Morphological and genetic diversity of breeding clones derived from Hippeastrum × chmielii

Abstract

Hippeastrum is among the most widely cultivated geophytes, utilized for both cut flowers and potted plants. The range of cultivars is expanding, with an average of 33 new cultivars registered annually in the international register maintained by the Royal General Bulb Growers’ Association in the Netherlands. This study focused on the morphological and genetic diversity of 15 Polish breeding clones derived from H. × chmielii and H. hybridum. The study encompassed phenotypic evaluation, assessment of the number and length of stomata, and the use of RAPD and ISSR markers across all genotypes examined. Phenotypic analysis, conducted in accordance with the descriptors of The International Union for the Protection of New Varieties of Plants (UPOV) and the RHS Colour Charts, facilitated the identification of three promising breeding clones. These clones (0037-13 with flowers of the greyed-purple colour group, 185A; 0021-10 with orange-red flowers, 34A; and 0023-11 with orange-red, 35B) produced a greater number of larger flowers than the maternal forms of H. × chmielii. The population exhibited significant variation in the number and length of stomata, with an inverse correlation between stomatal length and number. The use of RAPD and ISSR markers showed high genetic similarity between the initial forms and breeding clones in the range of 74–85%. The three selected Hippeastrum clones have the potential to become new cultivars originating from H. × chmielii.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-35035-1.

Introduction

The name of the genus Hippeastrum is a combination of two Greek words: hippeus meaning ‘cavalier’ and astron, which translates as ‘star’^1,2. However, plants often take on other names due to the variety of shapes, colours, and sizes of the perianth. In Persian, hipeastrum is called—Nasrin, which means the wild rose. “Butterfly Amaryllis”, “Green Amaryllis”, “Lily of the Palace” or “Mexican Lily” are some examples of the names by which these plants are known in different countries^3,4. Breeding programs for many ornamental plants, as well as the genus Hippeastrum, focus mainly on improving ornamental flower values, such as colour, perianth size, and shape, but also on compact forms and disease and pest resistance^5,6. Owing to advances in breeding and the multiplicity of cultivars, all cultivars were divided into nine groups of horticultural classification⁷. An analysis of the 2018–2023 cultivar registry data maintained in the Netherlands by the Royal General Bulb Growers’ Association (Koninklijke Algemeene Vereniging voor Bloembollencultuur, KAVB), shows that 196 new cultivars were registered (Fig. 1).

Fig. 1.

The number of new Hippeastrum cultivars registered in each cultivar group, in the KAVB international register in 2018–2023.

In Hippeastrum breeding, great emphasis has been placed on obtaining cultivars with large, colourful flowers^8,9. Analysis of the KAVB data showed that the highest number of registered cultivars (76) belonged to the Galaxy group (Fig. 1). In this group, the flowers are more than 16 cm in diameter. This was followed by 42 cultivars in the Diamond group, characterized by medium-sized, single flowers of 12–16 cm in diameter, and 7 cultivars with small, single flowers of less than 12 cm in diameter, belonging to the Colibri group. The same number of cultivars were also registered in the Butterfly group, where the flowers resemble the shape of a butterfly. Not only cultivars with single flowers, but also full-flower cultivars are popular. In the Double Galaxy group, in which representatives have large, full flowers, 55 cultivars have been registered^10–15.

It is also worth mentioning that the H. × chmielii hybrid was bred at the Warsaw University of Life Sciences (WULS), Poland, in 1993. Clones obtained by crossing H. pratense × selected cultivars of H. hybridum were characterized by vigorous growth, no visible dormant period, and repeated flowering. In addition, each H. × chmielii clone was characterized by a high vegetative reproduction rate, with the plant producing a large number of daughter bulbs. In the case of flowers, the petals can be red, red–orange, brick-orange, or bicoloured with a starry throat, and the inflorescences remain fully decorative for approximately 13–14 days¹⁶. Despite obtaining interesting characteristics that distinguished them from the rest of the species, the smaller flowers were not sufficiently attractive to be commercialized as cultivars. Therefore, 25 years later, the possibility of further generative propagation of two clones of this hybrid was tested, and after positive verification¹⁷, they were used for further creative breeding in crosses with three cultivars of H. hybridum. The completely new breeding clones obtained from further crossbreeding of H. × chmielii and their thorough genetic and morphological evaluations may provide material for new cultivars to be introduced to the market. A detailed assessment, especially morphological assessment, is extremely valuable for estimating the potential of a breeding clone to become a cultivar, and is helpful in the registration process.

The International Union for the Protection of New Varieties of Plants (UPOV) was conceived under the International Convention for the Protection of New Varieties of Plants adopted in Paris in 1961. UPOV’s mission is to provide and promote an effective system of plant variety protection that aims to encourage the development of new plant cultivars for the benefit of society¹⁸. The 2001 UPOV descriptor for the genus Hippeastrum (Amaryllis) is a guideline for conducting tests on the distinctiveness, uniformity, and persistence of new cultivars¹⁸.

Stomata are key structures that regulate leaf gas exchange by controlling diffusion conductance, thereby influencing water loss and carbon assimilation. Their function is essential for maintaining plant water balance^19–21.

Morphological markers are among the first traits that provide information on genetic diversity and can be used in breeding processes^22,23. It is on the basis of these that breeders are able to determine the value of seedlings in the juvenile stage and carry out initial selection. The first criterion for selecting representative hybrids is based on the morphological and agronomic traits of interest²⁴. These include leaf shape, growth strength, and flower colour. Today, owing to their high epigenetic variability, they have been applied in identifying breeding cultivars²⁵.

Molecular markers provide information regarding an individual’s genetic structure or are used to analyze the intraspecific distribution of variability²⁶. The use of DNA markers enables the assessment of the genetic diversity in breeding materials. Additionally, they facilitate the selection of suitable individuals for further crossing, hybrid identification, and subsequent selection²⁷. The application of such techniques allows the detection of polymorphisms by differentiating genotypes at the DNA level^6,25,28. A substantial number of molecular markers have been developed, permitting their selection based on intended applications. Among these, two notable types are Random Amplified Polymorphic DNA (RAPD markers) and Inter Simple Sequence Repeats (ISSR markers)²⁸.

Monocotyledonous and dicotyledonous plants differ in the structure of their leaves, especially in the distribution of their stomata. In dicotyledonous plants, the stomata are located mainly on the lower side of the leaves; therefore, they are not exposed to excessive light. In monocotyledonous plants, on the other hand, the stomata are located on both sides of the leaf, in addition, their number on both sides is the same and their distribution is even. This is due to the arrangement of the leaves in relation to light, since its effect on both sides of the leaf is the same^29–31.

This study aimed to evaluate the genetic and morphological differentiation of new breeding clones in relation to the maternal forms of clones 6 and 18 of H. × chmielii. It is necessary to select the most valuable clones in visual and production terms, representing potentially new Hipeastrum cultivars derived from the unique species H. × chmielii.

Results

Phenotype analysis

Analysis of the selected biometric parameters for the parental forms and the three selected clones obtained from the crossing of clone 6 of H. × chmielii with the cultivar ‘Gervase’ of H. hybridum showed that the leaves of the breeding clone 0033-10 were wider than those of the maternal form (Table 1, supplementary Table S1). As for the peduncle (inflorescence stem) length, the parent forms were longer than breeding clones 0033-10 and 0037-13. Clone 0040-31 had longer shoots than clone 0033-10 did. The cultivar ‘Gervase’ and clone 0037-13 were characterized by a higher number of flowers than clone 6. ‘Gervase’ and clones 0033-10, 0037-13 and 0040-31 had a longer perianth than the maternal form. The paternal form and clone 0037-13 had a wider perianth than that of the other genotypes. In addition, clones 0033-10 and 0040-31 had wider perianths than clone 6. The stamen colour of the four tested gnotypes was within the red group; only the stamens of clone 0040-31 had the colour (155C) of the CC RHS white group. As for anther colour, only clone 6 had a reddish colour (46A), whereas in the other genotypes studied, the anther colour before opening was purple.

Table 1.

Selected morphological characteristics of parental and breeding clones from crosses between clone 6 of H. × chmielii with H. hybridum ‘Gervase’.

	H. × chmielii clone 6♀	H. hybridum ’Gervase’♂	0033-10	0037-13	0040-31
Leaf: width [cm]	4.04 ± 0.29 a*	4.70 ± 0.44 ab	5.26 ± 0.29 b	4.68 ± 0.20 ab	4.62 ± 0.41 ab
Stem (Peduncle): length [cm]	53.35 ± 3.33 c	53.46 ± 1.95 c	42.14 ± 2.29 a	46.74 ± 1.56 ab	49.12 ± 3.07 bc
Inflorescence: number of flowers [pcs.]	3.2 ± 0.98 a	4.6 ± 0.80 b	4.0 ± 0.00 ab	5.0 ± 0.63 b	4.0 ± 0.00 ab
Flower: maximum length of perianth [cm]	10.50 ± 0.53 a	13.20 ± 1.81 b	13.88 ± 0.35 b	12.50 ± 0.58 b	13.96 ± 0.43 b
Flower: maximum width of perianth [cm]	12.34 ± 0.62 a	17.00 ± 0.34 c	15.22 ± 0.30 b	17.06 ± 0.33 c	14.40 ± 0.66 b
Flower: main colour of inner side [CC RHS]	45A	51B	33A	185A	45B
Stamens: colour [CC RHS]	45A	47A	45A	51B	155C
Anthers: colour (just before dehiscence) [CC RHS]	Reddish (46A)	Purplish (183D)	Purplish (187D)	Purplish (183D)	Purplish (184B)

*Means ± standard deviation (SD) in a verse followed by the same letter does not differ significantly at α = 0.05 (Tukey test); averages compared across traits.

Three genotypes selected from crossing H. × chmielii clone 6 with H. hybridum ‘Royal Velvet’ had higher leaf widths in the breeding clone 0072-17, compared than in the parental forms (Table 2, supplementary Table S2). Longer stems were possessed by the parent forms than clones 0050-15 and 0050-16. Also clone 0072-17 had longer inflorescence stems than clone 0050-16 did. A higher number of flowers was obtained for clone 0072-17 compared for clone 0050-16 and its parent form. For maximum perianth length, a higher value was achieved by clone 0072-17 in relation to the other genotypes. Clone 0050-15 had a longer perianth than the parental forms and clone 0050-16. ‘Royal Velvet’ had a longer perianth in relation to the maternal form and had a wider perianth than all other genotypes. The stamen colour in all studied Hippeastrum genotypes was within the red group. As for the colour of the anthers before opening, the initial forms—clone 6 and cultivar ‘Royal Velvet’ had a reddish colour, while clones 0050-15, 0050-16 and 0072-17 had a purplish colour.

Table 2.

Selected morphological characteristics of parental and breeding clones from crosses between clone 6 of H. × chmielii with H. hybridum ’Royal Velvet’.

	H. × chmielii clone 6♀	H. hybridum ‘Royal Velvet’♂	0050-15	0050-16	0072-17
Leaf: width [cm]	4,04 ± 0.29 a*	4.04 ± 0.33 a	4.44 ± 0.37 ab	4.64 ± 0.37 ab	4.88 ± 0.17 b
Stem (Peduncle): length [cm]	53.35 ± 3.33 c	56.64 ± 3.71 c	45.46 ± 1.78 ab	42.66 ± 1.37 a	51.24 ± 3.73 bc
Inflorescence: number of flowers [pcs.]	3.2 ± 0.98 a	4.2 ± 0.40 ab	4.0 ± 0.00 ab	3.2 ± 0.75 a	4.8 ± 0.75 b
Flower: maximum length of perianth [cm]	10.50 ± 0.53 a	12.16 ± 0.45 b	13.36 ± 0.55 c	11.54 ± 0.38 ab	14.82 ± 0.80 d
Flower: maximum width of perianth [cm]	12.34 ± 0.62 a	17.50 ± 0.32 e	14.06 ± 0.44 b	15.06 ± 0.36 c	16.02 ± 0.43 d
Flower: main colour of inner side [CC RHS]	45A	46A	46B	35A	45A
Stamens: colour [CC RHS]	45A	46A	45A/49B	50B	46A
Anthers: colour (just before dehiscence) [CC RHS]	Reddish (46A)	Reddish (46A)	Purplish (187D)	Purplish (186A)	Purplish (187B)

*Means ± standard deviation (SD) in a verse followed by the same letter does not differ significantly at α = 0.05 (Tukey test); averages compared across traits.

Analysis of biometric parameters of crossed genotypes of clone 18 of H. × chmielii with the cultivar ‘Gervase’ and the three selected clones from this crossing showed no differences for leaf width (Table 3, supplementary Table S3). In the case of inflorescence stem length, a higher result was obtained for the mother form and breeding clone 0003-13, compared for clone 0004-05. Cultivar ‘Gervase’ had more and longer flowers than clone 18. Both cultivar ‘Gervase’ and clone 0004-04 had a wider perianth than clone 0003-13 and the mother form. In addition, clones 0003-13 and 0004-05 had wider perianths than clone 18. The stamen colour of clone 18 (32A) was a colour from the orange-leaved colour group. The other genotypes have colours from the red group. The stamen colour before opening in clone 18 and breeding clone 0004-05 was reddish, whereas in the other three genotypes it was purplish.

Table 3.

Selected morphological characteristics of parental and breeding clones from crosses between clone 18 of H. × chmielii with H. hybridum ’Gervase’.

	H. × chmielii clone 18♀	H. hybridum ’Gervase’♂	0003-13	0004-04	0004-05
Leaf: width [cm]	4.54 ± 0.36 a*	4.70 ± 0.44 a	4.60 ± 0.40 a	4.48 ± 0.29 a	4.94 ± 0.44 a
Stem (Peduncle): length [cm]	57.46 ± 2.93 a	53.46 ± 1.95 ab	56.94 ± 2.10 b	53.62 ± 1.56 ab	51.70 ± 1.72 a
Inflorescence: number of flowers [pcs.]	3.4 ± 0.49 a	4.6 ± 0.80 b	4.0 ± 0.00 ab	4.0 ± 0.00 ab	4.0 ± 0.00 ab
Flower: maximum length of perianth [cm]	10.70 ± 0.92 a	13.20 ± 1.81 b	12.04 ± 0.65 ab	12.10 ± 0.46 ab	11.80 ± 0.80 ab
Flower: maximum width of perianth [cm]	13.26 ± 0.93 a	17.00 ± 0.34 c	15.26 ± 0.51 b	16.68 ± 0.61 c	16.14 ± 0.44 bc
Flower: main colour of inner side [CC RHS]	32A	51B	53D	52A	38A
Stamens: colour [CC RHS]	32A	47A	46A	53C	39B
Anthers: colour (just before dehiscence) [CC RHS]	Reddish (48C)	Purplish (183D)	Purplish (183C)	Purplish (187D)	Reddish (47A)

*Means ± standard deviation (SD) in a verse followed by the same letter does not differ significantly at α = 0.05 (Tukey test); averages compared across traits.

Parameter analysis for the cross between clone 18 of H. × chmielii with the cultivar ‘Rio Negro’ and three selected clones obtained from this crossing showed that the leaves of clone 18 were wider than those of the cultivar ‘Rio Negro’ and two clones 0021-10 and 0062-10 (Table 4, Supplementary Table S4). The cultivar ‘Rio Negro’ had wider leaves than clone 0062-10 did. Longer stems were obtained for clone 0021-10 compared for the parental forms and clone 0062-10. Shorter stems compared to all the genotypes tested were observed for the cultivar ‘Rio Negro’. More flowers were recorded for clone 0021-10 relative for the parental form and clones 0023-11 and 0062-10. In addition, the cultivar ‘Rio Negro’ had more flowers than clone 18. A longer perianth was observed in the cultivar ‘Rio Negro’ and clone 0023-11 relative to clone 18 and clone 0021-10. In addition, a longer perianth was recorded in clone 0062-10 than in the parental form. A wider perianth was observed for cultivar ‘Rio Negro’ and clones 0023-11 and 0062-10 than for clone 18 and clone 0021-10, which at the same time had a wider perianth than the mother form. The colour of the stamens of clone 18 and one of the breeding clones 0023-11 was within the orange-red group of the CC RHS. The red group was represented successively by the stamen colour of clone 0021-10 (43B), the yellow-green group of clone 0062-10 (150D) and the gray-purple group of cultivar ‘Rio Negro’ (185C). As for anther colour before opening, reddish ones were possessed by the maternal form and breeding clone 0062-10, while the paternal form and clones 0021-10 and 0023-11—were purplish.

Table 4.

Selected morphological characteristics of parental and breeding clones from crosses between clone 18 of H. × chmielii with H. hybridum ’Rio Negro’.

	H. × chmielii clone 18♀	H. hybridum ’Rio Negro’♂	0021-10	0023-11	0062-10
Leaf: width [cm]	4.54 ± 0.36 c*	3.86 ± 0.21 b	3.80 ± 0.30 ab	4.18 ± 0.13 bc	3.24 ± 0.30 a
Stem (Peduncle): length [cm]	57.46 ± 2.93 b	49.75 ± 3.38 a	64.04 ± 2.16 c	58.48 ± 2.01 bc	57.84 ± 2.87 b
Inflorescence: number of flowers [pcs.]	3.4 ± 0.49 a	4.6 ± 0.49 bc	5.0 ± 0.63 c	4.0 ± 0.00 ab	4.0 ± 0.00 ab
Flower: maximum length of perianth [cm]	10.70 ± 0.92 a	12.38 ± 0.72 c	10.80 ± 0.58 ab	12.36 ± 0.47 c	12.22 ± 072. bc
Flower: maximum width of perianth [cm]	13.26 ± 0.93 a	16.86 ± 0.36 c	14.54 ± 0.76 b	15.84 ± 0.27 c	16.96 ± 0.37 c
Flower: main colour of inner side [CC RHS]	32A	149D	34A	35B	39B
Stamens: colour [CC RHS]	32A	185C	43B	35B	150D
Anthers: colour (just before dehiscence) [CC RHS]	Reddish (48C)	Purplish (183D)	Purplish (183C)	Purplish (184B)	Reddish (53B)

*Means ± standard deviation (SD) in a verse followed by the same letter does not differ significantly at α = 0.05 (Tukey test); averages compared across traits.

No differences in leaf width were noted for H. × chmielii clone 18, cultivar ‘Royal Velvet’, and the three genotypes selected from this crossing (Table 5, Supplementary Table S5). Longer inflorescence stems were recorded for clone 0017-01 compared for the parental forms and clone 0075-25. In addition, clone 0075-25 was characterized by a larger number of flowers than the breeding clone 0017-01 and the parental form. A longer perianth was recorded for clones 0015-03 and 0017-01, compared for the parental forms and clone 0075-25. The cultivar ‘Royal Velvet’ and clones 0015-03 and 0017-01 had a wider perianth than clones 18 and clone 0075-25. The red group is represented by the stamens of four genotypes (cultivar ‘Royal Velvet’ and clones 0015-03, 0017-01, 0075-25), and only the stamens of the mother form of clone 18 have a colour from the orange-red group (32A). The colour of the anthers before opening differentiated the study population into two groups. The reddish group is represented by anthers of the initial forms—clone 18 and the cultivar ‘Royal Velvet’. The purplish colour is represented by the anthers of the three studied breeding clones (0015-03, 0017-01, 0075-25).

Table 5.

Selected morphological characteristics of parental and breeding clones from crosses between clone 18 of H. × chmielii with H. hybridum ’Royal Velvet’.

	H. × chmielii clone 18♀	H. hybridum ’Royal Velvet’♂	0015-03	0017-01	0075-25
Leaf: width [cm]	4.54 ± 0.36 a*	4.04 ± 0.33 a	4.30 ± 0.42 a	4.36 ± 0.22 a	4.42 ± 0.29 a
Stem (Peduncle): length [cm]	57.46 ± 2.93 a	56.64 ± 3.71 a	60.20 ± 4.67 ab	64.80 ± 2.34 b	55.22 ± 3.05 a
Inflorescence: number of flowers [pcs.]	3.40 ± 0.49 a	4.2 ± 0.40 ab	4.0 ± 0.00 ab	3.8 ± 0.40 a	5.0 ± 0.89 b
Flower: maximum length of perianth [cm]	10.70 ± 0.92 a	12.16 ± 0.45 a	13.78 ± 0.87 b	14.02 ± 0.33 b	12.14 ± 0.76 a
Flower: maximum width of perianth [cm]	13.26 ± 0.93 a	17.50 ± 0.32 b	17.50 ± 0.45 b	16.70 ± 0.36 b	12.80 ± 0.61 a
Flower: main colour of inner side [CC RHS]	32A	46A	45A	44A	45A
Stamens: colour [CC RHS]	32A	46A	46A	51A	46A
Anthers: colour (just before dehiscence) [CC RHS]	Reddish (48C)	Reddish (46A)	Purplish (187B)	Purplish (185C)	Purplish (187D)

*Means ± standard deviation (SD) in a verse followed by the same letter does not differ significantly at α = 0.05 (Tukey test); averages compared across traits.

Among the entire studied Hippeastrum population in terms of perianth shape, the largest number of plants (15 plants) had star-shaped flowers, accounting for 75% of the population. The next group was formed by genotypes with triangular perianth shapes: clones 0037-13, 0040-31, 0050-16, 0075-25, which accounted for 20% of the population. The last group contained only one clone 0017-01 (5%) with a round perianth shape (Fig. 2, Supplementary Fig. S2).

Fig. 2.

Distribution of the population of Hippeastrum genotypes in terms of the shape of the perianth as seen from the front.

Red was the most abundantly represented colour in the study population (13 pieces, 65%). Another group was formed by an orange-red color, which included five genotypes: clone 18 of H. × chmielii and four breeding clones: 0033-10, 0050-16, 0021-10, 0023-11 (25%). One genotype each (5%) was in the yellow-green and greyed-purple groups, ‘Rio Negro’ and clone 0037-13 (5%), respectively (Fig. 3, Supplementary Fig. S3).

Fig. 3.

Distribution of the population of Hippeastrum genotypes in terms of main perianth colour according to CC RHS.

Evaluation of the number and length of the stomata of the studied genotypes

Microscopic analysis of stomatal length showed differences in the number and length of Hippeastrum population (Fig. 4). It was found that 15 genotypes were characterized by the longest of the stomata—above 49.57 µm. Another 4 genotypes had shorter stomata with lengths ranging from 42.48 to 46.76 µm. Only one genotype tested, ‘Rio Negro’, had stomata with an average length of 35.73 µm (Table 6, Supplementary Tab. S6).

Fig. 4.

Microscopic image of the stomata visible on the abaxial side of the leaf of cultivar ’Rio Negro’ (A), of breeding clones 0050-16 (B) and 0017-17 (C).

Table 6.

Length and number of stomata of the 20 Hippeastrum genotypes.

Genotype	Lenght of Stomata [µm]	Stomatal Density [pcs.]	Genotype	Lenght of Stomata [µm]	Stomatal Density [pcs.]
H. × chmielii clone 6	49.88 ± 2,88 e*	39.20 ± 1.80 efg	0072-17	51.77 ± 3.68 f	35.47 ± 3.20 de
H. × chmielii clone 18	54.10 ± 3.25 hi	37.47 ± 1.54 efg	0003-13	49.81 ± 2.29 e	25.67 ± 3.53 ab
’Gervase’	55.40 ± 2.80 ij	39.53 ± 1.96 fg	0004-04	56.91 ± 1.06 jk	28.40 ± 2.58 bc
’Rio Negro’	35.73 ± 3.29 a	59.07 ± 2.54 i	0004-05	56.03 ± 1.24 j	24.40 ± 2.78 a
’Royal Velvet’	50.15 ± 3.97 e	31.93 ± 2.77 cd	0021-10	42.99 ± 1.83 bc	45.47 ± 4.24 h
0033-10	56.84 ± 1.92 jk	23.40 ± 2.15 a	0023-11	42.48 ± 3.41 b	41.07 ± 4.37 g
0037-13	49.57 ± 2.86 e	26.20 ± 3.58 ab	0062-10	44.51 ± 2.09 c	40.40 ± 1.54 g
0040-31	58.28 ± 3.90 k	24.27 ± 2.74 a	0015-03	53.41 ± 4.33 h	38.93 ± 2.86 efg
0050-15	56.91 ± 3.00 jk	29.53 ± 3.38 bc	0017-01	51.06 ± 3.67 ef	36.27 ± 3.71 ef
0050-16	46.76 ± 1.94 d	40.07 ± 1.84 fg	0075-25	52.96 ± 2.43 gh	31.67 ± 3.07 cd

*Means ± standard deviation (SD) in a column followed by the same letter do not differ significantly (Tukey’s test) at α = 0.05; averages compared across traits.

In the case of the number of stomata per 1 mm², microscopic analysis showed that the cultivar ‘Rio Negro’ has the highest average number of stomata—59.07 pcs. For the other genotypes studied, this value ranged from 23.40 to 41.07 pcs. (Table 6, Supplementary Table S6).

Correlation analysis of the relationship between the length of the stomata and their number per 1 mm², showed that Pearson’s correlation coefficient was − 0.78, which indicates a strong linear relationship. The coefficient of determination (r²) was 0.60, which means that the variation in the average length of stomatal apparatuses was 60%, which is explained by the variation in the number of apparatuses per 1 mm² (Fig. 5, Supplementary Fig. S5).

Fig. 5.

Graph of the scatter of the mean stomatal length in relation to the number of stomata per 1 mm² for the 20 Hippeastrum genotypes.

Assessment of genetic relatedness using RAPD and ISSR methods

RAPD analyses were performed (Fig. 6, Supplementary Fig. S6) yielded 3926 products using 30 primers (Table 7, supplementary Fig. S7), while ISSR analysis (Fig. 7), using the same number of primers, produced 4612 products. RAPD analysis yielded 221 polymorphic bands, representing 5.6% of the amplified products. Similarly, ISSR analyses yielded 147 polymorphic bands, accounting for 3.1% of the amplified products. In the RAPD analysis, a single primer generated an average of 7.4 polymorphic fragments, whereas in ISSR it was 4.9. The genetic similarity coefficient, according to Jaccard³², obtained for both techniques (RAPD and ISSR) was 0.768.

Fig. 6.

DNA polymorphism of 20 Hippeastrum genotypes after RAPD-PCR reactions with primers A-08 and A-12; 1—H. × chmielii clone 6, 2—H. hybridum ’Gervase’, 3—H. hybridum ’Royal Velvet’, 4—clone 0033-10, 5—clone 0037-13, 6—clone 0040-31, 7—clone 0050-15, 8—clone 0050-16, 9—clone 0072-17, 10—H. × chmielii clone 18, 11—H. hybridum ’Rio Negro’, 12—clone 0003-13, 13—clone 0004-04, 14—clone 0004-05, 15—clone 0021-10, 16—clone 0023-11, 17—clone 0062-10, 18—clone 0015-03, 19—clone 0017-01, 20—clone 0075-25.

Table 7.

Characteristics of PCR products for 20 Hippeastrum genotypes obtained using RAPD and ISSR markers.

Parameters	Marker type
	RAPD	ISSR
Total number of primers used	30	30
Total number of bands obtained	3926	4612
Sizes of generated products	200–2000 bp	100–2500 bp
Number of polymorphic bands	221	147
Number of monomorphic bands	62	124
Average number of polymorphic bands per primer	7.4	4.9
Percentage of polymorphic products	5.6	3.1
Average Jaccard similarity coefficient	0.768

Fig. 7.

DNA polymorphism of 20 Hippeastrum genotypes after ISSR-PCR reactions with primers 840 and 848; 1—H. × chmielii clone 6, 2—H. hybridum ’Gervase’, 3—H. hybridum ’Royal Velvet’, 4—clone 0033-10, 5—clone 0037-13, 6—clone 0040-31, 7—clone 0050-15, 8—clone 0050-16, 9—clone 0072-17, 10—H. × chmielii clone 18, 11—H. hybridum ’Rio Negro’, 12—clone 0003-13, 13—clone 0004-04, 14—clone 0004-05, 15—clone 0021-10, 16—clone 0023-11, 17—clone 0062-10, 18—clone 0015-03, 19—clone 0017-01, 20—clone 0075-25.

Based on the RAPD and ISSR analyses, a dendrogram was constructed using the Unweighted Pair Group Method with Arithmetic Mean (UPGMA) clustering method, from which five groups were distinguished (Fig. 8, Supplementary Fig. S8). The first group of similar genotypes included clone 6 of H. × chmielii, the cultivars ‘Gervase’ and ‘Royal Velvet’, as well as clones 0033-10, 0037-13, 0040-31, and 0050-15 (Figs. 9A,C,E and 10A–D). Interestingly, clone 6, as one of the maternal forms, shows 79% similarity with two paternal forms—‘Gervase’ and ‘Royal Velvet’, while the cultivars themselves exhibit 82% similarity between each other. Additionally, similarity with clones 0033-10, 0037-13, and 0040-31, which originated from crossing this maternal form with the cultivar ‘Gervase,’ and clone 0050-15, resulting from crossing with the cultivar ‘Royal Velvet’, is 78%.

Fig. 8.

UPGMA clustering dendrogram based on RAPD and ISSR marker polymorphism for 20 Hippeastrum genotypes.

Fig. 9.

Parental forms of H. × chmielii—clone 6 (A) and 18 (B) and cultivars ‘Gervase’ (C), ‘Rio Negro’ (D) and ‘Royal Velvet’ (E) of H. hybridum.

Fig. 10.

Breeding clones 0033-10 (A), 0037-13 (B), 0040-31 (C) 0050-15 (D), 0050-16 (E), 0072-17 (F) 0003-13 (G), 0004-04 (H), 0004-05 (I) 0021-10 (J), 0023-11 (K), 0062-10 (L) 0015-03 (M), 0017-01 (N), 0075-25 (O).

The second maternal form used in the previous crosses was clone 18 of H. × chmielii. It forms the largest group, clustering clones derived from crosses with the cultivars ‘Gervase’, ‘Rio Negro’, and ‘Royal Velvet’ (Figs. 9B and 10G–M,O). Breeding clones 0003-13, 0004-04, and 0004-05, resulting from crosses with the cultivar ‘Gervase’ show 80% similarity to clone 18. Clones 0021-10, 0023-11, and 0062-10, with ‘Rio Negro’ as the paternal form, exhibit 79% genetic similarity to clone 18, while genotypes 0015-03 and 0075-25, whose paternal form was ‘Royal Velvet’ display 77% and 79% similarity to clone 18, respectively.

The cultivar ‘Rio Negro’ and breeding clones 0050-16 and 0072-17 form three separate groups and are theoretically the least related to the entire population, although objectively these are still high values of genetic similarity (Figs. 9D and 10E,F). The similarity of the cultivar ‘Rio Negro’ to clones 0021-10, 0023-11, and 0062-10 is 77%, while the similarity of clones 0050-16 and 0072-17 to clone 6 is 76%. At the same time, clones 0050-16 and 0072-17, obtained from the same cross combination (clone 6 H. × chmielii crossed with ‘Royal Velvet’), showed the lowest relatedness (76%) compared to the other genotypes derived from the respective crosses. Comparing these two genotypes revealed differences in flower morphology and coloration, as well as in the entire plant.

The highest genetic similarity—85%—occurs between clones 0021-10 and 0023-11, which originated from the cross of clone 18 H. × chmielii with ‘Rio Negro’ (Fig. 10J,K). Both plants exhibited similar traits in terms of fertility and flower morphology in the paternal form.

Discussion

The aim of phenotypic evaluation of new breeding clones obtained from crosses of two old clones of H. × chmielii with cultivars of H. hybridum was to select genotypes combining features of the initial forms and simultaneously distinguishing them from the mother forms. These could become new cultivars derived from H. × chmielii, which was bred 30 years ago at WULS by Professor Henryk Chmiel³³. For a new genotype to be recognized as a new cultivar, it must meet the conditions of distinctiveness, alignment, and persistence of the traits. Therefore, for many plants, including Hippeastrum, UPOV has developed detailed descriptors of morphological traits, based on which a description of nine selected major plant traits out of a total of 22 has been made, which includes a descriptor for single flowers¹⁸.

The magnificent flowers of hipeastrum are one of the most important attributes for their decorativeness, and their size and shape determine the allocation of a cultivar to a particular horticultural classification group. Van Scheepen et al.⁹ evaluated 340 cultivars of over 1400 in the KAVB register in terms of flower size and shape, belonging to one of nine classification groups. Half of the cultivars (170) were classified in Galaxy group, as they had flowers over 16 cm in diameter. Our own analysis of the registration data of Hippeastrum cultivars in the Dutch KAVB for the last six years shows that the largest number of cultivars are registered in the groups where the flowers are the widest (> 16 cm) (Galaxy and Double Galaxy groups) and slightly smaller with a diameter in the range of 12–16 cm^10–15. In our study, the width of the perianth ranged from 12 cm for clone 6 of H. × chmielii to 17.5 cm for ’Royal Velvet’. In addition, relating the width of the flowers to the horticultural classification and disregarding their shape, it can be concluded that 50% of the studied population belongs to the Galaxy group and the other 50% to the Diamond group. It is worth noting that crossing clone 6, with smaller flowers, with the cultivar ’Gervase’, with larger flowers, produced hybrids with flowers of the same size as the paternal form or slightly smaller, but nevertheless larger than the maternal form clone 6 H. × chmielii. This is not the only example. Therefore, it is worth emphasizing that, with regard to the size of the flowers as measured by their width, it was possible in most cases to obtain breeding clones with larger flowers than the mother forms, both clones of H. × chmielii.

Leaf width is an important parameter affecting the assimilative area of plants. Owing to the increased photosynthetic area, carbohydrate accumulation increases, which is used in further plant growth and development processes^34,35. Our studies showed that the smallest leaf width (3.2 cm) was obtained by breeding clone 0062-10, and the largest (5.3 cm) by clone 0033-10. All the tested breeding clones of the two crosses (clone 18 and ’Royal Velvet’ and clone 18 and ’Gervase’) had leaves of the same width as the parental forms. This trait varied in clones derived from crosses of clone 6 with cultivars. In most cases, the breeding clones had leaves of the same width as one or both parental forms; in two cases (clone 0033-10 and 0072-17) the leaves were wider, and one clone (0062-10) had leaves narrower than both parental forms. Of the 16 genotypes evaluated by Azimi and Alavijeh³⁶, the cultivar ’Red Lion’ had the widest leaves at 7.5 cm, while the cultivar ‘Dancing Queen’ had the narrowest leaves at 3.7 cm. Interestingly, Shi et al.³⁴ showed that leaves of the same cultivar ’Red Lion’ had a width of 4.3 cm and ’Apple Blossom’ 4.8 cm. The leaf width of H. puniceum is 3.6 cm³⁷. According to the Pacific Bulb Society (PBS)³⁸, H. aulicum has leaves ranging from 3.8 to even 5.0 cm wide. Thus, it can be assumed that leaf size (including leaf width) is determined not only by genetic factors but also by the methods and conditions of cultivation³⁹.

Inflorescence stem length is a very important trait for cut-flower plants. Our research showed that breeding clone 0033-10 had the shortest inflorescence stems at 42 cm, while clone 0017-01 had the longest inflorescence at 65 cm. Clones with shorter stems (42–50 cm long) may be more valuable for pot cultivation and taller clones (55–65 cm) for cut flower cultivation. Nascimento et al.³⁷ reported that H. puniceum has equally long stems, averaging 61 cm in length. Other species, according to Traub⁸, Brickell⁴⁰, and PBS³⁸, can have stem lengths of 100 cm (H. vittatum), 60 cm (H. leopoldii), 50 cm (H. pardinum), 45–60 cm (H. aulicum), and 30–50 cm (H. reginae). Cultivars also vary in their stem length⁴¹. The breeding clones tested were mostly characterized by shorter or the same stem length as the parental forms; only clones 0021-10 and 0023-11 (both from a cross between clone 18 and ’Rio Negro’) had longer stems than the parental forms.

Flower number is a very important plant trait that influences the decorativeness of the whole plant and draws the attention of consumers. Therefore, cultivars with a higher flower number per inflorescence are more desirable and are more often used in breeding processes than those with a lower number³⁶. For the breeding clones tested, the average flowers number ranged from 3.2 units for clone 6 of H. × chmielii and clone 0050-16 to 5 pieces for clones 0037-13, 0021-10 and 0075-25. In the case of these three clones producing up to five flowers per stem, it was possible to transfer this trait to the breeding clones from the mother forms from the cultivars ’Gervase’, ’Rio Negro’, and ’Royal Velvet’, respectively. It is worth mentioning that in several other cases, an increase in the flower number was observed to an average of approximately 4 units compared to the mother forms (3.2 for clone 6 and 3.4 for clone 18). Although the difference was mostly statistically insignificant, it is important from a practical point of view when evaluating a potential new cultivar. Azimi and Alavijeh³⁶ reported that of the 16 genotypes evaluated, the cultivar ’Hercules’ had the highest number of flowers (5.3), whereas the lowest was observed in clone OPRC-206 (3.3). Meerow⁴² studied cultivars grown in Florida and showed that the cultivar ’Maiami’ had an average of 3.3 and the cultivar ’Jax’ 4.3 flowers. Nascimento et al.³⁷ showed that hybrids of Hippeastrum have at least three flowers per shoot, but rarely exceed six flowers, while species can have one or more flowers.

The perianth length of H. puniceum was determined to be 12.6 cm³⁷. For H. aulicum according to the PBS³⁸, it did not exceed 10 cm. In our study, clone 6 of H. × chmielii had the smallest perianth length (10.5 cm), which correlated with the smallest perianth width (12.3 cm). The longest perianth (14.8 cm) was characterized by clone 0072-17, a breeding clones of clone 6, and cultivar ’Royal Velvet’ whose perianth was also long (16.0 cm). In most cases, the clones had longer flowers than their mothers did.

According to the UPOV descriptor¹⁸, flower shape as seen from the front was divided into three categories: round, triangular, and star-shaped. The round shape was only observed in one clone, 0017-01, resulting from a cross between clone 18 and the cultivar ‘Royal Velvet’. Interestingly, both parental forms of this clone had a star-shaped perianth. A similar relationship can be observed in the triangular shape of the second category. The triangular-flowered clones (0037-13, 0040-31, 0050-15, 0075-25) also had a different perianth shape when viewed from the front relative to their parent forms. In the case of star-shaped flowers, which were found in the ten clones tested, no differences were observed relative to the parental forms.

The main colour of the inner side of the perianth of the studied Hippeastrum population is denoted by a specific number(s) according to the RHS CC colour catalogue, which breaks down into colour groups. The evaluated breeding clones changed colour slightly with respect to the maternal forms, although it can be noted that clones 0033-10 and 0050-16 which are breeding clones of clone 6 (red colour group), acquired an orange-red colour. Clone 0037-13, which is the breeding clone of clone 6 and the cultivar ’Gervase’, where both parental forms have a basic colour belonging to the red colour group, produces beetroot-coloured flowers, belonging to the grey-purple colour group. As a result of this crossbreeding, red colour enhancement occurred. In many cases in the clones, the perianth colour was different from that of the mother forms, but belonged to the same red colour group. Very interesting examples of such a colour change are the flowers of clone 0050-15, relative to those of clone 6, and clones 0003-13, 0004-05 and 0062-10, relative to those of clone 18. Genotype 0050-15 has raspberry-fuchsia coloured flowers (red group 46B) with white discoloured petal tips, which additionally have a dark perianth. In the case of clone 0003-13, the dominant colour was amaranth (red group 53D), but a delicately marked yellow-green throat was also visible. Clone 0004-05 has salmon-coloured flowers in various shades, with slight discolouration of the petals at their apex and a strongly marked veining (red group 38A). A small yellow-green throat is also observed. Clone 0062-10 has distinctive wavy-edged petals in a slightly smoky watermelon pink colour (red group 39B). The entire surface of the petals, excluding 3 lower petals, shows dark peach-coloured innervation. The flower has a pronounced yellow-green throat that runs from the base to the top of the petals. In addition, the dark peach-coloured border of the throat creates a significant contrast.

Although the finer flowers of hipeastrum, as in clones 6 and 18 of H. × chmielii, have their supporters, it is the large and showy flowers with interesting colours that arouse the greatest consumer interest in the market. Comparing phenotypically the breeding clones to clone 6, one can distinguish clone 0037-13 formed from a cross with the cultivar ’Gervase’, which produces more and, in addition, larger flowers on the shoot. In addition, the flowers are distinguished by their colour in the shade of beet and/or heavily ripe raspberries (greyed-purple group 185A) with a delicate green throat, and the edges of the petals are white. Closed buds are shaded white with delicate raspberry spotting running toward the top of the bud. Petals of the perianth curl-back. Interesting distinctive clones also include two genotypes: 0021-10 and 0023-11, which are derived from a cross between clone 18 and the cultivar ’Rio Negro’. Clone 0021-10 relative to its mother form, has longer stems with more and broader cinnabar-coloured flowers and prominent veining (orange-red group 34A). The petals of the perianth were narrow and sharply pointed, forming a characteristic star. A strongly marked throat with a flared creamy-white colour running from the base to ¾ of the height of the perianth petals was visible. The buds were creamy green-orange in clone 0023-11, which, relative to clone 18, had a larger number of flowers, which were wider and longer with a salmon colour (orange-red group 35B) with a prominent large and green throat, peachy at the base. In addition, the shades of the throat are lined with a dark peach-coloured border running parallel to the edges of the throat. Veining was observed throughout the flowers. The buds had a line along their entire length in the green peach shade. Venation was observed throughout the flower. The buds had a line throughout their length in green peach shade. Therefore, considering this aspect as the main one, it is worth distinguishing the clones mentioned above as potential new cultivars.

Analysis of stomatal number and size provides information on the traits that differentiate parental forms and hybrids derived from them, as well as variation in ploidy levels. This is one of the morphological markers used in breeding programs for initial hybrid evaluation⁴³. According to Shi et al.³⁴, the stomatal apparatus of Hippeastrum consists of two sickle-shaped cells with protective functions. Stomatal apparatuses are distributed over the entire leaf surface, but with a higher density on the abaxial side^44,45, which was also confirmed by our observations. A higher number of stomata on the lower side of the leaf makes it easier for plants to adapt to harsh environmental conditions. Our study allowed to detect morphological variations in the stomatal length and number. This may be due to the natural variability in the size of the apparatus in relation to the species, growth and cultivation conditions, and plant age⁴⁶. A study conducted by Shi et al.³⁴ on two cultivars of Hippeastrum—‘Red Lion’ and ’Apple Bloosom’ also confirmed the variation in their number and length.

The variation in the number and length of stomata in the studied population may be due to the different ploidy levels. Haist et al.⁴⁷ showed differences in stomatal length in their study of the diploid and autotetraploid forms of H. papilio. Analysis of stomatal size and number is a simple method for determining putative polyploids, which can be used in breeding programs based on Hippeastrum polyploidization methods^48–50. The use of this method at the initial stage of polyploid selection has many advantages, such as speed, low cost, and ease of execution, compared to chromosome number determination or flow cytometry analyses. In a study conducted by Haist et al.⁴⁷ on H. papilio and its polyploids, the size of the stomata on the underside of the leaf was a reliable indicator of the ploidy level of this genotype.

Regardless of the phenotypic evaluation of the studied Hippeastrum populations, an analysis of the genetic relatedness between parental forms and their breeding clones was conducted. Over the years, numerous studies have been conducted concerning the genetic relatedness of plants within species, as well as the relationships between botanical genera in the Amaryllidaceae family. Focusing on Hippeastrum, these studies have enabled delineation of this genus. Additionally, these methods, employing various genetic markers, are now widely used in identifying parental forms and their hybrids, which may eventually be registered as new cultivars^38,51–54. In the case of Hippeastrum, RAPD and ISSR are two examples of cost-effective and straightforward markers that have been successfully applied to assess genetic diversity^55–57. Ilczuk⁵¹ demonstrated that clone 18 of H. × chmielii is closely related to the cultivars ’Red Lion’ and ’Apple Blossom’. Additionally, the small flower size suggested a potential relationship with Rhodophiala pratensis (formerly Hippeastrum pratense), which might have been used as a parental form in the breeding process of this clone, or with the cultivar 'Scarlet Baby’. However, DNA polymorphism analysis using RAPD markers has not yet been confirmed. Huong et al.⁵⁸ employed RAPD markers to assess the degree of relatedness among five Hippeastrum clones derived from three crosses. They successfully identified the origin of each hybrid and confirmed the parentage of the clones. Varkulevičienė and Žukauskienė⁵⁹ utilized RAPD markers to analyze the genetic diversity of 20 H. hybridum cultivars (4 Dutch and 16 Lithuanian). Their results indicated that Lithuanian cultivars exhibited a 5% genetic difference from Dutch cultivars, with Lithuanian cultivars showing the highest DNA polymorphism. Datta⁶⁰ emphasized how DNA-based technologies using RAPD markers can facilitate the selection of appropriate parental lines based on similarity indices to develop desired cultivars. This information is crucial in targeted breeding programs.

The polymorphism level obtained using ISSR markers was considerably higher than that obtained using RAPD markers. However, this correlation was not observed in the present study. This may be due to the close genetic relatedness and/or low genetic diversity of the obtained genotypes. Zhang et al.⁵⁶ also employed ISSR markers in their studies. The population of 62 Hippeastrum cultivars (60 Dutch and 2 Chinese cultivars) was characterized by high variability and genetic diversity. Cluster analysis using the UPGMA method, similar to the present study, allowed the identification of seven groups, with cultivars exhibiting similar morphological clustering. All the 62 examined cultivars were identified using these markers. A similar pattern was observed in the current study, although with some exceptions. ISSR markers were used in the final stage of the study conducted by Yu et al.⁶¹ on callus induction from immature peduncles and the efficiency of Hippeastrum plant regeneration. These studies were carried out because calli are unstable structures that are prone to genetic changes. Ex vitro plants were subjected to morphological observations, and specimens showing no morphological differences were analyzed using ISSR markers, confirming the absence of DNA-level variation. Assessments of genetic diversity using ISSR markers in two populations of Rhodophiala bifida, a species related to the studied Hippeastrum, were performed by Schwedersky et al.⁶². Research conducted on 93 R. bifida individuals revealed DNA polymorphisms at the intrapopulation level. Further analysis divided the study population into two groups. This indicates that the gene flow between populations is relatively low and occurs mainly within individuals of the same population.

Genetic relatedness analysis among 20 Hippeastrum genotypes based on 30 RAPD and ISSR markers was confirmed by the constructed dendrogram for the original forms and breeding clones. The differences resulting from variability within the studied population were small but significant, as this is a highly hermetic group. The range of genetic similarity values was between 74 and 85%, suggesting that the genotypes were neither highly uniform nor extremely divergent. This is important for breeding strategies because moderate similarity allows for effective crossing without excessive inbreeding. Additionally, genotypes with higher similarity values may require careful evaluation of their morphological traits to ensure distinctness in cultivar registration.

An important observation is that both H. × chmielii clones share the same original forms. Regarding the cultivars ‘Gervase’ and ‘Royal Velvet’, Wang et al.⁶³ constructed a phylogenetic tree for 104 cultivars using 21 simple sequence repeat (SSR) markers and confirmed that these are two Dutch cultivars⁶⁴, which clustered together within the same group.

The results of the cluster analysis based on RAPD and ISSR markers scientifically confirm the previous crosses of clones 6 and 18 H. × chmielii with the paternal forms ‘Gervase’, ‘Rio Negro’, and ‘Royal Velvet’. The breeding clones grouped together with the maternal forms, indicating their genetic similarity, and consequently, relatedness.

In the context of Hippeastrum, the results of the present study highlight how RAPD-based technology can facilitate the selection of appropriate parental lines based on similarity indices for the development of breeding programs. The more frequent use of the RAPD method compared with ISSR in the referenced studies by other authors may be attributed to the advantages of this technique. Simple process optimization, the generation of diverse banding profiles, and the relatively low cost of the technology may explain the more frequent application of RAPD in studies of this type. It is also worth mentioning the disadvantages of this method, such as low repeatability of results, dominant nature of markers, low specificity of primers, and difficulties in interpreting bands.

Materials and methods

The Hippeastrum plants tested were all from the breeding populations and collection of the Section of Ornamental Plants, WULS, Warsaw, Poland (Figs. 9, 10). Clones 6 and 18 of H. × chmielii used for crossbreeding held in the collection of the WULS and in the Botanical Garden in Niegoszcz (Poland).

Phenotypic analysis of the studied genotypes

The analysis included 15 selected breeding clones obtained from crossbreeding Hippeastrum × chmielii Chm. clone no. 6 and clone no. 18 with Hippeastrum hybridum Hort. 'Gervase‘, ‘Rio Negro’ and ‘Royal Velvet’ as a result of crossings carried out in 2018 along with 5 initial forms (Table 8). Material for analyses was obtained by clonal propagation using adventitious bulbs.

Table 8.

Breeding clones obtained in 2018 from crossing H. × chmielii with cultivars of H. hybridum.

Crossbreeding	Clone number
H. × chmielii clone 6 × H. hybridum ’Gervase’	0033-10, 0037-13, 0040-31
H. × chmielii clone 6 × H. hybridum ’Royal Velvet’	0050-15, 0050-16, 0072-17
H. × chmielii clone 18 × H. hybridum ’Gervase’	0003-13, 0004-04, 0004-05
H. × chmielii clone 18 × H. hybridum ’Rio Negro’	0021-10, 0023-11, 0062-10
H. × chmielii clone 18 × H. hybridum ’Royal Velvet’	0015-03, 0017-01, 0075-25

The first observations and evaluations of the selected hybrids were conducted in 2019 and 2020, and the first flowering of some plants was recorded in 2020. At the end of February 2021, plants grown in the greenhouse of the WULS flowered en masse and were comprehensively evaluated and measured according to the selected 9 traits of the 2001 UPOV descriptor for the genus Hippeastrum¹⁸ (Figs. 11, 12):

Fig. 11.

Localization of descriptor trait 1 for leaf phenotypic evaluation for the genus Hippeatrum.

Fig. 12.

Localization of descriptor traits 2–9 for phenotypic evaluation of inflorescence stems and flowers for the genus Hippeastrum.

1. Leaf width [cm].

2. Peduncle (inflorescence stem) length [cm].

3. Number of flowers [pcs.]

4. Maximum length of perianth [cm].

5. Maximum width of perianth [cm].

6. Main colour of inner side [by CC RHS].

7. Stamens colour [by CC RHS].

8. Anthers colour (just before dehiscence) [by CC RHS].

9. Shape of the perianth at the front.

According to UPOV regulations¹⁸, all observations were made on plants that had fully open flowers soon after the anthers opened. Because daylight varies, colour determinations made from the international colour catalog (Colour Charts of the Royal Horticultural Society, CC RHS) were performed under artificial lighting or in the middle of the day in a room without direct sunlight. Measurements were taken at the bud and full bloom stages for 5 plants per replicate. The biometric measurements were performed using a ruler.

Analysis of stomata

Epidermal preparations were made from fragments of young, well-developed leaves taken 4.5 cm from their apex using the method described by Dyki and Habdas⁶⁵.

The number and length of stomata were measured using a Nikon Eclipse 80i microscope and NIS-Elements BR 2.30 software (Nikon, Tokyo, Japan). The apparatuses were counted per 1 mm² at × 10 objective for five fields of view for three prepared slides. Their lengths were determined at a magnification of 200 × , and measurements were performed in one repetition of 30 cameras for three prepared slides.

Determination of genetic relatedness using RAPD and ISSR markers

DNA from all Hippeastrum genotypes was isolated from microbulbs obtained from in vitro culture on Murashige and Skoog medium including vitamins⁶⁶ without plant growth regulators. The pH of the medium was set at 5.8. Five microbulbs were ground in a mortar with the addition of liquid nitrogen. From the homogenate, 0.1 g aliquots were prepared and placed in 1.5 ml Eppendorf tubes. The obtained material was stored at -86 °C. DNA was extracted using a GeneMATRIX Plant and Fungi DNA Purification Kit (EURx ® Sp. z o.o., Gdańsk, Poland).

To assess the efficiency of DNA extraction, electrophoresis was performed on a 1.5% agarose gel (Prona Agarose, ABO, Gdańsk, Poland) stained with 10 µl ethidium bromide.

A 25 µl reaction mixture was prepared for polymerase chain reaction (PCR) (Table 9).

Table 9.

Composition of the reaction mixture for polymerase chain reaction (PCR).

Components of the reaction mixture	Volume [µl] in a total volume of 25 µl
H2O	16.875
Buffer 10 × Pol Buffer C	2.5
dNTPs	2.5
Primer	2
Taq polymerase 1000 U	0.125
DNA	1

PCR was performed in triplicate for 30 RAPD and ISSR markers. Both RAPD-PCR and ISSR-PCR were carried out using an Eppendorf thermal cycler (Mastercycler® Nexus Gradient), and amplification was conducted according to the protocols presented in Table 10. The primers used in the reactions are listed in Tables 11 and 12.

Table 10.

PCR reaction profile—RAPD/ISSR.

RAPD/ISSR PCR stage	Temperature [°C]	Time [min]	Number of cycles
Initial denaturation	94.0/94.0	5.0/5.0	1/1
Denaturation	92.0/92.0	1.3/1.3	40/38
Primer annealing	35.0/34.1–55.9	1.3/1.3
DNA fragment elongation (extension)	72.0/72.0	1.3/1.3
Final elongation of DNA fragments	72.0/72.0	10.0/10.0	1/1

Table 11.

Primers used in the RAPD PCR reaction.

No	Primer ID	Primer sequence 5’—3'	No	Primer ID	Primer sequence 5’—3'
1	A-01	CAG GCC CTT C	16	A-16	AGC CAG CGA A
2	A-02	TGC CGA GCT G	17	A-17	GAC CGC TTG T
3	A-03	AGT CAG CCA C	18	A-18	AGG TGA CCG T
4	A-04	AAT CGG GCT G	19	A-19	CAA ACG TCG G
5	A-05	AGG GGT CTT G	20	A-20	GTT GCG ATC C
6	A-06	GGT CCC TGA C	21	B-01	GTT TCG CTC C
7	A-07	GAA ACG GGT G	22	C-02	GTG AGG CGT C
8	A-08	GTG ACG TAG G	23	C-06	GAA CGG ACT C
9	A-09	GGG TAA CGC C	24	C-11	AAA GCT GCG G
10	A-10	GTG ATC GCA G	25	D-05	TGA GCG GAC A
11	A-11	CAA TCG CCG T	26	D-10	GGT CTA CAC C
12	A-12	TCG GCG ATA G	27	D-20	ACC CGG TCA C
13	A-13	CAG CAC CCA C	28	E-01	CCC AAG GTC C
14	A-14	TCT GTG CTG G	29	E-03	CCA GAT GCA C
15	A-15	TTC CGA ACC C	30	F-02	GAG GAT CCC T

Table 12.

List of primers and annealing temperatures for the ISSR PCR reaction.

No	Primer ID	Primer sequence 5’—3'	Temperature [°C]
1	807	AGA GAG AGAG AGA GAG T	45.6
2	808	AGA GAG AGA GAG AGA GC	48.1
3	809	AGA GAG AGA GAG AGA GG	48.1
4	810	GAG AGA GAG AGA GAG AT	45.6
5	811	GAG AGA GAG AGA GAG AC	48.1
6	812	GAG AGA GAG AGA GAG AA	46.6
7	816	CAC ACA CAC ACA CAC AT	45.6
8	818	CAC ACA CAC ACA CAC AG	48.0
9	819	GTG TGT GTG TGT GTG TA	45.6
10	825	ACA CAC ACA CAC ACA CT	45.6
11	827	ACA CAC ACA CAC ACA CG	48.0
12	835	AGA GAG AGA GAG AGA GGC	51.3
13	836	AGA GAG AGA GAG AGA GCA	49.0
14	840	GAG AGA GAG AGA GAG ACT	49.0
15	841	GAG AGA GAG AGA GAG ATC	51.0
16	842	GAG AGA GAG AGA GAG ACG	51.3
17	847	CAC ACA CAC ACA CAC AGC	53.3
18	848	CAC ACA CAC ACA CAC AGG	53,3
19	856	ACA CAC ACA CAC ACA CCA	49.0
20	857	ACA CAC ACA CAC ACA CCG	51.3
21	861	ACC ACC ACC ACC ACC ACC	55.9
22	872	GAT AGA TAG ATA GAT A	34.1
23	873	GAC AGA CAG ACA GAC A	44.4
24	876	GAT AGA TAG ACA GAC A	39.3
25	878	GGA TGG ATG GAT GGA T	44.4
26	884	ATA AGA GAG AGA GAG AG	43.2
27	888	GGG CAC ACA CAC ACA CA	50.5
28	890	CCA GGT GTG TGT GTG TGT	51.3
29	899	CAT GGT GTT GGT CAT TGT TCC	54.0
30	900	ACT TCC CCA CAG GTT AAC ACA	54.0

To separate the PCR products, electrophoresis was performed on a 1.5% agarose gel (BIO STANDARD agarose, Prona Agarose, Gdańsk, Poland) containing 10 µl ethidium bromide. Twenty-five microliters of the reaction mixture were loaded into the wells, and the Perfect™ 100 bp DNA ladder (EURx ® Sp. z o.o., Gdańsk, Poland) was used as a molecular size/weight marker (control). Electrophoretic separation for both RAPD-PCR and ISSR-PCR reactions was conducted for 75 min at 120 V. Amplification products were documented using the Syngen Imagine system and Syngen Imagine 1.1 software (Syngen Biotech Sp. z o.o. Sp. k., Wrocław, Poland).

Statistical analyses

The results were analyzed using Statistica software version 13.3, and Tukey’s HSD test was used to determine significant differences between the averages at a significance level of α = 0.05. The experimental results were compiled using 1-factor analysis of variance and descriptive statistics. Pearson’s correlation coefficient and a regression model were used to determine the level of linear dependence between the variables.

The results of the molecular analyses were processed using the XLStat 2018 software. A binary matrix was created, where 1 indicated the presence of a DNA fragment and 0 indicated its absence. Genetic similarity was determined using the Jaccard³⁰ coefficient for RAPD and ISSR marker reactions. Similarity values were used for cluster analysis and dendrogram construction.

Conclusions

Phenotypic evaluation of the population derived from Hippeastrum × chmielii showed that the crosses led to the creation of new genotypes with improved ornamental characteristics. Most hybrids had larger and more intensely coloured flowers than the maternal forms, and some also exhibited favorable morphological traits, such as increased flower number, longer inflorescence stems, or broader leaves, confirming their potential as pot plants or cut flowers. Analysis of the stomatal apparatus revealed diversity and an inverse correlation between number and length, which probably reflects the differences in ploidy levels.

Genetic analysis using RAPD and ISSR markers revealed a moderate but significant level of genetic variability within the studied population, confirming both the accuracy of the crosses and the possibility of identifying hybrids. The cluster dendrogram demonstrated logical relationships between the parental forms and their clones and enabled the delineation of genotype groups with the highest degree of genetic similarity. The use of molecular markers has confirmed their usefulness in the selection and identification of new Hippeastrum cultivars. Three promising genotypes (0037-13, 0021-10 and 0023-11) were identified as the first cultivars derived from H. × chmielii.

These results provide a solid basis for further Hippeastrum breeding, the identification of promising parental forms and breeding clones traits, and highlight the value of integrating phenotypic and molecular methods in ornamental plant breeding.

Author contributions

Conceptualization, P.M. and D.S.; methodology, P.M., D.S., M.Z., K.N. and A.M.-C.; formal analysis, P.M. and M.Z.; writing—original draft, P.M.; writing—review and editing, P.M., D.S., J.R.-Ś. and A.M.-C.; visualization, P.M. and M.Z. All authors have read and agreed to the published version of the manuscript.

Funding

The publication was financed by the Science Development Fund of the Warsaw University of Life Sciences – SGGW, Warsaw, Poland.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.