Studies on Distant Hybridization Compatibility between the Azalea (Rhododendron × hybridum hort.) and the Rhododendron decorum Franch. Native to China

Abstract

Rhododendron resources are abundant in China, and hybridization breeding technology is the primary method for cultivating Rhododendron varieties. In order to optimize the utilization of wild Rhododendron resources for distant hybridization, this study took three horticultural varieties of Rhododendron subgenus Tsutsusi and the Rhododendron decorum Franch. of the subgenus Hymenanthes as the research objects, and the cross-compatibility between subgenera was analyzed from the aspects of pollen tube growth and ovary and seed development. At the same time, the statistics of ovary swelling rate and fruit bearing rate, numbers of capsule seeds, 1000 seed weight of hybrid seeds, germination rate, green seedling rate, and other indexes were analyzed to comprehensively evaluate hybrid fertility. The results showed that there was obvious pre-fertilization and post-fertilization barriers existing in the hybridization of Rhododendron × hybridum hort. and R. decorum. During the growth of pollen tubes, a large amount of callus appeared, which led to the entanglement, distortion, and abnormal development of the pollen tubes; only part of the pollen tubes entered into the ovary or ovule. The pre-fertilization barriers can be overcome by early pollination and delayed pollination. According to the observation of the ovary paraffin section, 45 d after pollination, the seed was shriveled and developed abnormally. The comprehensive evaluation of hybrid fertility showed that ‘Sima’ × R. decorum was fertile, ‘Yin Taohong’ × R. decorum was a weakly fertile, and ‘Little Taohong’ × R. decorum was sterile. This study provided a scientific basis for intergeneric hybridization breeding between the subgenus Tsutsusi and the subgenus Hymenanthes.

1. Introduction

Rhododendron is the largest genus of woody plants in the Northern Hemisphere and an important component of montane ecosystems, comprising eight subgenera and more than 1000 species [1]. The subgenus Tsutsusi and the subgenus Hymenanthes are important research subjects in the breeding work of the Rhododendron, and most of the Rhododendron cultivars in China are from the subgenus Tsutsusi and the subgenus Hymenanthes. The subgenus Tsutsusi is a medium-sized subgenus of the Rhododendron with about 115 species, and the subgenus Hymenanthes contains about 302 species, most of which are diploid and can interbreed with each other [2].

The Azalea (Rhododendron × hybridum hort.) is a Rhododendron horticultural hybrid of the subgenus Tsutsusi from Belgium. It was selected by the repeated hybridization of R. indicum, R. simsii, and R. mucronatumand [3]. They are popular in the market because of their dwarf size, large flowers, and long flowering period, but do not have a scent. R. decorum belongs to the subgenus Hymenanthes [4], which is an endemic species of China; its inflorescence is shaped like hydrangeas and is fragrant; and it is a rare source of fragrant flowers among Rhododendron [5].

Breeding of Rhododendron in the subgenus Tsutsusi is based on hybridization and bud mutation, and the horticultural varieties have been cultivated with a wide variety of flower shapes and colors, but there is a lack of fragrance. Intersubgeneric hybridization promises breakthrough new varieties, but compared with the hybridization within the subgenus, the cross compatibility between the subgenera is weak. Within a certain range, the greater the genotypic difference between the parents and the more distantly related they are, the stronger the hybrid advantage will be. Hybridization fertility is generally judged from indicators, such as the rates of ovary swelling and fruit bearing, germination, and green seedlings. The selection of hybrid parents plays an important role in the rate of success of crosses, and Zheng et al. [6] found that the distance of parental affinity of crosses influenced the fruiting rate, which was lower among different subgenus and higher within the same subgenus. Sorokhaibam et al. [7] found that the compatibility of hybridization between Rhododendron was also highly correlated with the morphology and growth habits of hybrid parents.

In the study of hybridization compatibility of Rhododendron, there are usually pre-fertilization and post-fertilization barriers after hybridization between the subgenus Tsutsusi and the subgenus Hymenanthes. At present, many researchers have conducted studies on hybrid compatibility in the aspects of hybrid incompatibility orientation, pollen tube growth, and gene identification affecting hybrid compatibility. Among them, Xie et al. [8] identified physiological and biochemical factors that might lead to hybrid incompatibility by observing pollen tube growth, measuring ovary hormones, and analyzing RNA-Seq of interspecific hybridization. In recent years, a large number of studies have found that the pre-fertilization barriers can be overcome by different pollination methods [9], while the post-fertilization barriers are mainly completed by embryo rescue technology [10].

The R. hybridum is an excellent hybrid of the subgenus Tsutsusi, for which the degree of cross-breeding research in China is not high at present, and the research results that can be found are very limited, so it is of certain exploration to use it as a parent for cross-breeding. In this study, the fragrant R. decorum was crossed with an excellent variety of the subgenus Tsutsusi to overcome the hybrid fertilization barrier by different pollination methods and comprehensively evaluate their cross fertility. At the same time, by observing the pollen tube growth and seed development, we can determine the time point and specific types of hybridization barriers. This hybridization is projected to provide a scientific basis for the improvement of Rhododendron varieties and is expected to be cultivated with excellent ornamental and resistant varieties of Rhododendron, as well as providing fragrance.

2. Materials and Methods

2.1. Plant Material

The experiment was conducted in a plastic greenhouse at the School of Landscape Horticulture, Yunnan Agricultural University (Kunming, China). In the hybridization experiment, the R. hybridum varieties ‘Sima’, ‘Yin Taohong’, and ‘Little Taohong’ were used as the hybrid female parent (Figure 1). The pollen of R. decorum was harvested in Driving Township, Huize County, Qujing City, China (103.36° N, 25.98° E). At the same time, the Rhododendron hybrid ‘Fuchsia Parasol’ and the Rhododendron hybrid ‘Red Tiara’ were used as the control male parent. All the materials used were more than three years old and were healthy with good reproductive status.

2.2. Pollination

Before pollination, the pollen vitalities of R. decorum, ‘Fuchsia Parasol’, and ‘Red Tiara’ were detected by 0.5%MTT staining, and the results were 89.33%, 90.05%, and 88.2%, respectively, which were suitable for pollination.

The hybridization experiment was conducted in April 2021. During the blooming period, flowers at the tight bud stage were carefully removed the stamens and individually bagged to ensure isolation and prevent cross-pollination. We collected fresh pollen from 8:00–11:00 a.m. for artificial pollination, and we continuously pollinated for 2 d and bagged for 2–3 d after pollination. There should be at least 30 pollinated flowers per hybrid combination. Considering Yin et al.’s [11] experiment with R. hybridum as the female parent, the three better pollination methods were selected for pollination in this experiment, which are as follows: (1) The method for conventional pollination includes pollinating when the stigma begins to secrete mucus during the flowering period. (2) The early pollination method includes pollinating at the first flowering stag when the stigma does not secrete mucus. (3) The delayed pollination method includes pollinating at the end of flowering when the stigma secretes a substantial amount of mucus.

2.3. Evaluation of Cross Fertility

The ovary development was observed regularly, the number of ovary enlargements was counted 1 month after pollination, and the number of fruit set was recorded 2 months later. Fruit ripening was observed 5 months after pollination, and the expanded fruits were harvested when the pericarp turned brown. The number of seeds in the fruit and the number of plump seeds were counted, and the weight of 1000 seeds was measured.

The hybrid seeds were sown by the aseptic seeding method (Growth Medium: WPM medium + 2 mg·L⁻¹ 6-BA + 0.1 mg·L⁻¹ NAA + 7.5 g·L⁻¹ agar + 30 g·L⁻¹ sucrose). Both the germinating seeds and seedlings were incubated under a 12 h photoperiod at 25 °C and an illumination intensity of 2000 lx. The ratio of seed germination was determined one month after the seeds were cultivated on the medium. Additionally, the green seedling rate was calculated 60 d after sowing.

The fertility index levels and weights were assigned as shown in

Table 1. Ranking and weighting of cross-fertility indices.

Index	Sterile		Weakly Fertile		Fertile
	Threshold	Value	Threshold	Value	Threshold	Value
Green seedling rate (%)	0 < Gs < 10	1.0	10 ≤ Gs < 50	3.0	Gs ≥ 50	5.0
Green seedling coefficient	0 < Gc < 0.6	0.5	0.6 ≤ Gc < 0.9	1.5	Gc ≥ 0.9	2.0
Fruit setting rate (%)	0 < St < 20	0.5	20 ≤ St < 40	1.0	St ≥ 40	1.5
Number of fertile seeds per unit	0 < Sf < 20	0.5	20 ≤ Sf < 200	1.0	Sf ≥ 200	1.5

Gs, green seedling rate; Gc, green seedling coefficient; St, setting rate; Sf, fertile seeds.

[12], and the sum of individual scores was the total score of the species, which was also the fertility value. Based on the fertile value, the fertile property was divided into four classes, which included 0 ≤ none < 2.5, 2.5 ≤ low < 5, 5 ≤ medium < 8, and high ≥ 8, corresponding to sterile, weakly fertile, fertile, and highly fertile types. The fertility indicator was calculated using the following formulas:

$$\mathrm{Ovary} \mathrm{swelling} \mathrm{rate}={\displaystyle \frac{\mathrm{the} \mathrm{number} \mathrm{of} \mathrm{hybrid} \mathrm{fruit} \mathrm{expansions}}{\mathrm{the} \mathrm{number} \mathrm{of} \mathrm{hybrid} \mathrm{flowers}}} \times 100\%$$

(1)

$$\mathrm{Fruit} \mathrm{bearing} \mathrm{rate}={\displaystyle \frac{\mathrm{the} \mathrm{number} \mathrm{of} \mathrm{hybrid} \mathrm{fruits} \mathrm{with} \mathrm{seeds}}{\mathrm{the} \mathrm{number} \mathrm{of} \mathrm{hybrid} \mathrm{flowers}}} \times 100\%$$

(2)

$$\begin{array}{c}\mathrm{T}\mathrm{h}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{a}\mathrm{n}\mathrm{d} \mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t} \mathrm{o}\mathrm{f} \mathrm{h}\mathrm{y}\mathrm{b}\mathrm{r}\mathrm{i}\mathrm{d} \mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s}=\mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t} \mathrm{p}\mathrm{e}\mathrm{r} 100 \mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s}\times 10\end{array}$$

(3)

$$\mathrm{Germination} \mathrm{rate}={\displaystyle \frac{\mathrm{the} \mathrm{number} \mathrm{of} \mathrm{germinating} \mathrm{seeds}}{\mathrm{the} \mathrm{number} \mathrm{of} \mathrm{seeds} \mathrm{for} \mathrm{testing}}} \times 100\%$$

(4)

$$\mathrm{Green} \mathrm{seedling} \mathrm{rate}={\displaystyle \frac{\mathrm{Green} \mathrm{seedling} \mathrm{count} \mathrm{from} \mathrm{germinating} \mathrm{seeds}}{\mathrm{the} \mathrm{number} \mathrm{of} \mathrm{seeds} \mathrm{for} \mathrm{testing}}} \times 100\%$$

(5)

$$\mathrm{Green} \mathrm{seedling} \mathrm{coefficient}={\displaystyle \frac{\mathrm{green} \mathrm{seedling} \mathrm{rate}}{\mathrm{germination} \mathrm{rate}}} \times 100\%$$

(6)

$$\mathrm{Number} \mathrm{of} \mathrm{fertile} \mathrm{seeds} \mathrm{per} \mathrm{unit}=\mathrm{Average} \mathrm{seed} \mathrm{count} \mathrm{per} \mathrm{capsule} \times \mathrm{green} \mathrm{seedling} \mathrm{rate}$$

(7)

2.4. Fluorescence Observation of the Pollen Tube

The style compression method described by Williams et al. [13] involves the harvesting of three pistils at 4 h, 8 h, 1 d, 2 d, 3 d, 4 d, 5 d, 6 d, 7 d, 8 d, 9 d, and 10 d after pollination. The pistils were fixed in FAA solution (70% ethanol/glacial acetic acid/methanol [18:1:1]) for 24 h and then placed in 70% ethanol and stored at 4 °C for later use. Before pressing, the pistils were rinsed three times each with 70%–50%–30% ethanol-distilled water in successive stages of rehydration. The fixed pistils were softened for 4 h at 60 °C in 4 N NaOH and then stained for 24 h with 0.1% water-soluble aniline blue in 0.1 mol/L K₃PO₄. The pistil was unfolded longitudinally with 1–2 drops of glycerol and then pressed under a Leica fluorescence microscope (Wetzlar, Germany) to observe the growth and fertilization of the pollen tube.

2.5. Anatomical Analysis

Five ovaries of ‘Yin Taohong’ × R. decorum and ‘Yin Taohong’ × ‘Red Tiara’ were collected at 15 d, 30 d, 45 d, 60 d, and 75 d after pollination; soaked in FAA solution; and stored at 4 °C. Then, the fixed material was taken out and placed in glycerol–ethanol softener and soaked in an oven at 50 °C for 48 h for softening treatment. After being dehydrated and transparent, immersed in wax, and embedded, the slices were dyed by the safranin staining method and then decolorized for solid green staining. Finally, the transparent seal was observed, and the images were collected under a microscope (Olympus BX50, Tokyo, Japan).

2.6. Statistical Analysis

Microsoft Excel 2016 (Redmond, WA, USA) was used for routine statistics on the experimental data, and SPSS 26.0 (IBM, Inc., Armonk, NY, USA) was used to analyze the difference in variance and make multiple comparisons with a significance level of 0.05.

3. Results

3.1. Effect of Pollination Methods on Fruit Setting

As shown in

Table 2. Effect of pollination methods on the ovary swelling and fruit bearing rates.

Cross Combination (Female × Male)	Pollination Methods	Ovary Swelling Rate	Fruit Bearing Rate	Number of Seeds per Fruit	Weight of 1000 Seeds
‘Sima’ × R. decorum	Conventional	5.88	1.96	113.67 ± 21.13 c	0.0683 ± 0.0081 bc
	Early	30.61	18.37
	Delayed	13.73	5.88
‘Sima’ × ‘Fuchsia Parasol’	Conventional	63.46	55.69	376.33 ± 43.47 a	0.082 ± 0.0056 a
	Early	71.43	65.71
	Delayed	83.33	77.78
‘Yin Taohong’ × R. decorum	Conventional	13.16	5.26	75 ± 23 c	0.0587 ± 0.0055 c
	Early	11.11	7.41
	Delayed	52.83	22.64
‘Yin Taohong’ × ‘Red Tiara’	Conventional	68.42	55.26	272.67 ± 52.70 b	0.081 ± 0.0035 a
	Early	74.29	65.71
	Delayed	83.78	72.97
‘Little Taohong’ × R. decorum	Conventional	70.59	0.00	60 ± 2.64 c	0.0423 ± 0.0093 d
	Early	76.92	3.85
	Delayed	75.00	2.08
‘Little Taohong’ × ‘Red Tiara’	Conventional	63.64	52.27	304.67 ± 50.14 b	0.0793 ± 0.0035 ab
	Early	74.51	68.63
	Delayed	69.57	65.22

Same letters indicate no significant difference. p = 0.05.

, the rates of ovary swelling (30.61%) and fruit bearing (18.37%) of ‘Sima’ × R. decorum with early pollination were significantly higher than those of conventional and delayed pollination. The rates of ovary swelling and fruit setting of ‘Yin Taohong’ × R. decorum were 13.16% and 5.26%, 11.11% and 7.41%, and 52.83% and 22.64% for the conventional, early, and delayed pollination methods, respectively, with ovary expansion and fruit set rates of the delayed pollination method that were significantly higher than those of the other two pollination methods. All three pollination methods used for the cross between ‘Little Taohong’ × R. decorum resulted in excellent ovary swelling rates, with percentages of 70.59%, 76.92%, and 75% respectively. However, despite these high swelling rates, the fruit bearing rates were very low, and for the conventional pollination method, no fruits were produced at all. The control combinations of ‘Sima’ × ‘Fuchsia Parasol’, ‘Yin Taohong’ × ‘Red Tiara’, and ‘Little Taohong’ × ‘Red Tiara’ all had rates of ovary swelling and fruit bearing of all three pollination methods that reached over 55% with the early and delayed pollination methods, which were higher than those of the conventional pollination method. For the intergeneric hybrid of R. hybridum horticultural varieties × R. decorum, early and delayed pollination were both effective at increasing ovary swelling and fruit bearing.

Regarding the number of seeds per fruit and the weight of 1000 seeds, there was no significant difference in the number of seeds per capsule between the three intergeneric crosses. The ‘Sima’ × R. decorum capsules had the highest average number of seeds with 113.67, while ‘Yin Taohong’ × R. decorum and ‘Little Taohong’ × R. decorum had fewer at 75 and 60 seeds per fruit, respectively. The weights of 1000 seeds among the three intergeneric crosses were 0.0683 g, 0.0587 g, and 0.0423 g, respectively, with ‘Little Taohong’ × R. decorum having significantly fewer seeds than the other crosses. In contrast, the number of seeds per fruit and weight of 1000 seeds in the three controls were significantly higher than those of the three intergeneric crosses.

3.2. Fertility Evaluation

The seed germination of R. hybridum × R. decorum is shown in

Table 3. Cross compatibility indicators and comprehensive scores.

Cross Combination (Female × Male)	Germination Rate (%)	Green Seedling Rate (%)	Green Seedling Coefficient	Number of Fertile Seeds per Unit	Comprehensive Score
‘Sima’ × R. decorum	29 ± 7.94 b	23 ± 8.89 b	0.79	26.14	5.5
‘Sima’ × ‘Fuchsia Parasol’	80.33 ± 1.57 a	80.33 ± 1.57 a	1.00	302.31	10.0
‘Yin Taohong’ × R. decorum	2.67 ± 1.15 c	2.00 ± 2 c	0.88	1.50	3.0
‘Yin Taohong’ × ‘Red Tiara’	76.33 ± 6.81 a	76.33 ± 6.81 a	1.00	208.13	10.0
‘Little Taohong’ × R. decorum	3.33 ± 2.31 c	0.00	0.00	0.00	0.0
‘Little Taohong’ × ‘Red Tiara’	84 ± 10.58 a	84 ± 10.58 a	1.00	255.92	10.0

Same letters indicate no significant difference. p = 0.05.

and Figure 2. It was shown that the rates of seed germination and green seedlings of the intergeneric hybrid combinations were 29% and 23%, 2.67% and 2%, and 3.33% and 0, respectively. The rates of germination and green seedlings of ‘Yin Taohong’ × R. decorum and ‘Little Taohong’ × R. decorum were significantly lower than those of ‘Sima’ × R. decorum. Among them, the intergeneric hybrid ‘Sima’ × R. decorum (Figure 2D) grew moderately well with a few sectional chimeras of green and white or albino seedlings. In contrast, ‘Yin Taohong’ × R. decorum (Figure 2E) and ‘Little Taohong’ × R. decorum (Figure 2F) had low germination rates and poor growth. The green seedling coefficients for ‘Sima’ × R. decorum and ‘Yin Taohong’ × R. decorum were calculated to be high at 0.79 and 0.88, respectively. The rates of seed germination and green seedlings in the control combinations were approximately 80%, which were significantly higher than those of the intergeneric crosses, and the hybrid seedlings grew very well, with the seedlings basically free from yellowing and albinism (Figure 2A,B), while ‘Little Taohong’ × ‘Red Tiara’ (Figure 2C) had yellowish-green leaves. Based on fertility index scores, the comprehensive evaluation showed that the control combinations had a score of 10, which indicated that they belonged to the high-fertility type. ‘Sima’ × R. decorum had a score of 5.5, which was fertile. ‘Yin Taohong’ × R. decorum had 3 points and was a weakly fertile type. ‘Little Taohong’ × R. decorum scored 0 points and was sterile.

3.3. Pollen Tube Analysis

According to the fluorescence observations, no abnormalities were found during the growth of pollen tubes in the control group ‘Sima’ × ‘Fuchsia Parasol’. After 4 h of delayed pollination, a large amount of germination began at the stigma (Figure 3A3-4 h), and after 5 d, it entered the ovary and ovule (Figure 3A3-5 d). More pollen sprouted in the stigma after 8 h by conventional pollination (Figure 3A1-8 h), and the pollen tube extended toward the style after 1 d and entered the ovary after 7 d (Figure 3A1-7 d). The plants that were subjected to the early pollination method began to germinate after 1 d of pollination and reached 1/2 of the style after 5 d (Figure 3A2-5 d) and entered the ovary after 7 d.

The intergeneric hybrids ‘Sima’ × R. decorum and ‘Yin Taohong’ × R. decorum had callose and twisted pollen tubes with abnormal pollen tube development, and only partial entry into the ovary or ovules was observed. A large number of pollen grains started to germinate in the stigma after 4 h of delayed pollination method for ‘Sima’ × R. decorum, 8 h for conventional pollination, and 2 d for early pollination. After 1 d of conventional pollination, the pollen sprouted heavily and elongated at the stigma (Figure 3B1-1 d). The pollen tube began to twist when it reached 1/3 of the styles after 3 d. It approached the ovary after 7 d (Figure 3B1-7 d), and after 8 d, a small number of pollen tubes entered the ovary. However, no entry into the ovules was observed. Early pollination reached 2/3 of the style after 6 d (Figure 3B2-6 d), approached the ovary after 7 d and revealed a small amount of callus plug around the pollen tube, entered the ovary after 9 d, and was not found to have entered the ovule. After 1 d of delayed pollination, the pollen tube began to extend toward the ovary in an orderly manner (Figure 3B3-1 d). Callose was found close to the ovary after 5 d (Figure 3B3-5 d) and entered the ovary after 6 d and the ovule after 7 d (Figure 3B3-7 d).

After ‘Yin Taohong’ × R. decorum was conventionally pollinated for 1 d, the pollen germinated in large numbers at the stigma, and a large amount of callose appeared (Figure 3C1-1 d), reached 1/3 of the style after 4 d, and was still accompanied by callose (Figure 3C1-4 d). After 6 d, it approached the ovary but was not found to enter the ovary or ovule. A large number of pollen grains began to germinate in the stigma 2 d after early pollination, they reached 1/4 of the style after 3 d of pollination, and a small amount of callus was found (Figure 3C2-3 d). The pollen reached 1/2 of the style after 6 d of pollination; a large amount of callus was found, and the pollen tube was twisted (Figure 3C2-6 d). No entry into the ovary or ovules was observed. After 4 h of delayed pollination, a small number of pollen grains germinated in the stigma (Figure 3C3-4 h), they reached one-half of the style after 3 d, and a large number of callus plugs appeared (Figure 3C3-6 d). A small amount entered the ovary after 6 d and did not enter the ovule.

3.4. Observation of the Process of Embryonic Development by Paraffin Section

The development of ‘Yin Taohong’ × R. decorum seeds are shown in Figure 4A with neatly arranged and full seeds in the ovary 15 d after pollination (Figure 4A-15 d). Most of the seeds began to shrink, and the shape of the seeds developed abnormally after 30 d of pollination with only a few of the remaining seeds developing normally (Figure 4A-30 d). After 45 d of pollination, the seeds developed abnormally and formed a rod shape without endosperm or another irregular shape (Figure 4A-45 d). After 60 d of pollination, the seeds basically shrank, and the very few remaining abnormally shaped seeds continued to grow, although they had not observable embryos (Figure 4A-60 d).

The developmental process of the seeds of ‘Yin Taohong’ × ‘Red Tiara’ is shown in Figure 4B. At 15–30 d after pollination, the seeds in the ovary were neatly arranged and developed normally; the seed structure was intact, and the endosperm and embryo could be observed (Figure 4B-15 d; Figure 4B-30 d). After 45 d of pollination, the seeds developed more mature, but some seeds did not form embryos or endosperm (Figure 4B-45 d). After 60 d and 75 d of pollination, the seeds developed to maturity, while a few heteromorphic embryos were observed (Figure 4B-60 d; Figure 4B-75 d).

Therefore, it was apparent that the seeds of ‘Yin Taohong’ × R. decorum began to shrink and develop abnormally in shape after 30 d of pollination, and although most of the seeds of control ‘Yin Taohong’ × ‘Red Tiara’ developed normally, there were also heteromorphic embryos and some sterile seeds at 45 d after pollination. The shrinkage and abnormal shapes of the seed of Rhododendron hybrids early in development could be important in affecting the low rate of fruit set, number of seeds per fruit, and seed weight per 1000 in Rhododendron hybrids.

4. Discussion

4.1. Hybridization Compatibility Analysis

In the hybridization of the different subgenus of Rhododendron, there were many cases of unbalanced cross-compatibility. The research results of Eeckhaut et al. [14] showed that unidirectional sterility existed in the subgenus Tsutsusi and the subgenus Hymenanthes. Xie et al. [15] also found that there was unidirectional incompatibility between different hybrid populations of evergreen Rhododendron. In the study of Xiong et al. [16] and Yin et al. [11], they found that the fruit bearing rate of the hybrid with the subgenus Hymenanthes as the paternal parent was about 20%, while the conventional pollination fruit bearing rate of the hybrid with the R. hybridum as the female parent was 0. In this study, after the positive cross between the R. hybridum of subgenus Tsutsusi and the R. decorum of subgenus Hymenanthes, it was found that the fruit bearing rate of conventional pollination was very low in the three distant cross combinations, and the pollination methods applied in different combinations were different. The effect of the cross direction of the parents on the cross compatibility could be studied by reverse cross.

There are both pre-fertilization and post-fertilization barriers in Rhododendron hybridization [17]. Pre-fertilization barriers are primarily caused by abnormal enlargement, deformation, or even the cessation of pollen tube growth to the point where it fails to reach the ovule [18,19]. This was shown by the fact that the ovary did not expand or set fruit after pollination. In this study, there were obvious pre-fertilization barriers in three distant hybridization combinations. The callus deposition played a dominant role, which resulted in the disordered growth of pollen tubes, as well as the appearance of the phenomenon of twisted and tangled pollen tubes. At the same time, the pollen tube is unable to penetrate the ovule and fertilization is not possible, and a small number of penetrating ovules leads to a small number of fertilized ovules. Studies have shown the hybridization barriers encountered in combination with the subgenus Tsutsusi as the maternal parent in the low percentage of pollen tube penetration into ovules and the inhibition of ovule penetration [20]. In the hybrid combination of R. hybridum × R. decorum, a few pollen tubes were observed to enter the ovary, but only a very small number were observed to enter the ovule, suggesting that the inability of pollen tubes to penetrate the ovule was a direct determinant of incompatibility. In addition, Xie et al. [21] found that the pollen tubes of the R. lapponicum ‘XXL’ were elongating from 1 to 14 d after pollination when they were crossed with R. cyanocarpum. Thus, it was assumed that this phenomenon could be owing to the short sampling time, which could be extended to verify these previous findings.

Post-fertilization barriers manifest as the cessation of ovule development, failure of seeds to germinate, and albino progeny seedlings [22,23]. In this experiment, all six crosses in this trial were found to have a rate of fruit set lower than that of the rate of ovary expansion, that is, ovule abortion or incomplete development of mature seeds. The main causes of ovule abortion are endosperm abortion and embryo abortion. In the paraffin sections of ‘Yin Taohong’ × R. decorum, the results showed that the number of seeds decreased, and seeds development was incomplete, deformed, or atrophied at 30 d after pollination. The low rates of seed germination and hybrids with albino seedlings of the 3 distant hybridization combinations indicated that incomplete embryo development was an important factor leading to the albino or death of seedlings.

4.2. Comprehensive Evaluation of Intersubgeneric Hybridization

Plant cross compatibility may be related to chromosome ploidy of hybrid parents [24,25]. In distant hybridization, the selection of parental materials is crucial. Studies have found that the chromosome ploidy of hybrid parents is closely related to hybridization compatibility, and the closer the chromosome ploidy of parents, the easier the hybridization will be successful, and the greater the difference, the more serious the reproductive disorders [26,27]. In this study, R. decorum was diploid, while the ploidy and phylogeny of the three varieties of R. hybridum needed further exploration. To comprehensively evaluate the hybrid ability and the rates of ovary swelling and fruit bearing, the number of capsule seeds, the 1000 seed weight of hybrid seeds, and the rates of germination and green seedlings were selected as indicators of fertility in this experiment. The combination ‘Sima’ × R. decorum had a score of 5.5, indicating that it was fertile. ‘Yin Taohong’ × R. decorum had 3 points and was a weakly fertile type. ‘Little Taohong’ × R. decorum scored 0 points and was sterile. Therefore, when breeding Rhododendron in distant hybridization, we can consider selecting materials with close or the same ploidy and high affinity as parents.

4.3. Overcoming the Barrier of Distant Hybridization

The compatibility of distant hybridization could be overcome by using different pollination methods. Yin et al. [11] adopted the NAA smear stigma method, repeated pollination method, heated pollen method, and delayed pollination method to effectively improve the hybrid success rate of the R. hybridum ‘Sima’ × Rhododendron spinuliferum combination and effectively overcome the pre-fertilization barriers. In this study, its ovary expansion rate and fruit set rate can be improved by early and delayed pollination methods to overcome the pre-fertilization barriers to a certain extent. Post-fertilization barriers are usually overcome by the rescue of young hybrid embryos. It was observed in this experiment that the seeds of ‘Yin Taohong’ × R. decorum began to suffer severe shrinkage and degeneration at 30 d after pollination, so if embryo rescue is conducted, the optimal time should be within 15–30 d of pollination.

5. Conclusions

Rhododendron has high ornamental value, medicinal value, and social and cultural value, playing an important role in social development. This study not only provides a scientific basis for the breeding of intersubgeneric hybrids between the subgenus Tsutsusi and the subgenus Hymenanthes, helping Rhododendron breeders to cultivate new varieties with fragrance, but also provides new materials and ways for its development and utilization. In this experiment, the combinations of ‘Sima’ × R. decorum and ‘Yin Taohong’ × R. decorum had some degree of pre-fertilization and post-fertilization barriers. The suitable pollination methods to overcome hybridization barriers varied among different subgenus hybridization combinations. For distant hybridization combinations and inbred hybridization combinations within the subgenus Tsutsusi, early pollination and delayed pollination methods were found to be effective in overcoming certain pre-fertilization barriers. These results will provide more valuable information for the breeding, cultivation, and application of Rhododendron.