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.

Fragrant rhododendron varieties remain relatively limited in current germplasm resources, constraining the enhancement of ornamental and aromatic characteristics in rhododendron breeding—this limitation has emerged as a critical bottleneck in the genetic improvement of rhododendrons. This research takes fragrant flower breeding as the breeding objective and conducts hybridization between varieties of the subgenus Tsutsusi, which can flower in multiple seasons and exhibit relatively strong resistance, and the fragrant R. moulmainense. Parallel intraspecific hybridizations within the subgenus Tsutsusi were implemented as experimental controls to quantify hybridization affinity. This study combines floral tube ontogeny histomorphological analysis, ovary paraffin sectioning, and optimized pollination protocols to address hybridization constraints, providing new insights for rhododendron intersubgeneric distant hybridization to create fragrant varieties. The results showed varying fertility among combinations, with some showing sterility or weak fertility due to low pollen germination and abnormal embryo development. Both pre- and post-fertilization reproductive barriers were observed, and different pollination methods significantly influenced ovary expansion and fruit set rates. Regarding limitations, this study lacks an in-depth analysis of reproductive isolation mechanisms, only describing phenotypic characteristics through morphological and histological methods, and it does not employ molecular techniques. The fundamental causes of reproductive isolation between subgenera therefore remain unclear. Additionally, there was no long-term monitoring of seedling emergence rates, hybrid plant growth potential, or flowering traits. This limits the ability to comprehensively evaluate the breeding value and genetic stability of distant hybrids.

Rhododendron decorum Franch. has high ornamental value and emerging edible potential, but conventional propagation cannot meet the demand for high-quality, genetically uniform planting stock. This study optimized explant disinfection, adventitious bud induction, shoot proliferation, and rooting culture using stem segments and leaves from a winter-flowering naturally mutated individual of Rhododendron decorum, aiming to minimize contamination and browning. Results demonstrated that combined disinfection with 1% KMnO₄ (10 min) and 10% H₂O₂ (4 min) outperformed single-treatment with 0.1% HgCl₂. For primary culture, the optimal medium for adventitious bud induction was WPM + 30 g·L⁻¹ sucrose + 7.5 g·L⁻¹ agar + 0.56 g·L⁻¹ Ca(NO₃)₂ + 100 mg·L⁻¹ VC + 3 mg·L⁻¹ TDZ + 0.1 mg·L⁻¹ NAA, achieving an induction rate of 66.67%. During subculture, when the concentrations of TDZ and NAA were adjusted to 0.6 mg·L⁻¹ and 0.05 mg·L⁻¹, respectively, and 2 mg·L⁻¹ GA₃ was supplemented, stem nodes exhibited slight thickening but no significant shoot elongation. For rooting, the basal medium WPM + 20 g·L⁻¹ sucrose + 7.5 g·L⁻¹ agar + 0.56 g·L⁻¹ Ca(NO₃)₂ + 100 mg·L⁻¹ VC + 1 g·L⁻¹ activated charcoal supplemented with 0.8 mg·L⁻¹ NAA yielded superior rooting performance. Adding 2 mg·L⁻¹ 2iP tended to enhance rooting rate relative to the NAA-only treatment from an independent rooting experiment. For callus-mediated bud induction from leaves, 0.1 mg·L⁻¹ TDZ + 0.01 mg·L⁻¹ NAA was optimal. Notched fully expanded apical leaves from tissue-cultured plantlets showed highest adventitious bud induction under 0.3 mg·L⁻¹ TDZ + 0.01 mg·L⁻¹ NAA. The established micropropagation system supports bud sport variety selection, mass seedling production, and genetic transformation research.