Evaluation of the possibility of obtaining viable seeds from the cross-breeding Hippeastrum × chmielii Chm. with selected cultivars of Hippeastrum hybridum Hort.
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
INTRODUCTION
Hippeastrum sp. (commonly known as amaryllis in the global market) currently occupies the 11th position among cut flowers and foliage sold on the Dutch flower auctions (CBI Market Intelligence, 2016) and it's also a popular potted plant worldwide. The genus Hippeastrum is still not very well understood taxonomically (Tombolato et al., 2013), particularly that it was included in the genus Amaryllis till 14th International Botanical Congress in 1987. Hippeastrum is a native bulbous plant from America (Bryan, 2002), including 55–75 species distributed from Mexico to Argentina (Tombolato et al., 2013). Wild species – mainly from Brazil and Peru – gave origin to new hybrids and hundreds of them are now cultivated worldwide as Hippeastrum hybridum Hort. cultivars. According to Traub (1958), Hippeastrum obtained from the crossing of several selected species, specifically Hippeastrum vittatum with H. aulicum, H. reginae, H. leopoldii and H. pardinum. Breeding began with Hippeastrum × johnsonii (Amaryllis × johnsonii at that time), a species commonly regarded as the first hybrid form obtained in England in 1799, which was created by the crossing of H. vittatum (A. vittatum) with H. reginae (A. reginae) (Meerow, 2009; Okubo, 1993; van Dijk and Kurpershoek, 2002).
Then, for almost 200 years, the breeders had been searching for ideal forms and created new cultivars and new hybrid species and among them was the Hippeastrum × chmielii Chm. bred by Professor Henryk Chmiel in 1993 at the Department of Ornamental Plants, Warsaw University of Life Sciences (WULS) – SGGW, Poland. The selected clones of the inter-species hybrid (H. pratense (Poepp.) Baker (= Rhodophiala pratensis (Poepp.) Traub) × chosen cultivars of H. hybridum) were characterised by vigorous growth (formation of large green mass), no visible resting period and repeated flowering. However, the periods of late winter and early spring were observed as the most abundant flowering periods for this species (Chmiel, 2000; Chmiel and Ilczuk, 2002; Chmiel et al., 2002; Ilczuk, 2015). The flowers were characterised by new decorative values. They were smaller than those of the compact hybrids, which may give the reason for their classification under the Diamond Group or Colibri Group (Van Scheepen et al., 2007). The delicate and small flowers made the material easier to pack and transport. In floristry, the petite structure has also overcome the difficulty of arranging the large, massive and visually quite heavy flowers of the standard cultivars of H. hybridum.
According to the register reports of the Dutch Royal General Bulb Growers’ Association (KAVB), the biggest number of cultivars were registered for three groups with single flowers – Galaxy, Diamond and Colibri during the years 2016–2019. The reason for this may be the constantly growing demand for this type of flower (Bodegom and van Scheepen, 2017, 2018, 2019; Bodegom et al., 2020).
The bulbs of the H. × chmielii have a smaller circumference and are more flattened than those of the compact cultivars of H. hybridum, which makes it possible to plant more bulbs per square meter. Production for cut flowers is therefore more profitable (Chmiel and Ilczuk, 2002), as more flowers can be obtained from a smaller growing area. A high coefficient of vegetative propagation was observed for all clones of H. × chmielii. Depending on the circumference of the storage organ, the plants produced a large number of daughter bulbs, which make this species propagate easily (Chmiel and Mynett, 1997; Chmiel et al., 1998; Chmiel and Szymański, 2001; Chmiel et al., 2002; Ilczuk, 2005).
No one had so far tried to propagate H. × chmielii by seeds as well as checked the possibility of using the clones for further breeding. This study aimed to test the ability of two H. × chmielii clones – No. 6 and No. 18 to reproduce generatively by pollinating by three cultivars of H. hybridum – ‘Gervase’, ‘Rio Negro’ and ‘Royal Velvet’.
MATERIALS AND METHODS
Plant material
The research was conducted on Hippeastrum × chmielii Chm. clones No. 6 and No. 18 (Figure 1) and three cultivars of H. hybridum – ‘Gervase’, ‘Rio Negro’ and ‘Royal Velvet’ (Figure 2). The cultivars ‘Gervase’ and ‘Royal Velvet’ are of Dutch origin and they belong to the Galaxy group. The diameter of their flowers is >16 cm. The first cultivar has pink-red flowers while the second one is a dark burgundy. ‘Rio Negro’ is a cultivar from the Spider group whose flowers are similar to the shape of a spider. It also has an interesting flower colour and anthocyanin pigmentation of the inflorescence stem. Two clones of H. × chmielii Chm. were selected for the study. Clone No. 6 has a dark red colour, and clone No. 18 has a brick orange colour. The first one was typical for the group of red coloured clones, and the second one was typical for the group of orange coloured clones. Both clones have flowers with a diameter of 12–16 cm and are characterised by the production of a large number of daughter bulbs.
Figure 1
Clones of H. × chmielii – No. 6 (left) and No. 18 (right).
Figure 2
Cultivars of H. hybridum tested – ‘Gervase’, ‘Rio Negro’ and ‘Royal Velvet’ (from left to right).
The bulbs of H. × chmielii Chm. were collected from the Department of Ornamental Plants of Warsaw University of Life Sciences – SGGW, Poland (WULS), where this hybrid was bred by Henryk Chmiel. The bulbs of cultivars originated from the commercial stocks.
The bulbs were cultivated in the greenhouse in plastic boxes 60 × 40 × 22 cm in a medium consisting of peat substrate and perlite at 3:1, supplemented with 10 g per box of complete fertilizer for bulbous plants (Agrecol Sp. z o.o., Poland) (7% N, 24% K2O, 13% CaO) and adjusted to pH = 6.5. The plants were grown from the beginning of April to the end of June 2018 at 24–30°C (on some hot days the temperature rose to 32°C) at natural photoperiod (day length of 13–17 h) without light supplementation. All experiments were conducted from April to July 2018 at WULS.
Preparing flowers for pollination
First, stamens were mechanically removed by laboratory tweezers from the fully coloured but unopened flower buds for preparing the flowers of H. × chmielii for pollination. The stigmas were protected by aluminium foil against uncontrolled or self-pollination.
Evaluation of pollination and fertilisation ability of flowers
The pollination ability of flowers of H. × chmielii clones was checked by staining the stigmas and ovules with alizarin red on the microscopic slides for 5 min. Three repetitions of microscopic slides for two flowers of each clone were prepared. The solution was prepared according to McGee-Russel (1958) by dissolving 2% alizarin red (Sigma Aldrich) in a solution of ammonia adjusted to pH = 4.2. In the case of Hippeastrum cultivars, indirect pollen viability was assessed by staining with 1% acetocarmine (10 g of carmine dye dissolved in 1 L of 45% glacial acetic acid). From the prepared microscopic slides and after 5–10 min of staining, the number of stained and all visible pollen grains were counted under an AX Provis (Olympus) light microscope (Olympus Optical Co. Ltd., Japan) at a magnification of 100×. Properly developed viable pollen grains were coloured carmine while the non-viable ones remain non-stained. The percentage of visible pollen grains for each cultivar was calculated based on their numbers observed in three fields of view. The germination capacity of pollen grains was evaluated by germination of 150 grains (in three repetitions) on the 50 mL agar solidified medium on a Petri dish lined with filter paper soaked with distilled water, containing 10% sucrose, 100 ppm H3BO3, 300 ppm Ca (NO3)2·4H2O, 200 ppm MgSO4·7H2O and 100 ppm KNO3 (Brewbaker and Kwack, 1963). The ability of pollen grains to develop pollen tubes was observed after 24 h under an AX Provis (Olympus) light microscope at a magnification of 100×.
Pollination and fertilisation
In May 2018, selected flowers of both clones of H. × chmielii were cross-pollinated by pollens taken from three cultivars. The following crosses were made: H. × chmielii clone 18 ♀ × H. hybridum ‘Gervase’ ♂ (10 crosses), H. × chmielii clone 18 ♀ × H. hybridum ‘Royal Velvet’ ♂ (15 crosses), H. × chmielii clone 18 ♀ × H. hybridum ‘Rio Negro’ ♂ (17 crosses), H. × chmielii clone 6 ♀ × H. hybridum ‘Gervase’ ♂ (10 crosses), H. × chmielii clone 6 ♀ × H. hybridum ‘Royal Velvet’ ♂ (15 crosses), H. × chmielii clone 6 ♀ × H. hybridum ‘Rio Negro’ ♂ (10 crosses).
Further, callose staining was performed 24 h after pollination to determine the ability of the pollen tube in penetrating the ovary under a microscope. The collected test material was placed in 70% ethanol for 24 h, next macerated for 1.5 h in 10M NaOH in an incubator (40°C) and followed by three washes in distilled water for 20 min. Then, the material was dyed for 1 h in 0.05% aniline blue solution in 0.067M phosphate buffer in total darkness. The microscopic slides (in three repetitions for each cross combination) were prepared with 50% glycerol, the fluorescence of callose was observed in AX Provis (Olympus) microscope (Olympus Optical Co. Ltd., Japan) using NU filters (excitation filter BP360-370 nm, barrier filter BA420 nm, dichroic mirror DM400 nm) and then photos were taken using Olympus U-CMAD 3, Japan, by QuickPHOTO Pro.
One week after pollination, mesh insulators were applied to the expanding ovaries to prevent them from cracking uncontrolled and the seeds from falling to the ground.
Seed germination
In mid-June 2018, entire seed bags were collected. The seeds from each bag (from 10 pcs. to 17 pcs. depending on crossing) were counted and divided into two parts for two germination methods testing. The first one is covered in Petri dishes which were lined with filter paper soaked in distilled water (ISTA 2011) and the second one in open glass jars of 240 mL volume were filled with 100 mL of distilled water (according to Treder – personal communication). The germination was taken place in a cultivation room at 21°C with 16 h of cold white fluorescent light at 25 μM · m−2 · s−1, and 60% RH all the time. The conditions were the same for each seed germinating method. The number and percentage of germinated seeds were assessed after 28 days (according to ISTA 2011).
Results of pollen viability and seed germinating were analysed on transformed data using Bliss transformation by analysis of variance and Multiple Range Test at the 5% significance level, using Statgraphics Plus 4.1.
RESULTS
Evaluation of pollination and fertilisation ability of flowers
The pollination capacity of the flowers of H. × chmielii clones No. 6 and No. 18 was determined by staining the stigmas and ovules with alizarine red. The red colouring in all the prepared microscopic slides indicates the accumulation of Ca2+ ions in large quantities, which expresses their receptivity (Figures 3A,3B and 4A,4B). Pollen grain viability of H. hybridum cultivars assessed by acetocarmine staining method was found and it was at the level of 83.0% for ‘Gervase’, for ‘Royal Velvet’ it is 82.7% and for ‘Rio Negro’ it is 66.4% (Table 1) (Figure 5A–5C). Based on the microscopic observations, the pollen grains of all the examined cultivars, placed on an agar solidified medium, were characterised by their capacity to produce pollen tubes (Figure 6A,6B).
Figure 3
The receptive stigma of Hippeastrum × chmielii stained by alizarine red – clone No. 6 (A) and clone No. 18 (B).
Figure 4
The receptive ovule of Hippeastrum × chmielii stained by alizarine red – clone No. 6 (A) and clone No. 18 (B).
Percentage of viable/stainable pollen grains of cultivars H. hybridum – ‘Gervase’; ‘Royal Velvet’; ‘Rio Negro’.
Means ± standard deviation in a column followed by the same letter does not differ significantly at p = 0.05.
Figure 5
Pollen grains stained by acetocarmine – H. hybridum ‘Gervase’ (A), H. hybridum ‘Rio Negro’ (B) and H. hybridum ‘Royal Velvet’ (C). White arrows show viable pollen grains; black arrows show non-stainable/non-viable grains.
Figure 6
Sprouting pollen tubes on the germination medium – H. hybridum ‘Rio Negro’ (A) and H. hybridum ‘Royal Velvet’ (B).
Pollination and fertilisation
Microscopic observations under a fluorescence microscope made 24 h after the application of pollen grains of three tested cultivars to the stigmas of both H. × chmielii clones confirm the presence of pollen tubes on fragments of the stigmas (Figure 7).
Figure 7
Germinating pollen grains of H. hybridum ‘Rio Negro’ on the stigma (A) and pollen tubes of ‘Rio Negro’ penetrated the ovule (B).
Seed germination
Pollinated flowers formed a different number of seeds. A detailed summary of the number of seeds obtained from individual crossings is shown in Table 2. From one flower a minimum of 22 and a maximum of 94 seeds were collected, on average for the type of cross-breeding from 44 seeds to 71 seeds. The least number of seeds was collected from the flowers of H. × chmielii clone No. 18 × ‘Rio Negro’ – 305, and the largest number from the flowers of H. × chmielii clone No. 6 pollinated with pollen from ‘Royal Velvet’ – 993. In the case of crossing of H. × chmielii clone No. 6 with H. hybridum ‘Rio Negro’, all seed bags became yellow and died on the inflorescence shoots and so no seeds were obtained. Totally, from all 72 crossings in 6 types, 3,043 seeds were obtained.
The number of seeds obtained from individual crossings of H. × chmielii with three cultivars.
Crossing
Number of flowers
Number of collected seeds from one flower (min – max)
Average number of collected seeds per flower ± SD
Total number of seeds
H. × chmielii 18 × ‘Gervase’
10
31–76
57.7 ± 15.53
577
H. × chmielii 18 × ‘Royal Velvet’
15
24–86
59.1 ± 17.99
828
H. × chmielii 18 × ‘Rio Negro’
17
22–68
43.6 ± 17.68
305
H. × chmielii 6 × ‘Gervase’
10
25–94
48.6 ± 22.17
340
H. × chmielii 6 × ‘Royal Velvet’
15
53–83
70.9 ± 9.44
993
H. × chmielii 6 × ‘Rio Negro’
10
0
0
0
Total number of seeds from all crossings
3,043
The first germinating seeds appeared after 9 days in jars with water. The percentage of germinated seeds calculated after 28 days varied between 48.3% and 77.9%, depending on the type of cross-breeding (Table 3). No effect of the germination method was noticed for most of the crossings. Only in the case of seeds originated from cross-breeding of H. × chmielii 18 × ‘Gervase’, more seeds germinated in jars with water than on Petri dishes.
Percentage of germinated seeds of H. × chmielii pollinated by pollen of three cultivars depending on the germination method.
Means ± standard deviation in a row followed by the same letter does not differ significantly at p = 0.05.
DISCUSSION
Cross-pollination is the only method of obtaining seeds capable of germination in H. hybridum (Wóycicki, 1966; Almeida et al., 2019). Additionally, the flowers of Hippeastrum are proterandrous, which means that the anthers mature earlier than the stigmas (Almeida et al., 2019; Szlachetka, 2000). It confirms the necessity of artificial pollination. However, in the case of H. × chmielii, self-pollination can occur (Szlachetka, 2000). In practice, a common problem in the process of successful pollination and fertilisation is the occurrence of various types of cross-pollination barriers as reported by many authors (Bomblies and Weigel, 2007; Chen and Lin 2016; Kuligowska et al., 2015 – for Kalanchoe genus). In the present research, during the crossbreeding of the H. × chmielii clone No. 6 with H. hybridum ‘Rio Negro’, the seed bags died out, which may indicate the occurrence of prezygotic barriers.
In our study, high receptivity of the stigmas of both H. × chmielii clones was observed in alizarin red staining. Observed red colouring in all microscopic slides indicates the accumulation of Ca2+ ions in large quantities, which is related to the enzymatic decomposition of pectins (Brewbaker and Kwack, 1963; Reger et al., 1992). It can be confirmed that pectins are also present in the structures of the hippeastrum pistil, as described by Chudzik (2002) for other Amaryllidaceae species. The content of pectins is essential for overcoming the problems due to the distance from the stigma of the pistil to the ovule by pollen tubes. The distance from the stigma of the pistil to the ovules is different in different plants and ranges from a few millimetres in Zantedeschia to several centimetres in Colchicum (Wóycicki, 1966). The distance for the Hippeastrum was determined as ca. 15 cm by Traub (1958). The enzymes found in the pollen tube decompose pectins into free calcium ions, which are the chemo-attractants that direct the pollen tubes towards the receptive ovules (Malho et al., 2000). During the evaluation of the germination capacity of the pollen grains of tested cultivars on a medium that imitates the conditions of the stigma, the pollen tubes spread out in each direction without any visible reference point. According to Wóycicki (1966), the pollen tubes growing on artificial substrates are much longer than the pistils of some species. In the case of Hippeastrum, this was not confirmed. The absence of arabinogalactan proteins in the medium, which probably play a signalling role for pollen tube growth towards the mature ovary sac (Schultz et al., 1998; Śnieżko and Chudzik, 2003), was responsible for the lack of directed pollen tube growth under artificial conditions.
In our study, the viability of Hippeastrum pollen grains was evaluated after the opening of the anthers by acetocarmine staining. Similar methods were used in the studies conducted by Weryszko-Chmielewska and Chwil (2006) for Taraxacum officinale, Szklanowska (1992) for selected trees and shrubs of the Rosaceae family and by Chwil (2006) for Narcissus sp., which were the most related to the Hippeastrum. On the other hand, it is known, that the viability/stainability of pollen grains is dependent on the method of staining (Słomka et al., 2010). In all cultivars of H. hybridum in our research, the pollen viability did not exceed 83%, but the lowest was 66.4%. In Narcissus (Amaryllidaceae), Chwil (2006) determined the pollen grains viability up to 92% in cultivar ‘Hardy’ and only 22% in cultivar ‘The Sun’. The viability of pollen grains ranged from 62% to 77% (Chwil, 2006) for the remaining cultivars of Narcissus and was comparable to the level of viability observed for three Hippeastrum cultivars tested by us and at the level of 60–80% for nine Hippeastrum hybrids reported by Khaleel et al. (1991). Therefore, the pollen viability could be a cultivar feature. Many cultivars obtained by crossing show reduced pollen viability or even sterility. This was proven, among others, for nine lily genotypes in a study by He et al. (2017), where the percentage of germinating pollen 1 day after anthesis was 81% for Lilium sulphureum, 73.4–77.1% for three hybrid cultivars and only 17.8% for cultivar ‘Tiny Padhye’. The pollen of the cultivar ‘Jinghe’ did not germinate at all. For narcissus, Sanders (2014) reported that the number of germinated pollen grains from the samples collected was 400 for the cultivar ‘Gloriosus’ but only 20, 23 and 34 for ‘Magic Step’, ‘Silver Bells’ and ‘Problem Child’, respectively.
The seeds obtained in the experiment ripened for 3 weeks, after which the bags were cracked and the seeds started to fall down. A single seed bag contained a minimum of 22 seeds to a maximum of 94 seeds, and there is a literature report of seeds range of 40–80 in a single seed bag for H. hybridum (Kazimierczak, 1992; Okubo, 1993; Szlachetka, 2000). The first germinating seeds appeared after 9 days in jars with water, which meats ISTA standards (ISTA, 2011) of 7–10 days for germinating of H. hybridum. The percentage of germinated seeds after 28 days varied between 48.4% and 77.9% and was affected by the type of cross-breeding. Similar studies carried out by Arayakitcharoenchai (2012) on seeds, obtained from crossbreeding of hybrid forms, showed that the first seedlings appeared after 14 days. The germination capacity of seeds was evaluated after 30 days and the total percentage of germination was between 62.3% and 98.6%. The maximum viability of pollen in our trials did not exceed 78%. It could be negatively affected by the high temperature in the greenhouse during the flowering period of these plants. This is confirmed by Doijode (2001), which states that seed viability of Hippeastrum is lost completely within 3 months of storage at 25–35°C.
The above results concerning the propagation of H. × chmielii clone No. 6 and clone No. 18 by seeds as well as further research and selection work on the obtained seedling population may contribute to obtaining new, interesting cultivars of this species in the near future. An additional aspect can be the knowledge of physiological aspects of flower development, which is useful in practice and applied for breeding programs. This will help in the selection of forms to be crossed and in choosing the right time for pollination. The obtained results concerning seed germination are important for breeding and horticultural practices.
CONCLUSIONS
The results indicate the possibility of propagation by seeds of H. × chmielii, which is confirmed by the assessment of the receptivity of the stigmas and ovules of both tested clones (No. 6 and No. 18).
As a result of pollination of the flowers of H. × chmielii with viable pollen of three selected cultivars of H. hybridum, viable seeds were obtained in most of the studied combinations, which confirms the possibility of using the H. × chmielii as a maternal form in further breeding.
Seeds obtained by successful cross-breeding of H. × chmielii with H. hybridum were characterised by a shorter (by 1 week) ripening period for the standards given in the literature for the hybrid cultivars.
Both tested germination methods were positive, but germination in water in jars proved to be more effective for one of the cross-breeding (H. × chmielii clone No. 18 × H. hybridum ‘Gervase’) compared to germination on Petri dishes with filter paper soaked in distilled water.
