2025 年 94 巻 3 号 p. 374-381
Evergreen azalea cultivars are used as ornamental shrubs and pot plants not only in Japan, but also in Western countries. These cultivars were developed from wild species native to Japan, and through selection and crossing, cultivar groups such as the Ryūkyū-tsutsuji, Hirado-tsutsuji, Edo-kirishima, Satsuki and pot azalea have been developed. In this study, we focused on the genetic contribution of wild species to the development of evergreen cultivars and used microsatellite loci to examine the genetic involvement of wild species, particularly Rhododendron ripense. The utilization of seven microsatellite loci enabled cultivar identification and an estimation of bud mutation lines. The results of STRUCTURE analysis revealed that the R. ripense cluster was predominant in large-flowered cultivars such as Kishi-tsutsuji, Ryūkyū-tsutsuji, and Ōkirishima. The involvement of the R. ripense cluster was also observed in pot azalea developed in Western countries. Additionally, the genetic involvement of R. scabrum was confirmed in many cultivars of Ōkirishima and Hirado-tsutsuji. Regarding small-flowered cultivars, the Yama-tsutsuji and Kurume-tsutsuji cultivar groups were dominated by the R. kaempferi cluster, while the Satsuki cultivar group was dominated by the cluster containing R. indicum, R. eriocarpum, and others. These results enable us to reevaluate the classification of azalea cultivars by conducting a more detailed study of cultivar groups composed of the same clusters.
Evergreen azalea cultivars are used as ornamental shrubs and pot plants not only in Japan, but also in Western countries. These cultivars have primarily been selected since the Edo era (1603–1867) from collections of natural mutants and natural hybrids, as well as from crosses under cultivation (Ito and Creech, 1984; Kobayashi, 2020; Kurashige and Kobayashi, 2008). In particular, the establishment of large-flowered cultivar groups such as Ryūkyū-tsutsuji (Rhododendron × mucronatum G. Don), Ōkirishima, and Hirado-tsutsuji (R. × pulchrum Sweet) has been indicated to involve R. scabrum G. Don, R. macrosepalum Maxim., and R. ripense Makino. Meanwhile, the establishment of small-flowered cultivar groups such as Edo-kirishima and Kurume-tsutsuji (R. obtusum Planch.) has been indicated to involve R. kaempferi Planch. and R. kiusianum Makino, while Satsuki cultivars have been indicated to involve R. indicum (L.) Sweet and R. eriocarpum (Hayata) Nakai (Kunishige and Tamura, 1961; Miyazawa, 1918; Tamura, 1963). These azaleas were introduced to Europe between the 17th and 19th centuries and were crossed with R. simsii Planch., native to East Asia, to breed Belgian-Indian hybrids. These hybrids remain internationally important potted plants today (Galle, 1987).
The genetic participation of wild species in the development of evergreen azalea cultivars has been investigated using AFLP analysis (Scariot et al., 2007), PCR-RFLP analysis of chloroplasts (Kobayashi et al., 2003, 2021), and microsatellite marker analysis (Kobayashi et al., 2017; Miyano et al., 2013; Yamamoto et al., 2019). In particular, PCR-RFLP analysis of chloroplast DNA revealed that R. ripense was an important wild species contributing to the establishment of large-flowered cultivars in Japan (Kobayashi et al., 2021). However, the genetic contribution of R. ripense and other wild species to cultivars remains unclear. In a previous report, we investigated the genetic relationships among wild evergreen azalea species, which are primarily endemic to Japan, using seven microsatellite loci. The results showed that R. ripense is genetically distinct from other wild species, with significant genetic diversity observed among R. ripense populations across geographical distributions. Additionally, genetic contributions from R. ripense were confirmed in seven cultivars, including Kishi-tsutsuji ‘Wakasagi’, Ryūkyū-tsutsuji ‘Shiro-ryūkyū’, and ‘Murasaki-ryūkyū’ (Ohta et al., 2024).
In this study, we aimed to clarify the wild species that have contributed genetically to azalea cultivars, which are used as ornamental plants worldwide. Building on the genetic composition of wild species revealed through microsatellite markers at seven loci in a previous report (Ohta et al., 2024), we conducted Bayesian clustering analysis to evaluate the genetic contribution of wild species, particularly R. ripense, to the development of cultivars across diverse cultivar groups.
The evergreen azalea wild species included in this study consisted of 432 individuals from nine wild species (R. ripense, R. scabrum, R. macrosepalum, R. yedoense Maxim. ex Regel var. yedoense f. poukahanense (H. Lév.) Sugim. ex T. Yamaz., R. eriocarpum, R. indicum, R. simsii, R. kaempferi, and R. kiusianum), as examined in Ohta et al. (2024). The 145 cultivars from 13 cultivar groups, including the seven R. ripense related cultivars used by Ohta et al. (2024), were collected from the Niigata Prefectural Botanical Garden, Koishikawa Botanical Garden of the University of Tokyo, Kurume-shi World Azalea Center, and the Plant Breeding Laboratory, Faculty of Life and Environmental Sciences of Shimane University (Table 1). These also included samples used by Kobayashi et al. (2021).
Evergreen azalea cultivars used in this study.
Total genomic DNA was extracted from approximately 70 mg of −80°C freeze fresh leaves of each plant by a modified CTAB method (Kobayashi et al., 1998).
The microsatellite markers used in this study were the seven loci (AZA002, 003, 008, 010, 011, RM2D2, and N25) described in Ohta et al. (2024). PCR was performed in 10 μL reaction mixtures containing 5 ng of genomic DNA, 1 × reaction buffer (Contains MgCl2), 0.2 mM dNTPs, 0.25 U of Blend taq (Toyobo Co., Ltd., Osaka, Japan), and 0.2 μM of each primer. The PCR amplification of the seven loci followed the method described by Ohta et al. (2024). The fragment length of PCR-amplified products was determined by comparison with the GeneScan 500LIZ Size Standard (Applied Biosystems, CA, USA) using an ABI PRISM 3100xl Genetic Analyzer (Applied Biosystems) and Peak scanner ver. 3.0.3-PRC-build4 (Applied Biosystems).
Statistical analysesThe origin of the cultivars was estimated using STRUCTURE ver. 2.3.4 (hereafter STRUCTURE analysis; Pritchard et al., 2000), following the method of Ohta et al. (2024). The STRUCTURE analysis was performed in two steps. First, in order to obtain data suitable for estimating the origin of cultivars, we performed STRUCTURE analysis using only wild species individuals. Next, we examined the genetic relationship of wild species to cultivars using the USEPOPINFO option in the STRUCTURE analysis. STRUCTURE analysis was performed using pre-clustering information for all individuals of the selected wild species (POPFLAG = 1), and 145 cultivars were of unknown origin (POPFLAG = 0). K was fixed at the optimal values determined in the previous step, and the genetic composition of POPFLAG = 0 was assessed based on the genotype information of POPFLAG = 1. Individual wild species used for USEPOPINFO were those with q-values ≥ 80% in the most dominant cluster.
Microsatellite analysis was performed on the cultivars using seven microsatellite loci. The results yielded fragment sizes ranging from 131 bp (RM2D2) to 311 bp (AZA002), with the number of alleles at each locus ranging from 19 (AZA011, and N25) to 57 (AZA002). Allelic polymorphisms at seven loci enabled us to identify all the varieties used in this study, except for the bud mutation cultivar lines.
The genetic composition of cultivars was estimated through STRUCTURE analysis using the data from Ohta et al. (2024). The optimal K values were determined to be K = 2 and 4 based on ΔK, and K = 10 based on log-likelihood. We used 423, 375, and 271 individuals from a total of 432 wild species with q-values ≥ 80% at K = 2, 4, and 10, respectively.
At K = 2, the wild species cluster was divided into the R. ripense population and other wild clusters. At K = 4, the R. ripense population was further subdivided into two clusters, while the remaining wild species clusters were categorized into large-flower (R. scabrum and R. macrosepalum) and small-flower (R. yedoense, R. eriocarpum, R. indicum, R. simsii, R. kaempferi, and R. kiusianum) clusters. At K = 10, the R. ripense population was separated into five clusters, while the other wild species clusters were divided into five clusters, including the 1) R. scabrum cluster, 2) R. macrosepalum cluster, 3) R. indicum, R. eriocarpum and R. simsii cluster (in / er / si cluster), 4) R. kaempferi cluster from the Kyushu region and R. kiusianum cluster (ka-Ky / ki cluster), and 5) R. kaempferi cluster from the Kanto region (ka-Ka cluster).
The genetic composition of cultivars was estimated by referencing the genotypes of wild species, and a bar chart was generated to illustrate the results (Fig. 1; Table S1). The proportion of values (q-values) derived from each individual is summarized for each cultivar group (Table 2).
Bar chart for K = 2, 4, and 10 of cultivars based on wild species. The bar chart follows the same cultivar order as in Table S1, and the number on the left side of the figure represents the code.
The q values of K = 2, 4, or 10 for each evergreen azalea cultivar group estimated by STRUCTURE analysis.
At K = 2, large-flowered cultivar groups such as Kishi-tsutsuji and Ōkirishima exhibited dominance of the R. ripense cluster, with q-values ≥ 65% (Table 2).
At K = 4, the small-flowered cultivar groups, such as Edo-kirishima and Yama-tsutsuji showed q-values ≥ 70% dominated by the small-flowered wild species cluster for many cultivars (Table S1).
At K = 10, the results indicated a genetic contribution of species to each of the cultivar groups. In the Ryūkyū-tsutsuji cultivar group, cultivars such as ‘Fujiman-yo’, ‘Murasaki-ryūkyū’, and ‘Shiro-ryūkyū’ exhibited a predominant genetic contribution from the R. ripense cluster, with q-values ≥ 90%. However, some cultivars, including ‘Shiroman-yo’ and ‘Ryūkyūshibori’ displayed approximately 50% of their q-values in the ka-Ky / ki and ka-Ka clusters (Table S1). In the Mochi-tsutsuji cultivar group, the R. macrosepalum cluster accounted for approximately 40% of the q-values for ‘Kochōzoroi’ and ‘Seigaiha’ among the 10 cultivars. However, half of the cultivars, such as ‘Hanaguruma’ and ‘Surugaman-yō’ had q-values below 20% for the R. macrosepalum cluster (Fig. 1; Table S1). In the Ōkirishima and Hirado-tsutsuji cultivar groups, the R. ripense and R. scabrum clusters comprised q-values ≥ 85% of the total (Table 2). Ōyama-tsutsuji (R. transiens) and small-flowered cultivars, such as Edo-kirishima and Yama-tsutsuji, were predominantly composed of the R. kaempferi cluster. The q-values for the Satsuki cultivar group accounted for 27% of the in / er / si cluster and 46% of the ka-Ky / ki and ka-Ka clusters (Table 2). In pot azalea, q-values accounted for approximately 28% in the R. ripense cluster, 23% in the in / er / si cluster, and 32% in the ka-Ky / ki and ka-Ka clusters (Table 2).
Microsatellite analysis using the seven loci employed in the previous report (Ohta et al., 2024) yielded 19 (AZA011, N25) to 57 (AZA002) alleles per locus, providing essential genotype information for cultivar identification used in this study. We then utilized these markers to examine the contribution of wild species to cultivars and to investigate the relationships among cultivars. As a result, genotype matches were identified for eight cultivars within the Ōkirishima cultivar group, with the exception of ‘Yukidōji’. These cultivars are considered bud mutation cultivars, exhibiting variations in flower color and shape based on their morphological characteristics (Tamura, 1963). The results of this study further support the possibility of bud mutation cultivars, as no differences were observed at the examined loci. Additionally, genotype matches were identified in some Ryūkyū-tsutsuji and Kurume-tsutsuji cultivars (Table S2). Furthermore, the morphological characteristics of the Kurume-tsutsuji cultivars ‘Kokuryū’ and ‘Hinokuni’, as well as Edo-kirishima ‘Hinodekirishima’ and Kurume-tsutsuji ‘Hinodegiri’, were very similar. These findings indicate that these cultivars may be synonymous cultivars. However, the seven microsatellite loci used in this study are limited, and more accurate determinations could be made by increasing the number of loci or by developing SNP markers.
Genetic contribution of wild species to large-flower cultivarsThe genetic composition of wild species reported by Ohta et al. (2024) was applied to the analysis of cultivar data. Consistent with the previous study, the origins of cultivars in the Kishi-tsutsuji and Ryūkyū-tsutsuji cultivar groups were primarily influenced by the genetics of R. ripense.
At K = 2, half of the Mochi-tsutsuji cultivars showed genetic involvement of R. ripense in their genetic composition. Among these, ‘Hanaguruma’ was dominated by clusters of R. ripense with q-values ≥ 90%, confirming the low involvement of R. macrosepalum. Kobayashi et al. (2021) conducted cpDNA analysis, revealing that the Mochi-tsutsuji cultivar group possesses a non-R. ripense type cpDNA (Table S1). In contrast, the findings obtained from the nuclear-derived microsatellite markers used in this study indicate a different genetic composition. Therefore, it is possible that the genetic component of ‘Hanaguruma’ was derived from R. ripense, which is likely the pollen parent. The results at K = 10 also suggested that many Mochi-tsutsuji cultivars, such as ‘Surugaman-yō’ and ‘Gin’nozai’, had low genetic components exclusively from R. macrosepalum, indicating that these cultivars are likely interspecific hybrids.
The genetic components of Ōkirishima and Hirado-tsutsuji were found to be composed of R. ripense and R. scabrum at K = 10. Ōkirishima ‘Ōmurasaki’ has large flowers and exhibits strong environmental tolerance, including winter hardiness and salt tolerance. Its petals are thick, and the flower color is red-purple with a bright red blotch. It is thought to be derived from a cross between R. ripense, R. × mucronatum, and the large red-flowered R. scabrum (Tamura, 1963). Additionally, high adaptability to environmental conditions is reported to be a trait derived from R. ripense (Kobayashi et al., 2010a, b). Hirado-tsutsuji exhibits large flowers in a wide range of colors, including red, pink, purple, vermillion, and white, with petal shapes varying from sword-like to round forms. Similar to Ōkirishima, it is an interspecific hybrid obtained through natural hybridization of R. ripense, R. scabrum, and cultivars such as ‘Ōmurasaki’ and R. × mucronatum, which is believed to have contributed to the diverse flower colors and shapes (Aburaya and Kitamura, 1983). PCR-RFLP analysis of the trn L–trn F region of cpDNA revealed that all of the Ōkirishima cultivars possess R. ripense type cpDNA (Table S1). On the other hand, more than half of the Hirado-tsutsuji cultivars tested were found to have non-R. ripense type cpDNA (Kobayashi et al., 2021). The flower color pigment composition and anthocyanin biosynthesis genes of Hirado-tsutsuji suggest that it may be a cross between R. scabrum and ‘Shiro-ryūkyū’ that is genetically related to R. ripense (Meanchaipiboon et al., 2020). Additionally, clusters dominated by R. ripense were found in some cultivars, such as the Hirado-tsutsuji ‘Hakuō’. Therefore, the results of the STRUCTURE analysis support the proposed breeding history of Hirado-tsutsuji and Ōkirishima, suggesting that these cultivars are hybrids between R. ripense or related cultivars and R. scabrum.
Genetic contribution of wild species to small-flower cultivarsFor small-flowered cultivars such as Edo-kirishima, Kurume-tsutsuji, and Yama-tsutsuji, the K = 10 results indicated genetic involvement of R. kaempferi from the Kyushu region and R. kiusianum, R. kaempferi from the Kanto region, or both. In previous PCR-RFLP analysis of the 16S rRNA region of cpDNA, many Edo-kirishima and Kurume-tsutsuji cultivars showed a band pattern of the R. kaempferi type. Additionally, in some Kurume-tsutsuji cultivars, the band pattern of R. kiusianum has been confirmed (Kobayashi et al., 2003). However, the microsatellite markers used in this study could not determine the detailed origin of cultivars such as Edo-kirishima and Kurume-tsutsuji, since R. kaempferi from the Kyushu region and R. kiusianum belong to the same cluster. Among them, genetic involvement of R. ripense was supported in the Kurume-tsutsuji ‘Tago-no-ura’. The morphological characteristics of the dorsal leaf surface in the Kurume-tsutsuji cultivar group suggest that ‘Tago-no-ura’ has a genetic relationship with R. ripense or R. × mucronatum (Okamoto et al., 2000). Moreover, Kobayashi et al. (2021), in their PCR-RFLP analysis, also indicated that it possesses R. ripense type cpDNA. Genetic components of R. ripense were identified in several Kurume-tsutsuji cultivars, such as ‘Murasakimino’, ‘Wakakaede’ and ‘Tago-no-ura’, suggesting the genetic involvement of R. ripense and related cultivars.
Satsuki cultivars were developed based on R. indicum and R. eriocarpum (Yamazaki, 1996). The results of this study revealed that most of the cultivars, with the exceptions of ‘Chihiro’ and ‘Ōryu’, were genetically derived from the clusters to which R. indicum and R. eriocarpum belong.
R. × hannoense Nakai is presumed to be a hybrid between R. kaempferi and R. indicum (Yamazaki, 1996), and the results presented in this study support previous studies. The genetic composition of ‘Amagibenichōju’ and ‘Komachibotan’ was dominated by R. kaempferi and R. indicum. While the genetic composition of ‘Bungoshiki’ and ‘Sekimori’ showed a low presence of R. indicum, indicating that R. kaempferi from the Kyushu region constituted the genetic composition of these cultivars. In addition, the K = 2 result for ‘Otomekatsura’ suggested the genetic involvement of R. ripense.
Pot azalea is presumed to be a hybrid between a Japanese cultivar and R. simsii (Galle, 1987). Kobayashi et al. (2021) revealed that 12 of the 15 pot azalea cultivars used in this study, with the exceptions of ‘Ward’s Ruby’ and two Noria azalea cultivars, had cpDNA of the R. ripense type. These results showed that most of the pot azalea cultivars used here had genetic involvement of large-flowered species such as R. ripense, R. scabrum, and R. macrosepalum. In AFLP analyses R. × mucronatum and Hirado-tsutsuji were reported to contribute to the Belgian pot azalea (Scariot et al., 2007). The results of this study supported the findings of Scariot et al. (2007) and Kobayashi et al. (2021), suggesting that R. ripense and its related Hirado-tsutsuji have made a genetic contribution to the establishment of pot azaleas.
Conclusion and prospective researchIn this study, we evaluated the genetic contribution to cultivars using genetic information on wild species reported in previous studies. The results suggested that R. ripense contributed significantly to the establishment of large-flowered cultivars such as Ryūkyū-tsutsuji, Mochi-tsutsuji, and Hirado-tsutsuji, as well as to some pot azalea cultivars. Therefore, R. ripense is the most important wild species in the development history of evergreen azalea cultivars. Furthermore, since the genetic components of Hirado-tsutsuji and Ōkirishima were composed of the same clusters, a more detailed study is required to reassess the classification of azalea cultivars. In addition, the genetic contribution of wild species to small-flowered cultivars has not been clearly evaluated.
In the future, we plan to expand the number of materials for large-flowered and small-flowered cultivars, as well as Western pot azalea cultivars. Through the use of RAD-seq analysis to efficiently obtain SNP data, we aim to elucidate more detailed genetic relationships, determine the detailed origin of cultivars, and reconstruct their group classifications.
The authors also thank the faculty of Life and Environmental Sciences in Shimane University for financial support in publishing this report.
