996a) Cross compatibility,seed germinability, and embryo rescue following2x · 4x crosses in Hibiscus syriacus. J Korean SocHortic Sci 37:713–718

Kim J, Ham D, Lim H, Lee K (1996b) Development ofimmature embryo and abnormal endosperm afterreciprocal crosses between diploids and tetraploids inHibiscus syriacus. J Korean Soc Hortic Sci 37:462–467

Kuwada H (1964) The newly artificially raised amphidip-loid plant named ‘Ai-Fuyo’ (Hibiscus muta-moscheu-tos), obtained from the progeny of H. mutabilis · H.moscheutos (Studies on interspecific and intergenerichybridisation in Malvaceae VII). Jap J Breed 14:27–32

Kyung HY, Kim JH (2001a) Effects of flowering season onthe compatibilities of interspecific single, three wayand double crosses among Hibiscus syriacus, H.sinosyriacus and interspecific hybrids (H. syria-cus · H. paramutabilis). J Korean Soc Hortic Sci42:568–574

Kyung HY, Kim JH (2001b) Relationship between pistilreceptivity and cross compatibility in interspecificsingle, three way and double crosses among Hibiscussyriacus, H. sinosyriacus and interspecific hybrids (H.syriacus · H. paramutabilis). J Korean Soc Hortic Sci42:561–567

Kyung HY, Park SM, Kim JH (2001a) Effects of bud andold-flower pollination on the interspecific single andthree way cross compatibilities between cultivargroups of Hibiscus syriacus and H. sinosyriacus.J Korean Soc Hortic Sci 42:575–580

282 Euphytica (2007) 155:271–283

123

Kyung HY, Park SM, Kim JH (2001b) Embryo abortionand germination inhibition in interspecific single,three way and double crosses among Hibiscus syria-cus, H. sinosyriacus and interspecific hybrids (H.syriacus · H. paramutabilis). J Korean Soc HorticSci 42:737–742

Metcalfe DJ (2005) Biological flora of the British isles:Hedera helix L. J Ecol 93:632–648

Michishita A, Ureshino K, Miyajima I (2002) Plastome-genome incompatibility of Rhododendron serpyllifo-lium (A. Gray) Miq. to evergreen azalea speciesbelonging to series Kaempferia. J Jap Soc Hort Sci71:375–381

Murashige T, Skoog F (1962) A revised medium for rapidgrowth and bioassays with tobacco culture. PhysiolPlant 15:473–497

Niimoto DH (1966) Chromosome numbers of some Hibis-cus species and other Malvaceae. Baileya 14:29–34

Paek K, Hwang J, Jong S, Park S (1989a) In vitropollination, ovule culture of Hibiscus syriacus anddifferentiation of adventious bud and somatic embryofrom callus. J Korean Soc Hortic Sci 30:145–156

Paek K, Hwang J, Jong S, Park S (1989b) Plant regener-ation in tissue cultures initiated from immatureembryos of Hibiscus syriacus. J Korean Soc HorticSci 16:93–104

Paek K, Hwang J, Jong S, Park S (1989c) Somatichybridisation by protoplast fusion in Hibiscus syriacusand Hibiscus rosachinensis. Kor J Breed 21:95–102

Perkin-Elmer (1995) AFLPTM Plant Mapping Kit: ProtocolSharma H (1995) How wide can a wide cross be?

Euphytica 82:43–64Shim KK, Kim KH, Ha YM (1993) Characteristics of

triploid cultivars ‘Diana’ and ‘Helene’ in Hibiscussyriacus L. J Korean Soc Hortic Sci 34:54–67

Skovsted A (1941) Chromosome numbers in the Malva-ceae II. Comptes rendus des traveaux du laboratoireCarlberg. Serie Physiologique 23:195–242

Sneath P, Sokal R (1973) Numerical taxonomy: theprinciples and practice of numerical classification.W.H. Freeman, San Francisco

Tachibana Y (1958) An interspecific hybrid of H. mutabilisL. and H. moscheutos L. (Studies on Hibiscus III).J Hort Ass Japan 27:201–206

Tian HQ, Russell SD (1997) Micromanipulation of maleand female gametes of Nicotiana tabacum: I. Isolationof gametes. Plant Cell Rep 16:555–560

Van De Laar HJ (1997) Hibiscus syriacus. Dendroflora34:43–60

Van Huylenbroeck JM, De Riek J, De Loose M (2000)Genetic relationships among Hibiscus syriacus, Hibis-cus sinosyriacus and Hibiscus paramutabilis revealedby AFLP, morphology and ploidy analysis. GenetResour Crop Evol 47:335–343

Van Tuyl J, De Jeu M (1997) Methods for overcominginterspecific crossing barriers. In: Shivanna K, Sawh-ney V (eds) Pollen biotechnology for crop productionand improvement. University Press, Cambridge, UK.pp 273–292

Van Tuyl JM, Van Dien MP, Van Creij MGM, VanKleinwee TCM, Franken J, Bino RJ (1991) Applica-tion of in vitro pollination, ovary culture, ovuleculture and embryo rescue for overcoming incongru-ity barriers in interspecific Lilium crosses. Plant Sci74:115–126

Vervaeke I, Parton E, Maene L, Deroose R, De Proft M(2001) Prefertilisation barriers between differentBromeliaceae. Euphytica 118:91–97

Wareing PF, Frydman VM (1976) General aspects ofphase change with special reference to Hedera helix L.Acta Hortic 56:57–69

Yao JL, Cohen D (2000) Multiple gene control ofplastome-genome incompatibility and plastid DNAinheritance in interspecific hybrids of Zantedeschia.Theor Appl Genet 101:400–406

Yu TY, Yeam DY, Kim Y (1976) A study on the breedingof Hibiscus syriacus L. on hybridisation amongintroduced tetraploids and H. rosa-sinensis. J KoreanSoc Hortic Sci 17:107–112

Euphytica (2007) 155:271–283 283

123