Journal of Applied Botany and Food Quality 92, 49 - 56 (2019), DOI:10.5073/JABFQ.2019.092.007 1Albrecht Daniel Thaer-Institut, Humboldt-Universität zu Berlin 2Julius Kühn-Institut, Quedlinburg Hybridization between Pelargonium acetosum L’Hér. and Pelargonium ×peltatum R. Kamlah1*, I. Pinker1, S. Plaschil2, K. Olbricht1 (Submitted: December 11, 2018; Accepted: February 7, 2019) * Corresponding author Summary Pelargonium acetosum L’Hér. is a wild species from South Africa with decorative bluish foliage. Only few reports describe crossings between P. acetosum and P. peltatum L’Hér. (or P. ×peltatum). Therefore, information about hybridization barriers is limited. In this study, two different genotypes of Pelargonium acetosum (AC1 and AC2) were crossed with the diploid P. ×peltatum ‘Tornado Fuchsia’ (PTF). Embryos and F1 hybrids from the combination AC1 × PTF were hampered by chlorophyll deficiencies. Embryos and seeds of the combination AC2 × PTF were underdeveloped. The reciprocal combination PTF × AC1 did not show any fruit set. The combination PTF × AC2 resulted in low numbers of seeds, which were normally developed. Hybrids from seeds were only obtained from the combinations AC1 × PTF and PTF × AC2. Embryo rescue of the combinations AC1 × PTF and AC2 × PTF resulted in few but viable hybrids. Flowers of all hybrids had shrivelled anthers and proved to be sterile. The occurrence of most hybridization barriers varied strongly between the different combinations and depended on both the genotype and the direction of cross-breeding. The bluish leaf colour did not appear among the F1. To overcome hybrid sterility a polyploidization is suggested. Keywords: Embryo rescue, hybridization barriers, hybrid variega- tion, hybrid sterility, incomplete embryo development, interspecific hybridization, wild species introgression Introduction In Europe and North America, Pelargonium cultivars represent a significant fraction of the bedding plant market. In order to main- tain that position, breeders regularly have to come up with novelties. Interspecific hybridization is one of the main approaches in creat- ing new characteristics, and it counteracts a narrowing gene pool in Pelargonium breeding programs (Olbricht, 2013). Worldwide about 280 Pelargonium species (Albers and vAn der WAlt, 2007) embody valuable genetic resources that have not yet been fully ex- ploited. However, introgression of species is time-consuming and of- ten hampered by reproductive barriers, which depend, amongst other reasons, on the genetical distance between two species. Interspecific cross-combinations are usually more successful if both species be- long to the same section of Pelargonium. Other factors influencing interspecific crossability include the ploidy level, the basic chromo- some number, and the direction of cross-breeding (hOrn, 1994). DNA-based phylogenetic analyses (bAkker et al., 2004; Weng et al., 2012; röschenbleck et al., 2014) support a subdivision of the genus into 16 sections and help breeders to recognize possible candidates for introgression. The section Ciconium (Sweet) Harv. contains the horticulturally important species P. inquinans L’Hér., P. zonale L’Hér. (both ances- tors of P. ×hortorum Bailey, the ‘zonal geranium’), and P. peltatum L’Hér. (the main ancestor of ‘ivy-leaved’ cultivars, usually named P. ×peltatum). In recent decades, several species of this section have been introgressed into P. ×hortorum, such as P. tongaense Vorster (esenAlievA et al., 2012) and P. quinquelobatum Hochst. (denis- PeixOtO et al., 1997; hOndO et al., 2015). P. peltatum was formerly placed in the section Dibrachya (Sweet) Harv. but has been included in Ciconium (gibby et al., 1990). P. peltatum is characterized by a relatively large genetical distance to other species of the same section (JAmes et al., 2004; Weng et al., 2012). In hybridization experiments between various species of section Ciconium and P. peltatum, both prezygotic and postzygotic barriers have been observed. An inhibited pollen tube growth or a lack of fertilization (when pollen tubes do grow down the style) represent the main observed prezygotic barriers (cOffin and hArney, 1978; yu, 1985). Known postzygotic barriers include incomplete development of seeds, stunted plant growth, hy- brid sterility, chlorophyll deficiencies and hybrid variegation (cOffin and hArney, 1978; yu, 1985; hOrn, 1994). The latter is a conse- quence of the biparental inheritance of plastids in Pelargonium: If only one of the two inherited plastid types shows an incompatibility with the nuclear genome, the segregation of plastids often leads to variegated leaves with chlorophyll-deficient sectors (metzlAff et al., 1981, 1982; grieger, 2007; Weihe et al., 2009). In the case of an incomplete development or the abortion of the em- bryo, postzygotic disturbances may be overcome by the use of em- bryo rescue. In respect to the explant material, embryo rescue tech- niques can be distinguished into ovary, ovule, and embryo culture (WinkelmAnn et al., 2010). A first study about embryo rescue of P. ×hortorum was published by becker-zens (1983), in which both ovule and embryo culture were performed. scemAmA and rAquin (1990) developed a method to circumvent early embryo abortion us- ing a combination of ovary and embryo culture. bentvelsen et al. (1990) applied embryo culture in various crosses between P. ×pel- tatum and other species of the section Ciconium. All these studies used phytohormone-free nutrient media and aimed at the germina- tion of the embryo, while other studies applied growth regulators to induce callus and adventitious shoots (kAtO and tOkumAsu, 1983; kAkihArA et al., 2012). Pelargonium acetosum L’Hér. (sect. Ciconium) is a species from South Africa with 2n=2x=18 chromosomes. It stands out due to its decorative bluish foliage, which is not common among ornamental Pelargonium cultivars. Despite its distinct appearance, it is closely related to P. zonale (JAmes et al., 2004), and has been successfully introgressed into P. ×hortorum (hOndO et al., 2014). A hybridiza- tion between P. acetosum and P. ×peltatum using embryo rescue is documented by bentvelsen et al. (1990), resulting in F1 hybrids but no F2 or backcross (BC) generation. yu (1985) described crossings between P. acetosum and P. peltatum as not resulting in viable hy- brids, while hOrn (1994) reported viable but sterile hybrids from the combination P. peltatum × P. acetosum. The objective of this study was to examine hybridization barriers be- tween P. acetosum and P. ×peltatum and to obtain genotypes with a novel variability for further breeding purposes. The long-term bree- ding aim is to achieve P. ×peltatum cultivars with bluish foliage. 50 R. Kamlah, I. Pinker, S. Plaschil, K. Olbricht Materials and methods Plant material and cultivation Two different genotypes of Pelargonium acetosum (AC1 and AC2) and the diploid cultivar P. ×peltatum ‘Tornado Fuchsia’ (PTF) were used as crossing parents. AC1 was ordered from a nursery (Gärtnerei Schoebel, Germany), AC2 was provided by the Julius Kühn-Institut, Quedlinburg. Both genotypes were received as adult plants and pro- pagated by cuttings. PTF had to be grown from seeds (mail-ordered from Mary K’s Unique Seeds, USA). The cultivation of the plants took place from 08/2016 until 01/2018 in the greenhouse (Berlin- Dahlem). Plants were potted in a mix of organic substrate (Klasmann- Deilmann 5) with 10% sand and were fertilized with 0.6% Wuxal Super every 14 days. Assimilation lighting was added using sodium vapour lamps (11 h/d in 2016, 13.5 h/d from 01/2017 until 04/2017, 11 h/d until 07/2017). The temperature ranged between 17 and 35 ºC. Pollen viability test Pollen of AC1, AC2, and PTF was stained with an aqueous solution of 1% thiazolyl blue tetrazolium bromide (MTT) and 10% sucrose for 30 min at 24 ºC (nOrtOn, 1966; firmAge and dAfni, 2001). From each genotype 3 × 100 pollen grains were evaluated under a transmission light microscope (Olympus IX70-S8F2). The number of pollen grains showing a purple colour was determined. Pollination Flowers were emasculated in the bud stage using forceps (ca. three days before anthesis, when petals became visible at the bud tip). Emasculated inflorescences were covered in perforated propylene bags (Crispac, 11 × 25 cm), which were sealed with Tesa Velcro strips. About five days after emasculation, when the stigma lobes had unfolded, anthers of the crossing partner were pressed on the stigma for pollination. Pollinated flowers were bagged as described above. The number of pollinated flowers varied in each experiment (see be- low). Observation of embryo development For quantitative analysis of embryo development, unripe fruits were harvested two weeks after pollination and the number of embryos per pollinated flower was determined. In each performed combination (Tab. 1) 20 flowers were pollinated. For qualitative examination of embryo development, unripe fruits were harvested two, three, four, and five weeks after pollination (combinations see Tab. 1). In each variant (combination × point of time = 24 variants) at least five pollinations were carried out. Ovules were taken out of the fruit and were dissected under the stereo mi- croscope (Olympus SZ-PT) with the help of forceps and a dissecting needle. Isolated embryos were documented with a connected camera (Olympus UC30) and associated software (Cellsens Entry 1.6). Seed production and sowing The combinations performed in order to achieve seeds are listed in Tab. 1. After harvest, seeds were stored in glassine envelopes. Before sowing, ca. 0.5 mm of the seed tip was removed with fingernail clippers. With the scarified tip facing down, seeds were sown into multi-cell trays (cell size: 6 × 3 × 3 cm) filled with moistened sow- ing substrate (Klasmann-Deilmann 1). Seed trays were placed under a foil tunnel in the greenhouse for 14 days. Four days after sowing, the front end of the tunnel was opened. During the first two weeks after sowing the temperature ranged between 19 and 30 ºC (22.1 ºC mean). Thirteen days after sowing the number of germinated seeds was determined. Three weeks after sowing, seedlings were transplanted into multi- pot trays (5 cm Ø per cavity), which were filled with a mix of or- ganic substrate and perlite (Klasmann-Deilmann Stecklingssubstrat). Twelve weeks after sowing, plants were potted in a mix of organic substrate (Klasmann-Deilmann 5) with 10% sand and cultivated as described (see plant material and cultivation). Embryo rescue Fruits from selected combinations (Tab. 1) were harvested two weeks after pollination. They were surface sterilized in 50 ml disinfec- tion solution (3% calcium hypochlorite and one drop of Tween20) for 20 min and then rinsed three times in sterile deionized water. Under sterile conditions, embryos were excised from fruits using a stereo microscope, forceps, and a dissecting needle. Nineteen em- bryos of AC1 × AC1, 21 embryos of AC1 × PTF, and 20 embryos each of AC2 × PTF and PTF × PTF were placed in glass tubes (10 × 2.5 × 2.5), which were filled with 10 ml of solid embryo-rescue- medium (ER-medium, Tab. 2). This medium was supplemented with 1 mg L-1 (5.71 μM) indole-3-acetic acid (IAA) and 1 mg L-1 (4.4 μM) 6-benzylaminopurine (BAP). For the first two weeks, embryos were cultivated in the dark at 22-24 ºC and subsequently transferred to 10 μmol m-2s-1 photosynthetic active radiation (PAR). Five weeks later (eight weeks later in case of AC2 × PTF), the de- veloped callus was divided into smaller portions and then cultivated in glass tubes (10 × 2.5 × 2.5 cm) each filled with 5 ml phytohor- mone-free active-charcoal-I-medium (AC-I-medium, Tab. 2). The glass tubes were placed at 22-24 ºC and 30 μmol m-2s-1 PAR. Once adventitious shoots reached at least 1 cm, they were transferred into culture jars (Sigma-Aldrich, 66 × 59 × 59 mm, enclosed with Magenta B-cap) filled with 20 ml AC-II-medium. The culture jars were placed at 22-24 ºC and 30 μmol m-2s-1 PAR for 16 h d-1. Rooted and unrooted shoots that had reached at least 2 cm length were planted into 4 cm pots filled with a mix of organic substrate (Klasmann-Deilmann Stecklingssubstrat) and 1/3 perlite. Plants were sprayed with water and placed under a transparent cover (to keep air humidity above 60%) at 21 ºC and 60 μmol m-2s-1 PAR for 16 h d-1. Two weeks later plants were transferred into the greenhouse. Acclimated plants were cultivated as described (see plant material and cultivation). Characterization of plants Crossing parents and progeny were morphologically characterized. Hybrid status was determined based on the morphology and colour of flowers and leaves. Anthers of the hybrids were examined under a stereo microscope. To assess female fertility, stigmas of the hybrids were pollinated with pollen from PTF. Tab. 1: Performed combinations in each experiment Experiment Performed combinations Quantitative analysis AC1 × AC1, AC1 × AC2, AC1 × PTF of embryo development AC2 × AC2, AC2 × AC1, AC2 × PTF PTF × PTF, PTF × AC1, PTF × AC2 Qualitative analysis AC1 × AC1, AC1 × PTF of embryo development AC2 × AC2, AC2 × PTF PTF × PTF, PTF × AC2 Seed production AC1 × AC1, AC1 × PTF and sowing AC2 × PTF PTF × PTF, PTF × AC2 Embryo rescue AC1 × AC1, AC1 × PTF AC2 × PTF PTF × PTF Pelargonium hybridization 51 Tab. 2: Composition of culture media Culture medium Macro- and Growth regulators Vitamins Other components micronutrients (μM) (mg L-1) (g L-1) ER MS 5.71 IAA 2.5 Thiamine-HCl 30 Sucrose 4.40 BAP 0.2 Pyridoxine-HCl 7 Agar 0.2 Biotin 100.0 myo-Inositol AC-I MS - 2.5 Thiamine-HCl 30 Sucrose 0.2 Pyridoxine-HCl 7 Agar 0.2 Biotin 3 Active charcoal 100.0 myo-Inositol AC-II MS - 2.5 Thiamine-HCl 30 Sucrose 0.2 Pyridoxine-HCl 7 Agar 0.2 Biotin 20 Active charcoal 100.0 myo-Inositol ER = embryo rescue, AC = active charcoal, MS = murAshige and skOOg (1962), IAA = indole-3-acetic acid, BAP = 6-benzylaminopurine Statistics Statistical analysis was carried out with IBM SPSS 23. Post-hoc com- parisons were made using Tukey's Honestly Significant Difference (HSD) test. Results and discussion Characterization of the crossing partners The two genotypes of Pelargonium acetosum (AC1 and AC2) were characterized by a matte bluish-green foliage. In contrast to AC1, the leaves of AC2 were more strongly lobed and their bluish colour was more distinct. While the flowers of AC1 had a light salmon-pink co- lour, the flowers of AC2 were first pale yellow and became white one day after anthesis. Both genotypes exhibited a more or less upright growth habit. The cultivar P. ×peltatum ‘Tornado Fuchsia’ (PTF) was characterized by a trailing growth, glossy green foliage, and single fuchsia flowers. The leaf blade showed a horseshoe-shaped dark zone. Both AC1 and PTF were flowering vigorously during this study and proved to be robust. In contrast, AC2 showed a high susceptibility to thrips, which hampered the development of flowers and complicated the use of this genotype. MTT staining (Tab. 3) resulted in a low percentage of stained pol- len grains in the case of AC1 and a high portion of stained pollen in the case of AC2 and PTF. Stained pollen grains were interpreted as able to germinate on a stigma (firmAge and dAfni, 2001), and the percentage of stained pollen was understood as a measure of ferti- lity. Based on these assumptions, the male fertility of AC2 and PTF was considered relatively high (PlAschil et al., 2017). In contrast, AC1 could only be assessed as partially male-fertile, which makes this genotype less suitable as a pollen parent. Interspecific Hybridization Embryo development Fruits from selfings of AC1 (AC1 × AC1) contained a maximum of two embryos and a mean of 0.95 embryos (Tab. 4), most likely due to the low fertility of this genotype. Two weeks after pollination, em- bryos were in transition between the torpedo and the cotyledon stage (Fig. 1). Three weeks after pollination, embryos had already reached their maximum size. The interspecific combination AC1 × PTF re- sulted in a maximum of four and a mean of two embryos per pollina- ted flower, but differences to AC1 × AC1 were not significant (Tab. 4). Embryos from the combination AC1 × PTF were less green than em- bryos from the selfings of AC1 (Fig. 1). These apparent chlorophyll deficiencies are most likely due to an incompatibility between one of the two plastomes and the nuclear genome (metzlAff et al., 1981, 1982; Weihe et al., 2009). Embryos of the combination AC1 × PTF often exhibited a slightly irregular morphology. In particular, the coty- ledons were often unequally sized (Fig. 1). Tab. 3: Pollen viability of the crossing parents according to the percentage of stained pollen grains after thiazolyl blue tetrazolium bromide (MTT) treatment Genotype Mean Standard deviation stained pollen grains (%)1 (%) AC1 38z 16.09 AC2 86y 4.58 PTF 88y 4.00 1different letters indicate significant differences (Tukey-HSD, n = 3 × 100, α = 5%) Tab. 4: Influence of the combination on the number of embryos, two weeks after pollination Combination Mean embryo Standard Total embryo number per deviation number per pollinated flower1 combination AC1 × AC1 0.95 zy 0.83 19 AC1 × AC2 1.65 y 1.46 33 AC1 × PTF 2.00 y 1.38 40 AC2 × AC2 4.05 x 1.10 81 AC2 × AC1 0.00 0.00 0 AC2 × PTF 3.35 x 1.60 67 PTF × PTF 5.45 w 1.57 109 PTF × AC1 0.00 0.00 0 PTF × AC2 0.30 z 0.66 6 1different letters indicate significant differences (Tukey-HSD, n = 20, α = 5%) 52 R. Kamlah, I. Pinker, S. Plaschil, K. Olbricht Selfings of AC2 (AC2 × AC2) resulted in three to five embryos and a mean of 4.05 embryos per pollinated flower (Tab. 4). These num- bers can be considered normal because usually no more than five ovules are fertilized in a Pelargonium flower (yAnO et al., 1975). The interspecific combination AC2 × PTF resulted in a mean of 3.35 embryos per flower, which was not significantly lower compared to AC2 × AC2 (Tab. 4). Despite their relatively high number, these hybrid embryos developed poorly. Two weeks after pollination the embryos had reached, at most, an early torpedo stadium (Fig. 1) and did not develop much further in the following weeks. This delayed and then arrested embryo development might have been caused by an irregular endosperm formation resulting from an incompatibility between the parental genomes in the endosperm (lAfOn-PlAcette and köhler, 2015). Such postzygotic disturbances represent one of the most significant hybridization barriers, which potentially can be overcome by embryo rescue (kuligOWskA et al., 2016). Combinations using PTF as the seed parent contained considerably smaller embryos and seeds than the previously mentioned combina- tions (Fig. 1 and 2). In 11 of 20 fruits deriving from the selfings of PTF (PTF × PTF), the number of five embryos was exceeded because carpels contained ‘twin embryos’. The occurrence of twin embryos in Pelargonium was documented by yAnO et al. (1975) and kubbA and tilney-bAssett (1980). No fruit development and no embryos were observed in the combination PTF × AC1 (Tab. 4), whereas the combination PTF × AC2 resulted in four fruits per 20 pollinated flow- ers containing 1-2 embryos (a mean of 0.3 embryos per pollinated flower). It appears likely that fertilization was hampered by prezy- gotic hybridization barriers (and in the case of PTF × AC1 addition- ally by a low pollen viability). Prezygotic barriers are well-known from interspecific crossings using Pelargonium peltatum as the seed parent (yu, 1985). However, in order to confirm the occurrence of prezygotic barriers, it would be necessary to monitor the pollen tube growth within the pistil (WinkelmAnn et al., 2010). Despite their small number, embryos from the combination PTF × AC2 seemed to develop normally in comparison with PTF × PTF (Fig. 1). Seed morphology and germination While all the seeds resulting from the selfings of AC1 germinated, the germination percentage of the combination AC1 × PTF was only 44% (Tab. 5). Seeds from this combination were not as round as those from the selfings of AC1 (Fig. 2). The development of most seedlings was hampered by chlorophyll deficiencies. Seeds from the combination AC2 × PTF were small and under- developed (Fig. 2). This result agrees with yu (1985), who reported incompletely developed seeds deriving from the combination P. ace- Fig. 1: Morphology of embryos deriving from selfings (control) and interspecific crossings, two and three weeks after pollination (wap), scale bar: 1 mm Fig. 2: Morphology of seeds deriving from selfings (control) and interspecific crossings, scale bar: 1 mm Pelargonium hybridization 53 tosum × P. peltatum. The small and shrivelled seeds from the combi- nation AC2 × PTF (Fig. 2) did not germinate (Tab. 5). This combina- tion could only be achieved by the application of embryo rescue (see below). Out of 96 seeds deriving from the selfings of PTF, 73 germinated (Tab. 5). The interspecific combination PTF × AC2 resulted in only six seeds per 20 pollinated flowers (a mean of 0.3 seeds per pollina- ted flower). They appeared normal-shaped (Fig. 2), and four of them germinated (Tab. 5). hOrn (1994) reported also a low seed set from the combination P. peltatum × P. acetosum (a mean of 0.5 seeds per pollinated flower), but in his case the germination percentage was only 36%. F1 progeny Plants from seeds were only obtained from the combinations AC1 × PTF and PTF × AC2. Except two plants, all hybrids from the combination AC1 × PTF exhibited chlorophyll deficiencies or hybrid variegation, which severely hampered their development (Fig. 3). Consequently, only four plants survived the first six months after sowing. Severe chlorophyll deficiencies leading to a low survival rate in the F1 also occurred after cross-breeding between P. aceto- sum and P. ×hortorum (hOndO et al., 2014). However, the four hy- brids from the combination PTF × AC2 did not show any chlorophyll deficiencies. Compared to the plants originating from the selfings of AC1 and PTF, the development of all the interspecific hybrids was at first charac- terized by considerably short internodes. Then, about 20 weeks af- ter sowing, these hybrids showed elongated internodes and a normal growth habit. However, one hybrid from the combination PTF × AC2 continued growing in a severely stunted manner and did not develop any flowers. Hybrids with such stunted growth habit are known from a hybridization between P. peltatum and P. ×hortorum (cOffin and hArney, 1978) and can be a result of genomic conflict (bOmblies and Weigel, 2007). Hybrids from both interspecific combinations developed sterile flow- ers. Anthers were shrivelled, and no fruit development was observed after pollination with PTF. hOrn (1994) also reported sterile pro- geny from the combination P. peltatum × P. acetosum. Hybrid steri- lity can be a consequence of either karyotypical differences or ge- netic incompatibilities (bOmblies, 2010). If the hybrid sterility was caused by different karyotypes, a polyploidization could restore fer- tility and allow further generations on the tetraploid level (sAttler et al., 2016). Nevertheless, if the sterility was a consequence of ge- netical incompatibilities, a restoration of fertility would be rather unlikely (rieseberg and Willis, 2007). The matte bluish leaf surface of P. acetosum did not appear among hybrids from the combinations AC1 × PTF and PTF × AC2. All plants had leaves with a rather glossy surface comparable to PTF. However, the leaf bases of all hybrids were cordate (unlike the peltate leaf base of PTF), and leaf zonation was not observed. Leaf zonation is domi- nant over zoneless leaves in P. ×hortorum (AmOAtey and tilney- bAssett, 1993), but this does not seem to apply here. Hybrids from the combination PTF × AC2 showed a high susceptibility to thrips, which was most likely inherited from AC2. Information about the inheritance of the waxy leaf surface of P. acetosum is limited. In the hybridization between P. acetosum and P. ×hortorum (hOndO et al., 2014), leaves of the F1 generation showed a rather intermediary level of wax bloom, and hybrids with a leaf surface comparable to P. acetosum first appeared in the F2. In Tab. 5: Influence of the combination on the number of seeds and the germination percentage Combination Mean seed number Standard Total seed number Number of Germination percentage per pollinated flower1 deviation per combination germinated seeds (%) AC1 × AC1 0.80 zy 1.06 16 16 100.00 AC1 × PTF 1.25 y 0.64 25 11 44.00 AC2 × PTF 2.15 x 1.35 43 0 0.00 PTF × PTF 4.80 w 1.15 96 73 76.04 PTF × AC2 0.30 z 0.73 6 4 66.67 1different letters indicate significant differences (Tukey-HSD, n = 20, α = 5%) Fig. 3: Progeny from the combinations AC1 × PTF (left) and PTF × AC2 (right), 14 weeks after sowing. Pot diameters: 9 cm, 7 cm, 4 cm (AC1 × PTF), and 9 cm (PTF × AC2). 54 R. Kamlah, I. Pinker, S. Plaschil, K. Olbricht Fig. 4: Shoot regenerates from the interspecific combinations AC1 × PTF (left) and AC2 × PTF (right), cultured on AC-II-medium (six weeks after transfer). All regenerates within one jar (5.9 cm Ø) originated from the same rescued embryo. order to achieve hybrids with bluish leaves from the hybridization between P. acetosum and P. ×peltatum, a further generation would be necessary, which seems only possible through previous polyploidiza- tion (see above). Hybrid status was determined based on the above-mentioned mor- phological traits and was additionally supported by the occurrence of hybrid incompatibilities. Nevertheless, for an unambiguous con- firmation of hybridity, a molecular analysis of the hybrid character should be considered, such as AFLP (amplified fragment length poly- morphism) or RAPD (random-amplified polymorphic DNA) finger- printing (bArcAcciA et al., 1999; kuligOWskA et al., 2016). Embryo rescue Two weeks after the establishment of embryo culture, most embryos started forming callus. In the case of AC2 × PTF, only six of the 20 cultivated embryos formed callus while the rest died off. Embryos below the torpedo stage did not survive (scemAmA and rAquin, 1990). A later start of embryo culture may have increased the number of surviving embryos (becker-zens, 1983), as embryos of the com- bination AC2 × PTF were slightly more developed three weeks after pollination (Fig. 1). In contrast, embryos of the other combinations were developed enough to begin embryo culture two weeks after pollination because they were at least in a bent torpedo stage (Fig. 1) (bentvelsen et al., 1990; scemAmA and rAquin, 1990). From the fourth week after establishment of culture, countless tiny organoids developed on most samples (except for AC1 × AC1, which showed organoid formation only on less than half of the samples). Most samples showed substantial callus proliferation. In the seventh week after establishment, an increasing number of adventitious shoots showed signs of vitrification. These symptoms are known from the micropropagation of P. ×peltatum on culture media containing BAP (WOJtAniA and gAbryszeWskA, 2001; WOJtAniA, 2010). Some of the samples recovered after transfer to the phytohormone-free culture medium, but only a few adventitious shoots showed substantial elon- gation. Samples of AC1 × AC1 showed a particularly stunted growth and consequently senescence of callus and shoots. After transfer of single shoots into jars, shoots of all samples continued growing slow- ly, and only a few shoots rooted. A strongly inhibitory effect of BAP on shoot elongation of P. ×peltatum was documented by WOJtAniA and gAbryszeWskA (2001). This effect apparently continued a long time after samples were removed from the BAP containing medium. As discussed above, embryos and seedlings of the combination AC1 × PTF showed chlorophyll deficiencies most likely due to an incompatibility between one of the two plastomes and the hybrid genome. In vitro, samples of this combination had a strong tenden- cy towards a segregation of intact and chlorophyll-deficient tissue. Consequently, many shoot regenerates were completely white and some entirely green (Fig. 4). Samples of the combination AC2 × PTF exhibited chlorophyll deficiencies only on intercostal fields of the leaf blade (Fig. 4). Because these symptoms only occurred in vitro, it remains unclear whether these were an effect of genomic conflict or other factors. As a consequence of low shoot qualities, low rooting rates, and chlo- rophyll deficiencies, very few plants endured the transplantation of shoots into organic substrate and the following acclimatization to the greenhouse. From the combination AC1 × PTF, four geno- types had survived six months later, two of them exhibiting chlo- rophyll deficiencies and hybrid variegation. From the combination AC2 × PTF, plants of only one genotype was successfully relocated to the greenhouse. These had green and glossy leaves and showed a high susceptibility to thrips, comparable to hybrids from the com- bination PTF × AC2. All flowering hybrids deriving from embryo rescue exhibited shrivelled anthers. In this regard, they did not differ from hybrids deriving from seeds. Morphological traits indicating a spontaneous polyploidization have not been observed. The latter is a frequent result of in vitro regeneration (PlAschil et al., 2015). Conclusion In the hybridization between P. acetosum and P. ×peltatum, the oc- currence of most hybridization barriers varied strongly between the performed combinations and depended on both the genotype and the direction of cross-breeding. The combination AC1 × PTF was hampered by chlorophyll deficiencies and hybrid variegation. Viable hybrids were both achieved from seeds and via embryo rescue, but the latter did not considerably increase the number of hybrids. The combination AC2 × PTF was characterized by an incomplete embryo development. In this case, hybrids were only accomplished through the use of embryo rescue. The reciprocal combination PTF × AC2 appeared to be impeded by prezygotic barriers, and one hybrid was characterized by a permanently stunted plant growth. All flowering hybrids exhibited sterile flowers, and the occurrence of hybrid steri- lity was independent of both the genotype and the direction of the Pelargonium hybridization 55 cross. Hybrid sterility represented the major barrier in the hybridi- zation between P. acetosum and P. ×peltatum. To overcome hybrid sterility a polyploidization would be necessary (rieseberg and Willis, 2007; sAttler et al., 2016). References Albers, f., vAn der WAlt, J.J.A., 2007: Geraniaceae. In: Kubitzki, K. (ed.), The families and genera of vascular plants. IX Eudicots. Springer, Berlin, Heidelberg, New York, 157-167. AmOAtey, h.m., tilney-bAssett, r.A.e., 1993: Multiple alleles and the control of leaf zonation patterns in zonal pelargoniums. J. Hort. Sci. 68(1), 45-52. DOI: 10.1080/00221589.1993.11516327 bArcAcciA, g., Albertini, e., fAlcinelli, M., 1999: AFLP fingerprinting in Pelargonium peltatum: Its development and potential in cultivar identi- fication. J. Hort. Sci. and Biotech. 74(2), 243-250. DOI: 10.1080/14620316.1999.11511102 bAkker, f.t., culhAm, A., hettiArAchi, P., tOulOumenidOu, t., gibby, m., 2004: Phylogeny of Pelargonium (Geraniaceae) based on DNA se- quences from three genomes. Taxon 53(1), 17-31. becker-zens, r., 1983: Anther and embryo culture in Pelargonium zonale- hybrids. Acta Hortic. 131, 209-214. DOI: 10.17660/ActaHortic.1983.131.23 bentvelsen, g.c.m., stemkens, h.g.W., tJeertes, P., 1990: Interspecific crosses in Pelargonium and the application of embryo rescue methods. In: Jong, J. de (ed.), Integration of in vitro techniques in ornamental plant breeding, 104-109. Proceedings, Eucarpia Symposium Section Ornamentals, Wageningen, 10-14 November 1990. bOmblies, k., 2010: Doomed lovers: mechanisms of isolation and incompa- tibility in plants. Annu. Rev. Plant Biol. 61, 109-124. DOI: 10.1146/annurev-arplant-042809-112146 bOmblies, k., Weigel, d., 2007: Hybrid necrosis: autoimmunity as a poten- tial gene-flow barrier in plant species. Nature Reviews Genetics 8(5), 382. DOI: 10.1038/nrg2082 cOffin, J.l., hArney, P.m., 1978: Intersubgeneric crosses within the genus Pelargonium. Euphytica 27(2), 567-576. DOI: 10.1007/BF00043183 denis-PeixOtO, l., cAdic, A., renOu, J.P., 1997: Interspecific crosses be- tween Pelargonium ×hortorum and P. quinquelobatum using embryo rescue and molecular characterization of hybrids by an endogenous chs probe. Plant Breeding 116(2), 177-180. DOI: 10.1111/j.1439-0523.1997.tb02174.x esenAlievA, A., dreWes-AlvArez, r., ArnOld, r., POhlheim, f., Wiede- mAnn, m., meinl, k., Olbricht, k., 2012: Interspecific hybridisation between Pelargonium zonale hybrids and Pelargonium tongaense Vorster on the tetraploid ploidy level. Acta Hortic. 953, 149-153. DOI: 10.17660/ActaHortic.2012.953.20 firmAge, d.h., dAfni, A., 2001: Field tests for pollen viability; a compara- tive approach. Acta Hortic. 561, 87-94. DOI: 10.17660/ActaHortic.2001.561.13 gibby, M., Albers, f., PrinslOO, B., 1990: Karyological studies in Pelar- gonium sectt. Ciconium, Dibrachya, and Jenkinsonia (Geraniaceae). Plant Syst. Evol. 170, 151-159. DOI: 10.1007/BF00937700 grieger, P., 2007: Untersuchungen zur Züchtung variegater Pelargonium ×zonale-Hybriden auf tetraploider Stufe, PhD thesis, Humboldt-Uni- versität zu Berlin, 129 pages. DOI: 10.18452/15673 hOndO, k., iWAsAkA, c., kAkihArA, f., 2014: Characteristics of progeny obtained from the cross between geraniums (Pelargonium ×hortorum) and wild species in section Ciconium, including heat tolerance and odor of leaves. Environ. Control Biol. 52(1), 37-43. DOI: 10.2525/ecb.52.37 hOndO, k., sukhumPiniJ, P., kAkihArA, f., 2015: Flower color and pigments in yellow-flowered hybrid progeny raised from the interspecific cross Pelargonium quinquelobatum × white-flowered geraniums. Sci. Hortic. 195, 145-153. DOI: 10.1016/j.scienta.2015.09.014 hOrn, W., 1994: Interspecific crossability and inheritance in Pelargonium. Plant Breeding 113(1), 3-17. DOI: 10.1111/j.1439-0523.1994.tb00696.x JAmes, c.m., gibby, m., bArrett, J.A., 2004: Molecular studies in Pelar- gonium (Geraniaceae). A taxonomic appraisal of section Ciconium and the origin of the ‘Zonal’ and ‘Ivy-leaved’ cultivars. Plant Syst. and Evol. 243(3-4), 131-146. DOI: 10.1007/s00606-003-0074-2 kAkihArA, f., hOndO, k., kAtO, m., 2012: Production of interspecific hy- brids between Pelargonium crispum and P. rapaceum through ovule culture and their characteristics. Acta Hortic. 937, 697-702. DOI: 10.17660/ActaHortic.2012.937.84 kAtO, m., tOkumAsu, s., 1983: Characteristics of F1 hybrids produced by ovule-culture in ornamental Pelargonium. Acta Hortic. 131, 247-252. DOI: 10.17660/ActaHortic.1983.131.28 kubbA, A.J., tilney-bAssett, r.A.e., 1980: The occurrence and survival of twin embryos in Pelargonium ×hortorum Bailey. Ann. Bot. 46 (1), 113- 117. DOI: 10.1093/oxfordjournals.aob.a085887 kuligOWskA, k., lütken, h., müller, r., 2016: Towards development of new ornamental plants: status and progress in wide hybridization. Planta 244(1), 1-17. DOI: 10.1007/s00425-016-2493-7 lAfOn-PlAcette, c., köhler, C., 2015: Epigenetic mechanisms of postzy- gotic reproductive isolation in plants. Curr. Opin. Plant Biol. 23, 39-44. DOI: 10.1016/j.pbi.2014.10.006 metzlAff, m., börner, t., hAgemAnn, r., 1981: Variations of chloroplast DNAs in the genus Pelargonium and their biparental inheritance. Theor. Appl. Genet. 60 (1), 37-41. DOI: 10.1007/BF00275175 metzlAff, m., POhlheim, f., börner, t., hAgemAnn, r., 1982: Hybrid variegation in the genus Pelargonium. Curr. Genet. 5 (3), 245-249. DOI: 10.1007/BF00391813 murAshige, t., skOOg, f., 1962: A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol. Plantarum 15(3), 473-497. nOrtOn, J.d., 1966: Testing of plum pollen viability with tetrazolium salts. P. Am. Soc. Hortic. Sci. 89, 132-134. Olbricht, k., 2013: Klonzüchtung ausgewählter gärtnerischer Kulturen unter dem Aspekt der Nachhaltigkeit. Habilitationsschrift, Humboldt- Universität zu Berlin, 197 pages. PlAschil, s., budAhn, h., schrAder, O., Olbricht, k., WiedemAnn, m., hOfmAnn, C., 2015: Tetraploid male fertile Pelargonium crispum hybrids and their use in interspecific hybridization. Acta Hortic. 1087, 345-350. DOI: 10.17660/ActaHortic.2015.1087.45 PlAschil, s., kAlich, s., WiedemAnn, m., rietze, e., 2017: Evaluierung von tetraploiden S1-Populationen der Kreuzung Pelargonium ×domesti- cum × Pelargonium ×crispum, DGG Proceedings Vol. 7, No. 5, 1-5. DOI: 10.5288/dgg-pr-sp-2017 rieseberg, l.h., Willis, J.h., 2007: Plant speciation. Science 317(5840), 910-914. DOI: 10.1126/science.1137729 röschenbleck, J., Albers, f., müller, k., Weinl, s., kudlA, J., 2014: Phylogenetics, character evolution and a subgeneric revision of the genus Pelargonium (Geraniaceae). Phytotaxa 159(2), 31-76. DOI: 10.11646/phytotaxa.159.2.1 sAttler, m.c., cArvAlhO, c.r., clArindO, W.r., 2016: The polyploidy and its key role in plant breeding. Planta 243(2), 281-296. DOI: 10.1007/s00425-015-2450-x scemAmA, c., rAquin, c., 1990: An improved method for rescuing zygotic embryos of Pelargonium ×hortorum Bailey. J. Plant Physiol. 135(6), 763- 765. DOI: 10.1016/S0176-1617(11)80893-X Weihe, A., APitz, J., POhlheim, f., sAlinAs-hArtWig, A., börner, t., 2009: Biparental inheritance of plastidial and mitochondrial DNA and hybrid variegation in Pelargonium. Mol. Genet. Genomics 282 (6), 587-593. DOI: 10.1007/s00438-009-0488-9 Weng, m.l., ruhlmAn, t.A., gibby, m., JAnsen, r.k., 2012: Phylogeny, rate variation, and genome size evolution of Pelargonium (Geraniaceae). Mol. Phylogenet. Evol. 64(3), 654-670. DOI: 10.1093/molbev/mst257 WinkelmAnn, t., dOil, A., reinhArdt, s., eWAld, A., 2010: Embryo res- cue. In: Davey, M.R., Anthony, P. (eds.), Plant Cell Culture: Essential Methods, 79-95. John Wiley & Sons, London, UK. DOI: 10.1002/9780470686522.ch5 WOJtAniA, A., 2010: Effect of meta-topolin on in vitro propagation of Pelar- 56 R. Kamlah, I. Pinker, S. Plaschil, K. Olbricht gonium ×hortorum and Pelargonium ×hederaefolium cultivars. Acta Soc. Bot. Pol. 79(2), 101-106. DOI: 10.5586/asbp.2010.013 WOJtAniA, A., gAbryszeWskA, e., 2001: Effect of cytokinins and amino acids on multiplication of Pelargonium cultivars. Acta Soc. Bot. Pol. 70(3), 203-208. DOI: 10.5586/asbp.2001.026 yAnO, f., tOkumAsu, s., kAtO, m., 1975: The behaviour of pollen tubes and the developmental process of ovules in Pelargonium. Euphytica 24(1), 251- 259. DOI: 10.1007/BF00147194 yu, s.-n., 1985: Untersuchungen zur interspezifischen Kompatibilität und Biosystematik bei der Gattung Pelargonium. PhD thesis, Technische Uni- versität München, 270 pages. Address of the corresponding author: R. Kamlah, Lausitzer Platz 1, 10997 Berlin, Germany E-mail: rainer.kamlah@web.de © The Author(s) 2019. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creative- commons.org/licenses/by/4.0/deed.en).