Exploring genetic variability and floral trait inheritance in F₁ hybrids from climber × miniature rose crosses Authors: Ezgi Dogan Meral, Soner Kazaz, Tugba Kılıç, Veysel Aydin, Suheda Basire Akca Yilmaz, Emine Kirbay Published: 2026 ABSTRACT Potted miniature roses are widely cultivated as popular indoor ornamental plants. Owing to increasing consumer demand, new varieties are introduced to the market each year. Cross-breeding remains the most prevalent method for developing potted miniature roses. In the present study, Rosa noisettiana Thory, a climber rose species, was used as the seed parent, while ten commercially popular miniature rose cultivars (Rosa White Star, Hot Jewel, Rosa Sweet Star, Rosa Bling Love Star, Orange Jewel, Rosa Shining Star, Rosa Lady Star, Orange Romance, Red Romance, and Sparkling Jewel) served as pollen donors. The success of hybridization was evaluated, and both quantitative and qualitative traits were analyzed in 861 F₁ hybrid individuals to assess the breeding potential of miniature roses. The percentage of pollen viability ranged from 17.71% to 46.79%, while the pollen germination rate varied between 10.33% and 21.68%. In total, 250 controlled pollinations were performed, all of which successfully developed into 250 fruits (hips). These fruits collectively contained 7,249 seeds. The results revealed that R. noisettiana × miniature rose combinations exhibited a 99.5% fruit set rate, with an average of 29.23 seeds per fruit and a seed germination rate of 20.53%. However, the recurrent blooming rate ranged between 15% and 34%, and many tall individuals were obtained from these crosses. Key commercial traits, such as flowering time, plant height, flower diameter, and petal number, were found to vary depending on the parental genotypes. In contrast, flower color did not exhibit as much variation as expected in the hybrid combinations. Based on these findings, further crosses should be conducted with different rose species to enhance genetic diversity in breeding programs and achieve more effective outcomes in the ornamental plant sector. Moreover, early selection methods utilizing genetic markers could be developed to increase the recurrent blooming rate and to obtain more compact plants.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09355-3. Introduction Hybridization between miniature and climbing roses presents considerable potential for combining desirable traits, including compact growth, recurrent blooming, and climbing ability. However, the success of such crosses is influenced by factors such as genetic compatibility, environmental conditions, and breeding efficiency [1]. Gaining a deeper understanding of the genetic basis of key traits—such as flower color, disease resistance, and plant architecture—is essential for developing effective breeding strategies. Genetic diversity is a cornerstone of crop improvement, as it broadens the gene pool and enhances the potential for effective selection [2, 3]. Roses are generally heterozygous, which makes it challenging to ensure the stable expression of target traits across successive generations [4]. Consequently, breeders strive to develop genetically uniform plants in both an efficient and cost-effective manner [5]. The development of F₁ hybrids represents a crucial step in combining complementary traits from selected parents [6]. One of the initial challenges in hybridization is the limited availability of seeds, as each seed represents a unique genotype. This necessitates a detailed phenotypic evaluation before selecting stable lines [7]. Furthermore, reproductive traits such as pollination success and seed set play a critical role in determining breeding outcomes, underscoring the importance of controlled pollination and optimal growing conditions. Climbing roses represent a valuable genetic resource in rose breeding due to their vigorous growth, high fertility, large flowers, and broad adaptability. In contrast, miniature roses are widely preferred as potted ornamentals because of their compact growth habit and recurrent blooming ability. Combining these contrasting parental groups may facilitate the development of novel genotypes with improved ornamental characteristics and increased genetic diversity. Although miniature rose breeding has been extensively studied, information regarding the breeding performance and trait inheritance of climbing × miniature rose crosses remains limited. Morphological characterization of F₁ progeny is essential for evaluating hybrid performance and identifying promising individuals for subsequent development. The present study represents the first large-scale evaluation of hybridization potential between a climbing rose species (Rosa noisettiana) and ten miniature rose cultivars, resulting in the generation of more than 860 F₁ individuals. Through a comprehensive assessment of fertility, recurrent blooming, and key morphological traits, the findings underscore the innovative use of climbing roses as maternal parents to broaden the genetic base of miniature roses. The present study represents one of the few comprehensive investigations of hybridization between Rosa noisettiana and commercially important miniature rose cultivars. Specifically, this study aimed (i) to evaluate the reproductive compatibility and fertility performance of climbing × miniature rose combinations, (ii) to assess the inheritance and variability of key ornamental traits, including recurrent blooming, plant architecture, flower size, petal number, and flower colour, and (iii) to determine the breeding potential of climbing roses as a genetic resource for broadening the diversity of miniature rose breeding programs. Materials and methods The study was conducted in the rose breeding greenhouse of the Department of Horticulture, Faculty of Agriculture, Ankara University (Ankara, Türkiye) (39°57′40.2″N, 32°51′51.7″E; [8] between May 2019 and October 2022. The ploidy levels of all rose genotypes included in the study were determined at the Plant Genetics and Cytogenetics Laboratory at Namık Kemal University. The seed parent used in the present study was a climber rose species (Rosa noisettiana Thory, RN, red color). As pollen sources, ten popular commercial miniature potted rose varieties were utilized: ‘Red Romance’ (RR, red color), ‘Rosa Sweet Star’ (RSS, red color), ‘Hot Jewel’ (HJ, red color), ‘Rosa Bling Love Star’ (RBLS, dark pink), ‘Sparkling Jewel’ (SJ, pink color), ‘Rosa Lady Star’ (RLS, pink color), ‘Rosa White Star’ (RWS, white color), ‘Rosa Shining Star’ (RShS, yellow color), ‘Orange Romance’ (OR, orange color), and ‘Orange Jewel’ (OJ, orange color) (Supplementary Figure S1). These varieties were obtained from different commercial suppliers and were selected based on their popularity in the Turkish market. More than twenty plants of each genotype were cultivated individually in pots (26 × 28 cm) filled with a substrate composed of cocopeat and perlite (50:50, v/v). During the cultivation period, greenhouse temperatures were maintained at 23 ± 5 °C, with a relative humidity range of 60% to 70%. To protect the plants from excessive summer sunlight, a 55% thermal screen was used. Irrigation and fertilization were provided through a computer-controlled drip irrigation system. Evaluation of ploidy levels and pollen germination test Prior to hybridization, the ploidy levels of the plants were determined by analyzing nuclear DNA content using a PARTEC CyFlow Space flow cytometer. Young, fully expanded leaves were selected as the tissue source, as they are commonly used in flow cytometry and are less affected by endoreduplication. Each sample consisted of nuclear DNA isolated from three distinct plants of the same genotype, thereby minimizing biological variation. To ensure calibration and accuracy, Vicia sativa L. cv. Nilüfer, with a known 2 C DNA content, was employed as the internal reference standard. For each measurement, leaf nuclei from the rose samples and the standard were co-processed and stained under identical conditions using PARTEC’s commercial kit (CyStain PI Absolute P), following the protocol described by Tuna and Cabi [9]. The staining solution was prepared by mixing 6 µL of RNase stock solution, 12 µL of propidium iodide (PI) stock solution, and 2 mL of staining buffer. Nuclear DNA content was calculated by comparing the fluorescence intensity (G1 peak) of the sample with that of the standard, according to the following formula:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}\text{Nuclear DNA Content (pg)}&=\frac{\text{Fluorescence intensity of sample (G1 peak)}}{\text{Fluorescence intensity of standard (G1 peak)}}\\&\times \text{Known DNA content of standard (pg)}\end{aligned}$$\end{document} Chromosome counts were conducted to validate the ploidy levels determined by flow cytometry. Root tip meristems were pretreated, fixed, and stained according to standard cytogenetic protocols. To minimize error, at least 10–15 well-spread metaphase plates were examined per genotype, and only clearly distinguishable chromosomes were included in the analysis. The chromosome counting results were then compared with the nuclear DNA content obtained from flow cytometry to ensure consistency and accuracy. Initially, the nuclear DNA content of each plant was measured, after which a plant with a clearly distinguishable DNA content was selected for chromosome counting. This approach enabled the correlation of DNA content with chromosome number, thereby confirming the ploidy level. For the pollen germination test, anthers were collected from unopened flower buds of ten plants one day prior to anthesis. The collected pollen was transferred into a glass container and incubated for 24 h at 24 °C with a relative humidity of 60%–65%. Three petri dishes were used per replicate. The germination medium consisted of 1% agar supplemented with 20% sucrose and 10 ppm boric acid, following the method described by Eti [10]. Each pollen parent was tested in four replicates. Pollen grains were applied to the medium using a fine paintbrush, and the samples were incubated for 24 h under the same temperature and humidity conditions. Pollen grains were considered germinated if the pollen tube length exceeded 1.5 times the diameter of the grain [11] (Supplementary Figure S2). For the hybridization process, both pollen donors and seed parents were carefully prepared. Anthers were collected and incubated at 22 °C with 60%–65% relative humidity to allow pollen release. To prevent self-pollination, seed parents were emasculated by removing their male reproductive organs early in the morning, following the method described by Chimonidou et al. [12] When flower buds reached 50%–60% openness, petals were removed, and the buds were covered with paper bags. Pollen was then applied to the stigmas of emasculated flowers using a fine brush, also in the morning, as described by Spethmann et al. [13]. Each cross was labeled and protected with paper bags for five days to avoid contamination. Hybridization was carried out between June 1 and July 30, 2019. Mature rose fruits (hips) were harvested in November 2019. Immediately after collection, fruit weights (FW) were recorded. Seeds were then extracted from the fruits and counted to determine the number of seeds per fruit (SPF). Total seed weight (TSW) was recorded for the seeds obtained from each cross combination and expressed in grams.The F₁ hybrid seeds underwent cold stratification for 90 days at 4 ± 1 °C to break dormancy and promote germination. After this period, the seed germination rate (SG) was calculated. Germinated seeds were then sown in a 1:1 mixture of peat and cocopeat. Throughout the process, data were collected for fruit set (FS), total seed number (TSS), seeds per fruit (NSpF), fruit weight (FW), total seed weight (TSW), and germination rate (SG) for each cross from the time of pollination. The resulting seedlings were transplanted into pots and cultivated for one year. In April 2021, the F₁ progenies were budded to evaluate growth performance and morphological characteristics. For each parental combination, at least 20 pollinations (dustings) were performed. In total, 250 pollinations were carried out across 20 different crosses. Qualitative and quantitative traits The full-bloom stage was recorded when the flowers were completely open, and flower diameter (FD) was measured during active bloom using a digital caliper [14]. Plant height (PH) was measured in centimeters (cm) from the base of the plant to the tip of the terminal bud [15]. The number of petals (PN) was counted on at least two flowers per genotype in the F₁ progeny once the reproductive organs (stigma and anthers) became visible, following the method described by Koning-Boucoiran et al. [16]. Only petals were considered in this evaluation. Recurrent blooming (RB) was expressed as a percentage, calculated as the proportion of individuals exhibiting recurrent flowering relative to the total number of evaluated plants for each hybrid combination. Petal color (PC) was visually assessed and reported as a percentage of the progeny expressing each color type. Leaf color (LC) was evaluated using a three-point scale (1 = matte green, 2 = glossy light green, and 3 = glossy dark green) [14], and this scale was consistently used in Table 4. The total number of flowers (NFpP) per progeny stem was recorded from the first bloom until just before pruning. The time elapsed from planting hybrid individuals until the appearance of their first flowers was referred to as the “duration of blooming” (DFB) [17]. Morphological traits of one-year-old F₁ plants derived from hybridization were evaluated over three flowering cycles between May and October 2021. Observations were carried out continuously on all flowers of each genotype throughout the flowering period. Statistical analysis Statistical analyses were performed to evaluate phenotypic variation among genotypes and to determine the relationships between hybrids and their respective parents. Principal Component Analysis (PCA) was conducted using standardized morphological data to reduce dimensionality and visualize genotype distribution based on trait similarities. The first two principal components were used to construct a biplot, allowing the identification of phenotypic divergence and grouping patterns among hybrid combinations and parental lines. Pairwise Pearson correlation analysis was applied to quantify the phenotypic resemblance between each hybrid genotype and its corresponding parents across all measured traits. A separate trait-based comparative analysis was conducted to calculate absolute differences between hybrids and their parents for each trait, and the results were further visualized using a heat map. For the evaluation of trait differences among genotypes, analysis of variance (ANOVA) was performed. Where significant differences were observed (p < 0.05), means were compared using Duncan’s Multiple Range Test. All statistical analyses and visualizations were conducted using the R and Python programming environments (version 3.11). Results and discussions Ploidy levels and in vitro germination rates Using flow cytometry, the nuclear DNA content of the miniature rose varieties ranged from 1.87 pg/2 C to 2.51 pg/2 C. All miniature rose genotypes included in the study were identified as tetraploid (2n = 4x) (Table 1). Table 1Nuclear DNA content and ploidy levels of species/ varieties [17]Species/VarietiesDNA (pg/2 C)Ploidy LevelRosa noisettiana Thory.1.952n=4xRed Romance1.952n=4xRosa Sweet Star2.412n=4xHot Jewel1.872n=4xRosa Bling Love Star2.222n=4xSparkling Jewel1.982n=4xRosa Lady Star2.512n=4xRosa White Star2.392n=4xRosa Shining Star2.372n=4xOrange Romance2.432n=4xOrange Jewel2.402n=4x Table 1Nuclear DNA content and ploidy levels of species/ varieties [17]Species/VarietiesDNA (pg/2 C)Ploidy LevelRosa noisettiana Thory.1.952n=4xRed Romance1.952n=4xRosa Sweet Star2.412n=4xHot Jewel1.872n=4xRosa Bling Love Star2.222n=4xSparkling Jewel1.982n=4xRosa Lady Star2.512n=4xRosa White Star2.392n=4xRosa Shining Star2.372n=4xOrange Romance2.432n=4xOrange Jewel2.402n=4x Nuclear DNA content and ploidy levels of species/ varieties [17] Breeding success in roses is generally higher when hybrid combinations involve parents with the same ploidy level. Recent studies on rose species and cultivars have revealed their nuclear DNA content, which serves as an indicator of ploidy levels. In this context, miniature roses with nuclear DNA contents ranging from 1.87 pg/2 C to 2.51 pg/2 C were identified as tetraploid. These findings are consistent with previous reports indicating that tetraploid roses typically possess DNA contents between 1.85 pg/2 C and 2.71 pg/2C [18]. Garden roses are commonly diploid or tetraploid, whereas commercial miniature and cut roses are predominantly tetraploid [19]. Tetraploid roses (2n = 4x) are particularly favored in breeding programs due to their high yield potential [20]. Despite the similarity in ploidy levels, our study revealed variability in fruit and seed set among hybrid combinations. This suggests that cross-incompatibility may still occur between tetraploid parents when their nuclear DNA contents differ. Such variation may arise from underlying genetic or physiological incompatibilities associated with differences in DNA content [21, 22]. With respect to pollen germination, ‘Rosa Bling Love Star’ exhibited the highest rate (21.68%), which was significantly greater than those of all other cultivars. ‘Rosa Shining Star’ (18.92%) and ‘Red Romance’ (18.89%) also demonstrated relatively high germination rates. In contrast, ‘Orange Jewel’ (10.33%) and ‘Sparkling Jewel’ (10.99%) showed the lowest pollen germination performances. While some cultivars may display strong pollen viability, their germination ability can remain limited, emphasizing the importance of assessing both traits when selecting pollen parents for hybridization (Table 2). Table 2Pollen viability and in vitro germination of pollen sourcesPollen SourcesPollen germination rate (%)Orange Romance14.31 ± 0.16dRosa Shining Star18.92 ± 0.13bRosa Bling Love Star21.68 ± 0.08aRosa White Star16.59 ± 0.08cSparkling Jewel10.99 ± 0.05eOrange Jewel10.33 ± 0.03fRed Romance18.89 ± 0.04bHot Jewel13.45 ± 0.05d,eRosa Lady Star14.65 ± 0.05dRosa Sweet Star12.09 ± 0.05d,eMeans followed by different letters are significantly different according to Duncan’s multiple range test (p < 0.05). Values are presented as mean ± standard deviation (SD) Table 2Pollen viability and in vitro germination of pollen sourcesPollen SourcesPollen germination rate (%)Orange Romance14.31 ± 0.16dRosa Shining Star18.92 ± 0.13bRosa Bling Love Star21.68 ± 0.08aRosa White Star16.59 ± 0.08cSparkling Jewel10.99 ± 0.05eOrange Jewel10.33 ± 0.03fRed Romance18.89 ± 0.04bHot Jewel13.45 ± 0.05d,eRosa Lady Star14.65 ± 0.05dRosa Sweet Star12.09 ± 0.05d,eMeans followed by different letters are significantly different according to Duncan’s multiple range test (p < 0.05). Values are presented as mean ± standard deviation (SD) Pollen viability and in vitro germination of pollen sources Means followed by different letters are significantly different according to Duncan’s multiple range test (p < 0.05). Values are presented as mean ± standard deviation (SD) Successful hybrid breeding depends mainly on the fertility of the parental genotypes, particularly the germination capacity of pollen from the donor plants. Pollen functionality is a critical determinant of successful fertilization and seed development. Previous studies have similarly reported wide variation in pollen germination rates among rose cultivars. Previous studies have reported a wide range of pollen germination rates (0–57%), which is consistent with our findings [11, 23–26]. Various fertility traits among cross combinations In the evaluation of fertility parameters across different cross-combinations, where the first listed cultivar represents the maternal parent and the second indicates the paternal parent, statistically significant differences were observed among the hybrids (Table 3). The average fruit weight (AFW) ranged from 6.39 g to 7.52 g, with the highest value recorded in R. noisettiana × Rosa White Star (7.52 g), which was significantly different from R. noisettiana × Rosa Sweet Star and R. noisettiana × Sparkling Jewel, as indicated by Duncan’s grouping letters. Table 3Various fertility traits among cross combinationsCross-combinationAFW(g)FS (%)TSSTSW(g)NSpFSG(%)R. noisettiana x Rosa White Star7.52 ± 0.03a100 ± 0.0a976 ± 1.00c1.66 ± 0.01a37.54 ± 0.08b21.0 ± 0.10cR. noisettiana x Hot Jewel7.06 ± 0.02ab100 ± 0.0a1113 ± 1.00a1.56 ± 0.01a53.00 ± 0.05a5.66 ± 0.05fR. noisettiana x Rosa Sweet Star6.44 ± 0.02b100 ± 0.0a633 ± 1.00g1.35 ± 0.01a28.77 ± 0.06d27.33 ± 0.06aR. noisettiana x Rosa Bling Love Star7.23 ± 0.03ab100 ± 0.0a1012 ± 1.00b1.52 ± 0.01a29.76 ± 0.05c17.89 ± 0.05dR. noisettiana x Orange Jewel7.42 ± 0.01ab100 ± 0.0a663 ± 1.00e1.46 ± 0.01a28.83 ± 0.08d10.41 ± 0.04eR. noisettiana x Rosa Shining Star7.38 ± 0.01ab100 ± 0.0a809 ± 1.00d1.36 ± 0.01a24.52 ± 0.06f9.27 ± 0.03eR. noisettiana x Rosa Lady Star6.71 ± 0.02ab100 ± 0.0a505 ± 1.00h1.18 ± 0.01a20.20 ± 0.05ı28.51 ± 0.04aR. noisettiana x Orange Romance6.62 ± 0.01ab100 ± 0.0a458 ± 1.00ı1.13 ± 0.01a19.91 ± 0.05ı23.80 ± 0.02bR. noisettiana x Red Romance7.33 ± 0.01ab100 ± 0.0a643 ± 1.00f1.57 ± 0.01a26.79 ± 0.08e27.37 ± 0.03aR. noisettiana x Sparkling Jewel6.39 ± 0.02b95 ± 0.10b437 ± 1.00i1.33 ± 0.01a23.00 ± 0.05g20.82 ± 0.03cMeans followed by different letters are significantly different according to Duncan’s multiple range test Values are expressed as mean ± standard deviation (SD)AFW Average fruit weight, FS Fruit set, TSS Total seed set, TSW Total seed weight (g), NSpF Number of seeds per fruit, and SG Seed germination (p < 0.05) Table 3Various fertility traits among cross combinationsCross-combinationAFW(g)FS (%)TSSTSW(g)NSpFSG(%)R. noisettiana x Rosa White Star7.52 ± 0.03a100 ± 0.0a976 ± 1.00c1.66 ± 0.01a37.54 ± 0.08b21.0 ± 0.10cR. noisettiana x Hot Jewel7.06 ± 0.02ab100 ± 0.0a1113 ± 1.00a1.56 ± 0.01a53.00 ± 0.05a5.66 ± 0.05fR. noisettiana x Rosa Sweet Star6.44 ± 0.02b100 ± 0.0a633 ± 1.00g1.35 ± 0.01a28.77 ± 0.06d27.33 ± 0.06aR. noisettiana x Rosa Bling Love Star7.23 ± 0.03ab100 ± 0.0a1012 ± 1.00b1.52 ± 0.01a29.76 ± 0.05c17.89 ± 0.05dR. noisettiana x Orange Jewel7.42 ± 0.01ab100 ± 0.0a663 ± 1.00e1.46 ± 0.01a28.83 ± 0.08d10.41 ± 0.04eR. noisettiana x Rosa Shining Star7.38 ± 0.01ab100 ± 0.0a809 ± 1.00d1.36 ± 0.01a24.52 ± 0.06f9.27 ± 0.03eR. noisettiana x Rosa Lady Star6.71 ± 0.02ab100 ± 0.0a505 ± 1.00h1.18 ± 0.01a20.20 ± 0.05ı28.51 ± 0.04aR. noisettiana x Orange Romance6.62 ± 0.01ab100 ± 0.0a458 ± 1.00ı1.13 ± 0.01a19.91 ± 0.05ı23.80 ± 0.02bR. noisettiana x Red Romance7.33 ± 0.01ab100 ± 0.0a643 ± 1.00f1.57 ± 0.01a26.79 ± 0.08e27.37 ± 0.03aR. noisettiana x Sparkling Jewel6.39 ± 0.02b95 ± 0.10b437 ± 1.00i1.33 ± 0.01a23.00 ± 0.05g20.82 ± 0.03cMeans followed by different letters are significantly different according to Duncan’s multiple range test Values are expressed as mean ± standard deviation (SD)AFW Average fruit weight, FS Fruit set, TSS Total seed set, TSW Total seed weight (g), NSpF Number of seeds per fruit, and SG Seed germination (p < 0.05) Various fertility traits among cross combinations Means followed by different letters are significantly different according to Duncan’s multiple range test Values are expressed as mean ± standard deviation (SD) AFW Average fruit weight, FS Fruit set, TSS Total seed set, TSW Total seed weight (g), NSpF Number of seeds per fruit, and SG Seed germination (p < 0.05) The FS percentage was generally high across all combinations, with all but one cross exhibiting a 100% fruit set. The lowest fruit set (95%) was recorded in R. noisettiana × ‘Sparkling Jewel,’ which differed significantly from the other combinations. The number of seeds varied considerably among combinations, ranging from 437 to 1,113. The average seed weight remained relatively consistent across combinations, with no significant variation observed. However, the NSpF varied significantly, with the highest value observed in R. noisettiana × ‘Hot Jewel’ (53.00), followed by R. noisettiana × ‘Rosa White Star’ (37.54), while R. noisettiana × ‘Orange Romance’ exhibited the lowest (19.91). The SG rate also showed marked differences among the crosses. The highest germination rates were recorded in R. noisettiana × ‘Rosa Lady Star’ (28.51%) and R. noisettiana × ‘Rosa Sweet Star’ (27.33%), whereas the lowest was observed in R. noisettiana × ‘Hot Jewel’ (5.66%). These findings indicate that although certain combinations perform well in seed production, this does not necessarily translate into high germination success. Among these results, the differences in seed germination rates are particularly significant to our main argument, as they provide a direct measure of hybridization efficiency. The correlation analysis among fruit and seed traits revealed several statistically significant relationships (Fig. 1). A strong positive correlation was observed between average fruit weight and average seed weight (r = 0.86, P < 0.01), indicating that heavier fruits tended to produce heavier seeds. Similarly, AFW was significantly correlated with FS (r = 0.54, P < 0.01) and TSS (r = 0.43, P < 0.05). FS also showed a significant positive correlation with the percentage of SG (r = 0.51, P < 0.01), suggesting that successful fruit formation may be associated with enhanced seed viability. TSS was positively correlated with NSpF (r = 0.82, P < 0.01), as expected. However, no significant correlation was found between TSW and seed germination (SG) (r = 0.04), nor between TSS and SG (r = 0.20). Interestingly, NSpF was negatively, though non-significantly, correlated with SG (r = -0.32). These results highlight the complex interactions among fruit and seed traits in rose hybridization, suggesting that improvements in one trait may not necessarily predict gains in others, particularly in terms of germination success (Fig. 1). Fig. 1Correlation of various fertility traits among cross combinations (Average Fruit Weight: AFW; Fruit Set: FS; Total Seed Set: TSS; Total Seed Weight: TSW; Number of seeds per flower: NSpF; Seed Germination: SG Correlation of various fertility traits among cross combinations (Average Fruit Weight: AFW; Fruit Set: FS; Total Seed Set: TSS; Total Seed Weight: TSW; Number of seeds per flower: NSpF; Seed Germination: SG Hybrid breeding success in roses is shaped by a combination of genetic, physiological, and morphological factors, with the fertility levels of parental genotypes playing a particularly central role. Fertility in hybrid combinations is primarily determined by viable pollen, efficient fertilization, and the successful development of fruits and seeds [11, 23]. In the present study, significant differences were observed among cross-combinations in key reproductive parameters, despite the use of genotypes with matching ploidy levels. These results indicate that factors beyond ploidy—such as variation in nuclear DNA content, genetic compatibility, and gametophytic–pistil interactions—are critical determinants of hybridization outcomes [21, 22, 27]. The variation observed in fruit set and seed production is consistent with previous reports, which show that even tetraploid roses may exhibit cross-incompatibility due to subtle genetic and cytological differences [28]. R. noisettiana, a climbing rose species, exhibited a consistently high fruit set rate across all hybrid combinations. This indicates that, in addition to its vigorous vegetative growth, R. noisettiana also demonstrates strong generative capacity. Climbing roses are generally recognized for their high fertility potential, which makes them advantageous as seed parents in hybridization programs. These results align with earlier findings highlighting the high fruit set capacity of climbing rose genotypes in breeding efforts [19, 29]. Although most hybrid combinations in the present study achieved high fruit set percentages, significant differences were observed in seed number, seed germination, and fruit characteristics. For example, R. noisettiana × ‘Hot Jewel’ produced the highest number of seeds per cross (1113) but exhibited the lowest seed germination rate (5.66%), suggesting that a high seed yield does not necessarily correspond with high germination success. Conversely, combinations such as R. noisettiana × ‘Rosa Lady Star’ and R. noisettiana × ‘Rosa Sweet Star’ produced moderate seed numbers but relatively high germination rates (28.51% and 27.33%, respectively), indicating superior embryo viability. The lack of correlation between SG and other reproductive parameters (e.g., FS, NSpF) may be attributed to abnormal meiotic divisions, chromosomal imbalances, dormancy mechanisms, or the presence of lethal alleles in specific crosses [30, 31]. Furthermore, rose seeds are known to exhibit both physical and physiological dormancy, often regulated by growth inhibitors such as abscisic acid (ABA) present in the pericarp and testa. These physiological barriers may delay or inhibit germination even in genetically viable seeds. In conclusion, the findings of the present study underscore that successful hybridization in roses is not solely dependent on ploidy compatibility but also requires careful consideration of genetic compatibility, pollen viability, and physiological factors influencing post-fertilization development. The complex interplay of these factors underscores the importance of judicious parental selection to optimize both seed production and germination efficiency in rose breeding programs. Morphological and floral traits of parents and progenies The morphological and floral characteristics of the parental cultivars involved in the hybridization process (Table 4), as well as the phenotypic traits observed in the 861 F₁ progenies derived from Rosa noisettiana crosses (Table 5), are presented below. According to the results, the parental genotypes exhibited considerable diversity in key traits, including plant height (PH), petal number (PN), flower diameter (FD), and petal color (PC). For instance, Rosa noisettiana exhibited the tallest plant stature (110.00 cm) and the largest flower diameter (12.87 cm). In contrast, pot roses such as Rosa Lady Star and Orange Romance displayed more compact growth and smaller floral dimensions, aligning with their ornamental breeding backgrounds. Table 4Morphological and floral traits of parentsSpecies/VarietiesRB(%)PH(cm)FD(cm)PNLCRosa noisettiana Thory.(RS)Yes110.00 ± 0.10a12.87 ± 0.03a27.0 ± 0.10f3 ± 0.00aRosa Sweet Star (RSS)Yes23.15 ± 0.05b7.18 ± 0.03b55.0 ± 0.10b3 ± 0.00aRosa Shining Star (RShS)Yes23.02 ± 0.03b6.7 ± 0.02b45.0 ± 0.10c3 ± 0.00aRosa White Star (RWS)Yes24.12 ± 0.03b6.5 ± 0.02b35.0 ± 0.10d3 ± 0.00aRosa Lady Star (RLS)Yes22.99 ± 0.04b5.92 ± 0.02b65.0 ± 0.10a3 ± 0.00aRosa Bling Love Star (RBLS)Yes25.86 ± 0.04b6.05 ± 0.01b45.0 ± 0.10c3 ± 0.00aRed Romance (RR)Yes23.45 ± 0.05b7.13 ± 0.01b35.0 ± 0.10d2 ± 0.00aOrange Romance (OR)Yes24.58 ± 0.03b7.5 ± 0.02b30.0 ± 0.10e3 ± 0.00aHot Jewel (HJ)Yes22.97 ± 0.03b6.33 ± 0.01b35.0 ± 0.10d3 ± 0.00aSparkling Jewel (SJ)Yes25.12 ± 0.03b6.29 ± 0.01b25.0 ± 0.10g2 ± 0.00aOrange Jewel (OJ)Yes23.49 ± 0.04b6.51 ± 0.01b30.0 ± 0.10e3 ± 0.00aMeans followed by different letters are significantly different according to Duncan’s multiple range test (P < 0.05) Values are expressed as mean ± standard deviation (SD)RB Recurrent Bloom, PH Plant height, FD Flower diameter, PN Petal number, LC Leaf color Table 4Morphological and floral traits of parentsSpecies/VarietiesRB(%)PH(cm)FD(cm)PNLCRosa noisettiana Thory.(RS)Yes110.00 ± 0.10a12.87 ± 0.03a27.0 ± 0.10f3 ± 0.00aRosa Sweet Star (RSS)Yes23.15 ± 0.05b7.18 ± 0.03b55.0 ± 0.10b3 ± 0.00aRosa Shining Star (RShS)Yes23.02 ± 0.03b6.7 ± 0.02b45.0 ± 0.10c3 ± 0.00aRosa White Star (RWS)Yes24.12 ± 0.03b6.5 ± 0.02b35.0 ± 0.10d3 ± 0.00aRosa Lady Star (RLS)Yes22.99 ± 0.04b5.92 ± 0.02b65.0 ± 0.10a3 ± 0.00aRosa Bling Love Star (RBLS)Yes25.86 ± 0.04b6.05 ± 0.01b45.0 ± 0.10c3 ± 0.00aRed Romance (RR)Yes23.45 ± 0.05b7.13 ± 0.01b35.0 ± 0.10d2 ± 0.00aOrange Romance (OR)Yes24.58 ± 0.03b7.5 ± 0.02b30.0 ± 0.10e3 ± 0.00aHot Jewel (HJ)Yes22.97 ± 0.03b6.33 ± 0.01b35.0 ± 0.10d3 ± 0.00aSparkling Jewel (SJ)Yes25.12 ± 0.03b6.29 ± 0.01b25.0 ± 0.10g2 ± 0.00aOrange Jewel (OJ)Yes23.49 ± 0.04b6.51 ± 0.01b30.0 ± 0.10e3 ± 0.00aMeans followed by different letters are significantly different according to Duncan’s multiple range test (P < 0.05) Values are expressed as mean ± standard deviation (SD)RB Recurrent Bloom, PH Plant height, FD Flower diameter, PN Petal number, LC Leaf color Morphological and floral traits of parents Means followed by different letters are significantly different according to Duncan’s multiple range test (P < 0.05) Values are expressed as mean ± standard deviation (SD) RB Recurrent Bloom, PH Plant height, FD Flower diameter, PN Petal number, LC Leaf color Table 5Morphological and floral traits of progenyHybrid combinationsRB(%)PH(cm)Range of PH(cm)FD(cm)Range of FD(cm)PNLCNFpPDFB(day)R. noisettiana x Rosa White Star28.68 ± 0.03b22.08 ± 0.03e9.80–50.04.27 ± 0.01a3.40–7.6747.36 ± 0.02a2.50 ± 0.01a1.81 ± 0.01a28.68 ± 0.02bcR. noisettiana x Hot Jewel18.18 ± 0.03e26.06 ± 0.02a13.50–56.704.80 ± 0.01a3.74–5.9830.67 ± 0.03e2.33 ± 0.01a1.50 ± 0.01a32.67 ± 0.03aR. noisettiana x Rosa Sweet Star22.08 ± 0.03cd24.17 ± 0.03bc8.20–43.903.76 ± 0.01a3.20–6.6239.15 ± 0.02c2.45 ± 0.01a1.74 ± 0.01a27.94 ± 0.02cR. noisettiana x Rosa Bling Love Star34.07 ± 0.03a21.86 ± 0.02e8.60–39.204.45 ± 0.01a1.55–11.4741.11 ± 0.01b2.50 ± 0.01a1.98 ± 0.01a25.98 ± 0.01dR. noisettiana x Orange Jewel15.00 ± 0.02f24.60 ± 0.02ab13.50–56.704.80 ± 0.01a3.74–5.9830.67 ± 0.03e2.33 ± 0.01a1.50 ± 0.01a32.67 ± 0.03aR. noisettiana x Rosa Shining Star20.59 ± 0.02d17.03 ± 0.02f11.0–21.03.69 ± 0.01a3.0-5.7140.21 b ± 0.01c2.41 ± 0.01a1.57 ± 0.01a24.57 ± 0.01dR. noisettiana x Rosa Lady Star22.92 ± 0.03c23.09 ± 0.01bc7.80–37.504.34 ± 0.01a3.42–7.3036.43 ± 0.01d2.59 ± 0.01a1.88 ± 0.01a28.73 ± 0.01bcR. noisettiana x Orange Romance21.43 ± 0.02cd25.58 ± 0.02ab11.20–52.03.86 ± 0.01a2.99–5.8335.19 ± 0.01d2.63 ± 0.01a1.83 ± 0.01a30.44 ± 0.01bR. noisettiana x Red Romance21.11 ± 0.01cd21.53 ± 0.02e14.60-35.754.57 ± 0.01a3.57–6.6935.54 ± 0.01d2.23 ± 0.01a2.79 ± 0.01a28.89 ± 0.01bcR. noisettiana x Sparkling Jewel28.79 ± 0.02b22.68 ± 0.02de12.50–34.204.68 ± 0.01a2.36–7.5731.59 ± 0.01e2.47 ± 0.01a1.98 ± 0.01a30.21 ± 0.01bMeans followed by different letters are significantly different according to Duncan’s multiple range test (P < 0.05) (RB (%): Percentage of individuals showing recurrent blooming in each hybrid population., PH Plant height, FD Flower diameter, PN Petal number, LC Leaf color, NFpP Number of flowers per plant, DFB duration of flower blooming) Values are expressed as mean ± standard deviation (SD) Table 5Morphological and floral traits of progenyHybrid combinationsRB(%)PH(cm)Range of PH(cm)FD(cm)Range of FD(cm)PNLCNFpPDFB(day)R. noisettiana x Rosa White Star28.68 ± 0.03b22.08 ± 0.03e9.80–50.04.27 ± 0.01a3.40–7.6747.36 ± 0.02a2.50 ± 0.01a1.81 ± 0.01a28.68 ± 0.02bcR. noisettiana x Hot Jewel18.18 ± 0.03e26.06 ± 0.02a13.50–56.704.80 ± 0.01a3.74–5.9830.67 ± 0.03e2.33 ± 0.01a1.50 ± 0.01a32.67 ± 0.03aR. noisettiana x Rosa Sweet Star22.08 ± 0.03cd24.17 ± 0.03bc8.20–43.903.76 ± 0.01a3.20–6.6239.15 ± 0.02c2.45 ± 0.01a1.74 ± 0.01a27.94 ± 0.02cR. noisettiana x Rosa Bling Love Star34.07 ± 0.03a21.86 ± 0.02e8.60–39.204.45 ± 0.01a1.55–11.4741.11 ± 0.01b2.50 ± 0.01a1.98 ± 0.01a25.98 ± 0.01dR. noisettiana x Orange Jewel15.00 ± 0.02f24.60 ± 0.02ab13.50–56.704.80 ± 0.01a3.74–5.9830.67 ± 0.03e2.33 ± 0.01a1.50 ± 0.01a32.67 ± 0.03aR. noisettiana x Rosa Shining Star20.59 ± 0.02d17.03 ± 0.02f11.0–21.03.69 ± 0.01a3.0-5.7140.21 b ± 0.01c2.41 ± 0.01a1.57 ± 0.01a24.57 ± 0.01dR. noisettiana x Rosa Lady Star22.92 ± 0.03c23.09 ± 0.01bc7.80–37.504.34 ± 0.01a3.42–7.3036.43 ± 0.01d2.59 ± 0.01a1.88 ± 0.01a28.73 ± 0.01bcR. noisettiana x Orange Romance21.43 ± 0.02cd25.58 ± 0.02ab11.20–52.03.86 ± 0.01a2.99–5.8335.19 ± 0.01d2.63 ± 0.01a1.83 ± 0.01a30.44 ± 0.01bR. noisettiana x Red Romance21.11 ± 0.01cd21.53 ± 0.02e14.60-35.754.57 ± 0.01a3.57–6.6935.54 ± 0.01d2.23 ± 0.01a2.79 ± 0.01a28.89 ± 0.01bcR. noisettiana x Sparkling Jewel28.79 ± 0.02b22.68 ± 0.02de12.50–34.204.68 ± 0.01a2.36–7.5731.59 ± 0.01e2.47 ± 0.01a1.98 ± 0.01a30.21 ± 0.01bMeans followed by different letters are significantly different according to Duncan’s multiple range test (P < 0.05) (RB (%): Percentage of individuals showing recurrent blooming in each hybrid population., PH Plant height, FD Flower diameter, PN Petal number, LC Leaf color, NFpP Number of flowers per plant, DFB duration of flower blooming) Values are expressed as mean ± standard deviation (SD) Morphological and floral traits of progeny Means followed by different letters are significantly different according to Duncan’s multiple range test (P < 0.05) (RB (%): Percentage of individuals showing recurrent blooming in each hybrid population., PH Plant height, FD Flower diameter, PN Petal number, LC Leaf color, NFpP Number of flowers per plant, DFB duration of flower blooming) Values are expressed as mean ± standard deviation (SD) In comparison, the progenies exhibited a broad range of variation across all recorded traits, reflecting successful genetic recombination. Plant height in the hybrids ranged from 7.80 to 56.70 cm, while flower diameter varied between 1.55 and 11.47 cm, indicating transgressive segregation in several crosses. Petal number also showed substantial variation, with hybrids inheriting intermediate to higher values depending on the specific combination. Although all parental genotypes displayed high recurrent bloom (RB) percentages (typically 100%), the F₁ progenies demonstrated lower values, ranging from 15.0% to 34.07% depending on the cross-combination. However, this observation should be interpreted with caution, as plants derived from seeds often exhibit a juvenile phase and irregular flowering behavior during the early years. In contrast to vegetatively propagated commercial cultivars, which show stable and uniform flowering, seed-derived individuals may require additional time to express their full reproductive potential. Therefore, the reduced RB values observed in the F₁ population may not solely reflect genetic inheritance but could also be influenced by developmental factors. These results indicate that the hybridization between R. noisettiana and modern miniature rose cultivars generated substantial variability in both vegetative and floral traits, reflecting the diverse genetic backgrounds of the parental genotypes and contributing to the enrichment of the progeny pool. The reduction in recurrent bloom (RB) percentages observed in the F₁ hybrids, despite the strong expression of this trait in the parental lines, should be interpreted with caution. Plants derived from seeds often undergo a juvenile phase and may exhibit delayed and irregular flowering in the early years, unlike vegetatively propagated commercial cultivars, which display stable, uniform blooming patterns. Although all parental genotypes exhibited recurrent blooming, only 15–34% of the F₁ progenies expressed this trait. This discrepancy suggests that recurrent blooming in roses is controlled by a complex genetic mechanism rather than simple inheritance. Previous studies have identified RoKSN, a TERMINAL FLOWER1/CENTRORADIALIS homolog, as a major floral repressor associated with once-flowering behavior, whereas RoFT functions as a promoter of floral transition and recurrent flowering [32, 33]. The expression and interaction of these genes play a critical role in determining flowering patterns in rose populations. The reduced frequency of recurrent blooming observed in the present study may therefore reflect segregation of flowering-related loci and variation in their expression among F₁ individuals. In addition to genetic regulation, recurrent blooming is influenced by developmental and environmental factors. Seed-derived rose plants commonly undergo a juvenile growth phase during which flowering behavior may not be fully expressed. Consequently, young F₁ individuals may exhibit irregular or delayed flowering compared with established vegetatively propagated cultivars. Environmental conditions such as temperature, photoperiod, and plant vigor may further influence the expression of recurrent blooming. Therefore, the relatively low proportion of recurrent-flowering individuals observed in the present study likely reflects the combined effects of genetic segregation and developmental regulation within the F₁ population. PH values among the hybrids ranged from 17.03 cm to 26.06 cm, which were notably lower than that of the maternal parent R. noisettiana, a tall climbing rose (~ 110 cm). This suggests that dwarfing genes from the pollen parents, most of which were modern and compact cultivars, exerted a dominant or partially dominant effect on the F₁ offspring. The significant reduction in PH aligns with previous findings indicating that miniature or compact rose types tend to reduce overall plant stature in progeny when crossed with tall-growing types [19]. FD ranged from 3.69 cm to 4.80 cm among the hybrids, representing intermediate values between the large-flowered maternal parent and the smaller-petaled pollen parents. This inheritance pattern suggests additive genetic effects or incomplete dominance for this trait. Similar trends were observed for petal number, with hybrids displaying intermediate or slightly elevated values depending on the cross. The highest petal numbers were recorded in combinations involving pollen parents characterized by dense floral structures, such as Rosa ‘Sweet Star’ and Rosa ‘Bling Love Star’. In terms of the number of flowers per plant, values ranged from 1.50 to 2.79, which are relatively low overall. This outcome may be attributed to the influence of the maternal genetic background (R. noisettiana), which, despite its vigorous vegetative growth, may not strongly promote floral proliferation when used as a seed parent. Additionally, the duration of flower blooming among hybrids varied between 24.57 and 32.67 days, with the most extended durations observed in crosses with ‘Hot Jewel’ and ‘Orange Jewel.’ This finding suggests that these pollen parents contribute positively to floral longevity, a desirable trait in ornamental breeding. The observed variation across these morphological and floral traits underscores the genetic diversity generated through interparietal hybridization, providing valuable insights into patterns of trait inheritance. Nonetheless, the relatively narrow expression ranges in traits such as flower number and petal count indicate possible dominance effects or restricted segregation in the F₁ generation. The parental cultivars used in the hybridization program displayed a broad spectrum of petal colors, including red, pink, white, yellow, orange, and dark pink tones (Table 1). Among them, Rosa noisettiana exhibited red petals. At the same time, the pollen sources presented diverse pigmentation patterns: white in Rosa White Star, pink in Rosa Lady Star and ‘Sparkling Jewel,’ orange in ‘Orange Romance’ and ‘Orange Jewel,’ yellow in Rosa Shining Star, and dark pink in ‘Rosa Bling Love Star.’ This diversity provided a suitable genetic background for examining petal color segregation in the F₁ progenies. Across all crosses, red and pink were the predominant petal colors among the progeny. Hybrids derived from red- or pink-colored parents (e.g., ‘Hot Jewel,’ ‘Sweet Star,’ ‘Lady Star’) primarily produced red and pink offspring, whereas crosses involving more distinct parental colors (e.g., white, yellow, or orange) showed limited transmission of those specific hues to the progeny. Notably, Light colors (white, yellow, orange) were absent or rarely expressed (Fig. 2). These findings suggest that red and pink petal colors display dominance in segregation, whereas lighter colors such as white and yellow may be recessive or polygenically inherited, with limited expression under the current genetic background. The absence of orange coloration in the progeny, despite its presence in certain pollen parents, further underscores the complexity of petal color inheritance in roses and indicates the involvement of potential epistatic or suppressive gene interactions. The predominance of red and pink phenotypes in the F₁ progenies can be attributed to the molecular regulation of pigment biosynthesis. Fig. 2Petal colour of F1 plants obtained according to hybrid combinations Petal colour of F1 plants obtained according to hybrid combinations A heat map with hierarchical clustering was generated to visualize the distribution and relationships of phenotypic traits among the genotypes, including both parental lines and their hybrid progenies. The analysis revealed distinct groupings based on similarities in key agronomic and ornamental traits, such as recurrent bloom percentage, plant height, flower diameter, petal number, leaf color, number of flowers per plant, and petal color. Genotypes were clustered into distinct groups, reflecting varying degrees of phenotypic resemblance. Notably, hybrid combinations generally clustered separately from their respective parental lines, suggesting the presence of transgressive segregation or intermediate trait expression in specific crosses. Both clustering and PCA analyses confirmed the phenotypic divergence between hybrids and their parents. The hierarchical structure observed on both axes (genotypes and traits) further revealed potential correlations among traits. For example, genotypes with larger flower diameters and higher petal numbers were generally associated with greater plant height, indicating a positive phenotypic correlation. In contrast, certain hybrids displayed lower values for traits such as flower count per plant or petal color intensity, underscoring the extent of genetic variability and the potential for trait recombination within the progeny pool. Hierarchical clustering analysis revealed that none of the hybrid combinations clustered within the same group as their respective parental lines. This indicates that the hybrids exhibit distinct phenotypic profiles that diverge from both parents. Such a pattern suggests either intermediate inheritance or transgressive segregation, in which hybrid progenies display trait values that exceed the parental range. This divergence highlights the emergence of novel trait combinations and underscores the genetic variability introduced through hybridization. From a breeding perspective, this finding is particularly significant, as it demonstrates the potential of hybrid genotypes to exhibit unique phenotypes not observed in either parent, thereby offering new opportunities for selecting superior ornamental traits (Fig. 3). Fig. 3The heat map with hierarchical clustering of the parameters analyzed for each progeny of cross-combinations The heat map with hierarchical clustering of the parameters analyzed for each progeny of cross-combinations The prevalence of red and pink flower colours in the F₁ population may be related to the genetic control of the anthocyanin biosynthesis pathway, which is responsible for the formation of red, pink and purple hues in rose petals. Anthocyanin accumulation is regulated by structural genes such as chalcone synthase (CHS), dihydroflavonol 4-reductase (DFR), anthocyanidin synthase (ANS) and flavonoid 3′-hydroxylase (F3’H), as well as transcription factors belonging to the MYB, bHLH and WD40 families [34–36]. The expression levels and interactions of these genes play a decisive role in the formation of pigment density and flower colour. The limited occurrence of white, yellow and orange-coloured individuals in this study may be related to epistatic gene interactions and regulatory mechanisms within pigment biosynthetic pathways. In particular, the presence of alleles supporting anthocyanin synthesis may have led to a higher proportion of red and pink colours in the population. Similarly, the limited expression of carotenoid-derived yellow and orange tones may be explained by the suppression of genes responsible for the formation of these pigments or by more complex inheritance patterns. The results obtained indicate that flower colour inheritance is not a simple trait controlled by a single gene, but rather possesses a complex genetic structure shaped by the interaction of numerous structural and regulatory genes. Principal Component Analysis (PCA) was performed to investigate the multivariate relationships among hybrid combinations and their parental lines based on morphological traits. The first two principal components (PC1 and PC2) explained a substantial proportion of the total variance, enabling an effective two-dimensional visualization of the genotype distribution. The resulting scatter plot clearly separated hybrid genotypes from their respective parents, indicating considerable phenotypic differentiation across most crosses. While some hybrids were positioned closer to one of their parental lines along specific principal components, the overall dispersion pattern reflects the effects of recombination and the occurrence of potential transgressive segregation in the F₁ hybrids (Fig. 4). Fig. 4PCA plot of hybrid combination and parents based on morphological traits PCA plot of hybrid combination and parents based on morphological traits Complementing the PCA visualization, pairwise Pearson correlation analysis was conducted to quantify the phenotypic resemblance between hybrids and their respective parents. The results revealed variable levels of correlation across different hybrid–parent pairs. For example, ‘R × HJ’ exhibited a relatively high correlation with the maternal parent R (r = 0.6868), indicating a stronger phenotypic affinity. In contrast, the same hybrid showed a lower correlation with the paternal parent HJ (r = 0.5313). In contrast, ‘R × RWS’ was more closely aligned with the paternal parent RWS (r = 0.6344) than with R (r = 0.5135). These findings suggest that, in some crosses, hybrid phenotypes tend to approximate the paternal line, while in others they more closely resemble the maternal line. This variability underscores the genetic complexity of the analyzed traits and the differential parental contributions to hybrid trait expression (Table 6). Table 6Pairwise correlation values between F₁ hybrids and parentsHybridParentCorrelationR x RWSR0.514R xRWSRWS0.634R x HJR0.686R x HJHJ0.531R x RSSR0.579R x RSSRSS0.698R x RBLSR0.697R x RBLSRBLS0.822R x OJR0.623R x OJOJ0.420R x RShSR0.438R x RShSRShS0.636R x RLSR0.647R x RLSRLS0.780R x ORR0.647R x OROR0.545R x RRR0.580R x RRRR0.593R x SJR0.724R x SJSJ0.719 Table 6Pairwise correlation values between F₁ hybrids and parentsHybridParentCorrelationR x RWSR0.514R xRWSRWS0.634R x HJR0.686R x HJHJ0.531R x RSSR0.579R x RSSRSS0.698R x RBLSR0.697R x RBLSRBLS0.822R x OJR0.623R x OJOJ0.420R x RShSR0.438R x RShSRShS0.636R x RLSR0.647R x RLSRLS0.780R x ORR0.647R x OROR0.545R x RRR0.580R x RRRR0.593R x SJR0.724R x SJSJ0.719 Pairwise correlation values between F₁ hybrids and parents Conclusion The present study provides important insights into the hybridization potential between R. noisettiana and ten miniature rose cultivars, highlighting promising outcomes for ornamental plant breeding. The high fruit set, seed yield, and phenotypic diversity observed in the F₁ populations demonstrate that such combinations hold considerable potential for the development of novel rose varieties. Traits such as plant height, flower diameter, petal number, and blooming duration exhibited transgressive segregation, indicating the emergence of new phenotypes beyond the parental range. However, the relatively low expression of recurrent blooming in the F₁ progeny (15–34%), compared to the parental genotypes, should be interpreted as preliminary, as trait expression may vary during early growth stages and requires confirmation through extended evaluation.Additionally, the expected variation in flower color, a trait of high commercial importance, was not fully realized, with red and pink hues predominating across hybrid combinations. This outcome suggests a dominant inheritance pattern for these colors, as well as possible epistatic interactions that suppress the expression of lighter or distinct colors. To more effectively transfer ornamental traits in breeding programs, future efforts should prioritize marker-assisted selection targeting recurrent blooming QTLs (e.g., RoKSN, RoFT) and extend evaluations to BC₁ and F₂ populations in order to clarify additive and epistatic genetic effects. In conclusion, the findings demonstrate that intergroup hybridization between miniature and climbing roses is a valuable strategy for generating genetically diverse and morphologically novel genotypes. The present study provides a solid foundation for further breeding efforts aimed at enhancing trait expression and broadening the genetic base of ornamental rose cultivars. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3.