Skip to main content
NCBI home page
PhytoKeys logoLink to PhytoKeys
. 2026 Mar 6;271:187–204. doi: 10.3897/phytokeys.271.175682

Morphological variation in the Rhododendron pseudochrysanthum Hayata (Ericales, Ericaceae) species complex from Taiwan

Ya-Rong Zheng 1, Yen-Hsueh Tseng 1,2, Hsy-Yu Tzeng 1,✉
PMCID: PMC12988429  PMID: 41835233

Abstract

Our study examines the morphological and statistical differentiation within the Rhododendron pseudochrysanthum species complex through comparative analyses of macro- and micro-morphological characters. Using significance testing and cluster analysis, our results demonstrate that R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki is distinct from other members of the complex, namely R. morii Hayata, R. pseudochrysanthum Hayata, and R. hyperythrum Hayata. This taxon is characterized by glabrous mature leaves with revolute margins, larger flower buds with an elongated conical shape, larger pollen and seed sizes, and distinct pollen and seed morphology. Furthermore, R. pseudochrysanthum ssp. morii var. taitunense exhibits a restricted and localized distribution, occurring exclusively in low-elevation mountainous areas of Northern Taiwan. Overall, our results indicate that this taxon has undergone clear morphological differentiation from other taxa within the R. pseudochrysanthum species complex.

Key words: Flower buds, pollen, seeds, statistical analysis, taxonomy

Introduction

Rhododendron subg. Hymenanthes (Blume) K.Koch belongs to the genus Rhododendron L. in the family Ericaceae Juss. They are also known as elepidote rhododendrons and they account for approximately one-quarter of the species diversity within the genus. This subgenus is classified into one section and 24 subsections (Chamberlain 1982), comprising around 270 species worldwide. It is primarily distributed in Asia, with additional occurrences in Europe and North America. China has 259 Rhododendron subg. Hymenanthes species (Fang et al. 2005). Most phylogenetic studies indicate that this subgenus forms a monophyletic group (Kron 1997; Kurashige et al. 1998, 2001; Milne 2004).

Rhododendron rubropunctatum Hayata was described by Hayata (1913) based on a specimen collected by S. Sasaki from Chihsingshan in Northern Taiwan. This taxon was later reported from Taiwan by several authors, including Kanehira (1917), Sasaki (1928), Masamune (1936), Kanehira (1936), and Masamune (1954). However, Wilson (1925), Li (1963), Hsu (1973), Ying (1976), and Li (1978) observed that both R. rubropunctatum Hayata and R. hyperythrum Hayata possess prominent glandular structures on the abaxial leaf surface, and consequently treated R. rubropunctatum Hayata as a synonym of R. hyperythrum Hayata.

In 1981, Yamazaki noted that a Rhododendron species occurring in the Tatun Volcanic Group of Northern Taiwan, including on the slopes of Chihsingshan and Tatunshan, was initially published as R. rubropunctatum by Hayata (1913). He pointed out that this name is a later homonym of R. rubropunctatum H. Lév. & Vaniot (1911), and thus, illegitimate. Yamazaki also noted that Wilson (1925) replaced the name with R. hyperythrum Hayata, which led to further taxonomic confusion (Yamazaki 1981). Nevertheless, due to the revolute leaf margins distinct from other taxa, Yamazaki renamed the taxon as R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki.

Later, Lu and Yang (1989) and Li et al. (1998) considered the reddish glands on the abaxial leaf surface to lack taxonomic significance and treated R. rubropunctatum Hayata, R. morii Hayata, and R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki as synonyms of R. pseudochrysanthum Hayata. In the same treatment, Li et al. (1998) recognized only three taxa in the Rhododendron subgenus Hymenanthes in Taiwan: R. formosanum Hemsl., R. hyperythrum Hayata, and R. pseudochrysanthum Hayata.

Furthermore, due to continuous morphological variation and the absence of clear genetic differentiation among R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki, R. morii Hayata, R. pseudochrysanthum Hayata, and R. hyperythrum Hayata, these four taxa have been collectively referred to as the Rhododendron pseudochrysanthum species complex (Lu and Yang 1989; Chung et al. 2007; Hsieh 2007; Liang 2011; Chen et al. 2014; Cao et al. 2022).

Pollen morphology is considered to exhibit a high degree of genetic conservatism, with intrinsic features such as shape, surface ornamentation, and the number of germination pores commonly used for plant classification and identification (Zhou et al. 2008). Moreover, it is informative for both taxonomic delimitation and phylogenetic analysis within certain groups of Ericaceae (Kron et al. 2002; Sarwar 2007). Within a single subgenus, pollen characters can be particularly valuable for the recognition of sections and subsections, for instance, in Rhododendron sect. Ponticum G. Don, although all taxa produce tetrad pollen grains, variations in pollen size and exine sculpture have been reported, reflecting meaningful interspecific differentiation (Huang 1972; Vasanthy and Pocock 1987; Fuhsiung et al. 1995; Mao et al. 2000; Terzioğlu et al. 2001; Gao et al. 2002; Wang et al. 2006; Zhang et al. 2009; Miyoshi et al. 2011). In Rhododendron species, seed morphology, size, and the presence or absence of appendages are closely related to plant habits, primarily reflecting the systematic differentiation among taxa within the genus (Geng 2014). Moreover, the width and shape of seed wings vary significantly across different groups (Ding et al. 1995; Wang et al. 2007).

Previous studies on the Rhododendron pseudochrysanthum species complex have focused primarily on external morphological descriptions. In this study, we re-examine morphological differentiation within the complex by incorporating a broader set of macromorphological characters, as well as pollen and seed morphological traits.

Materials and methods

Samples collection and morphological comparison

The materials used in this study were primarily collected from wild populations of the R. pseudochrysanthum species complex in Taiwan (Table 1). Fieldwork included photography, GPS point recording, and fixation of fresh leaves and flowers. External morphological traits, such as leaves, flowers, floral buds, and seeds, were observed and compared among the different taxa. Macroscopic features were documented using a digital camera equipped with a macro lens (Nikon D7500 + Nikon AF-S DX NIKKOR 40 mm f/2.8G Micro). At the same time, finer structures were examined under a high-magnification stereomicroscope (KEYENCE VHX-7020).

Table 1.

Materials collected of the R. pseudochrysanthum species complex.

taxon location code lat. (°N), long. (°E) altitude (m) Specimens examined
R. pseudochrysanthum ssp. morii var. taitunense Mt. Qixing taitun_QXS 25.17°N, 121.55°E 972 ZYR298
Mt. Wuzhi taitun_WZS 25.14°N, 121.62°E 635 ZYR257
Mt. Bijia taitun_BJS 24.97°N, 121.62°E 580 ZYR266
Mt. Caigongkeng taitun_CGK 25.19°N, 121.52°E 864 ZYR271
Mt. Banping taitun_BPS 25.10°N, 121.87°E 621 ZYR332
Mt. Canguangliao taitun_CGL 25.09°N, 121.87°E 717 ZYR270
Mt. Xin taitun_XS 25.13°N, 121.63°E 481 ZYR335
R. morii Mt. Jiali morii_JLS 24.51°N, 121.01°E 2,026 ZYR30
Yushan Trail 5 km morii_YS 23.47°N, 120.93°E 2,963 ZYR280
Tataka morii_TTK 23.48°N, 120.89°E 2,820 ZYR275
Mt. Taman morii_TMS 24.71°N, 121.45°E 2,109 ZYR222
Mt. Xinan morii_XNS 23.09°N, 120.81°E 2,439 ZYR345
Beinanzhu morii_BNS 23.07°N, 120.86°E 2,551 ZYR337
Tefuya morii_TFY 23.47°N, 120.82°E 2,580 ZYR375
Mt. Xue 4 km morii_XUE 24.39°N, 121.28°E 3,007 ZYR69
Mt. Taiping morii_TPS 24.49°N, 121.53°E 1,979 ZYR66
R. pseudochrysanthum Hehuan East Peak pseudo_HHE 24.14°N, 121.28°E 3,353 ZYR297
Hehuan Road Side pseudo_HHR 24.16°N, 121.28°E 3,084 ZYR350
Mt. Xue pseudo_XUE 24.39°N, 121.24°E 3,596 ZYR113
Yushan Main Peak pseudo_YS 23.47°N, 120.96°E 3,834 ZYR86
Nanhu Main Peak pseudo_NHM 24.36°N, 121.44°E 3,709 ZYR366
R. hyperythrum Nanhu Cirque hypery_NHC 24.37°N, 121.44°E 3,400 ZYR364
Nanhu 18.6 km hypery_NH18k 24.38°N, 121.44°E 3,460 ZYR367
Nanhu 17 km hypery_NH17k 24.38°N, 121.43°E 3,320 ZYR370
Shenmazheng hypery_SMZ 24.38°N, 121.44°E 3,457 ZYR362
Wuyan Peak hypery_WYF 24.38°N, 121.44°E 3,544 ZYR369
Open in a new tab

Herbarium resources

Herbarium acronyms from Index Herbariorum were used in this study (Thiers 2026, continuously updated). Voucher specimens collected for the current study were deposited in the herbarium of the Department of Forestry, National Chung-Hsing University, Taiwan (TCF). Physical and digital specimens in several herbaria were also examined; physical specimens: Taiwan Forestry Research Institute (TAIF), Herbarium of Taiwan Biodiversity Research Institute, Taiwan (TAIE), National Taiwan University, Taiwan (TAI), National Chung-Hsing University, Taiwan (TCF); and digital specimens: Department of Forestry and Natural Resources, National Chia-Yi University, Taiwan (CHIA), Provincial Pingtung Institute, Taiwan (PPI), the Royal Botanic Gardens, Kew (K), Herbarium of Academia Sinica, Taiwan (HAST), and Herbarium of the University of Tokyo, Japan (TI).

Pollen morphology

Fresh stamens were collected and fixed in 70% ethanol. During the collection, pollen was extracted from the anthers and placed into a 1.5 ml centrifuge tube containing 70% ethanol. After a single centrifugation step, the supernatant was removed, and a mixture of 30 ml 2,2-dimethoxypropane (DMP) and 1 drop of 1M HCl was added. The sample was incubated for approximately 2 hours, followed by an acetone substitution for at least 1 hour. Subsequently, the samples underwent critical point drying (CPD) (Halbritter 1998). The mounted samples were then gold-coated for 100 seconds at 15 mA using a Quorum SC7620 sputter coater (Quorum Technologies, Laughton, UK). Finally, observations and imaging were performed using a scanning electron microscope (Hitachi S-3400N) at 15 kV.

The morphological characteristics of pollen were described with reference to Erdtman (1952), Sarwar (2007), and Sarwar and Takahashi (2013). Observations included the shape and size of tetrad pollen, as well as the following measurements (in µm): Tetrad diameter (D), Polar length (P), Equatorial diameter (d for tetrads; E for monads), Colpus length (2f for tetrads; L for monads), P/d ratio (polar to equatorial diameter ratio), D/d ratio (tetrad diameter to equatorial diameter ratio), 2f/D ratio (colpus length to tetrad diameter ratio), and Apocolpial exine sculpture.

Seed morphology

Mature dehiscent capsules of each taxon were collected from various sampling sites. A total of 3–5 capsules were placed in an oven at 50 °C for at least 48 hours. Thirty mature seeds were randomly sampled for each taxon and observed under a high-magnification dissecting microscope (KEYENCE VHX-7020) to examine and document seed size and appendages. The morphological characteristics of the seeds were described with reference to Barthlott (1981, 1990) and Geng (2014), and the following measurements (in mm) were noted: seed length (L) and seed width (W).

Distribution map

The collection site information of this study, including GPS coordinates and elevation, was integrated. The distribution map was adapted from Lin (2018), and the distribution points of each taxon were plotted using QGIS ver. 3.4. Additionally, the study referred to the geographical and climatic regions of Taiwan as defined by Su (1985) and the montane vegetation zones to analyze the vertical and horizontal distribution patterns of each taxon.

Data analysis

Quantitative traits were measured using ImageJ software. Statistical analyses were conducted in R 4.3.2 using the stats package (R Core Team 2024). First, normality was assessed using the Shapiro-Wilk test (shapiro.test) to determine whether the quantitative trait values followed a normal distribution across taxa (p < 0.05 indicates non-normal distribution). If traits followed a normal distribution, one-way ANOVA was performed using the aov function, followed by a Tukey HSD post-hoc test (TukeyHSD, p ≤ 0.05) to assess significant differences among taxa. If traits did not follow a normal distribution, nonparametric analysis was conducted using the Kruskal-Wallis test (kruskal.test) to determine differences among taxa, followed by a pairwise Wilcoxon rank-sum test (pairwise.wilcox.test, p ≤ 0.05) for post-hoc comparisons. Boxplots were generated using the ggboxplot function from the ggplot2 package (Wickham 2016) to visualize data distribution.

To investigate the clustering results of taxa based on characteristic variables such as pollen and seeds, cluster analysis was conducted using R 4.3.2. The stats package (R Core Team 2024) was used with the dist function to construct a distance matrix based on Euclidean distance to determine the proximity between data points. Hierarchical cluster analysis was then performed using the hclust function with Ward’s minimum variance method. Finally, a dendrogram was plotted using the plot function from the graphics package (R Core Team 2024).

Results and discussion

Macromorphological differences

The leaf morphology of the R. pseudochrysanthum species complex is highly variable (Fig. 1, Table 2), exhibiting shapes such as narrow elliptic, lanceolate, oblanceolate, and narrow ovate. Leaf base morphology is also diverse, and many traits show considerable intraspecific variation. Although leaf length differs significantly among some taxa (Fig. 4, Table 2), leaf size and shape can be influenced by habitat conditions or climate, and even within a single individual, variation may occur due to light exposure (Geng 2014). Therefore, leaf size, shape, and base morphology alone cannot be used as reliable taxonomic criteria. In terms of epidermal appendages, glands were present across all four taxa, consistent with findings by Lu and Yang (1989), indicating that this trait is not a valid diagnostic character. However, trichomes exhibited high diversity among the four taxa, suggesting that this macromorphological feature may serve as a useful characteristic for taxonomic identification.

Figure 1.

Figure 1.

Open in a new tab

Leaf morphology of Rhododendron pseudochrysanthum species complex. A. R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki; B. R. morii Hayata; C. R. pseudochrysanthum Hayata; D. R. hyperythrum Hayata.

Table 2.

Macromorphological differences within the Rhododendron pseudochrysanthum species complex.

taxon R. pseudochrysanthum ssp. morii var. taitunense R. morii R. pseudochrysanthum R. hyperythrum
traits
Leaf length 9.482 ± 1.203a 9.264 ± 1.641a 5.272 ± 0.761b 5.370 ± 0.988b
Leaf width 2.465 ± 0.589ab 2.713 ± 0.484b 2.080 ± 0.408c 2.371 ± 0.393a
Leaf length/width 4.141 ± 1.186a 3.454 ± 0.367b 2.634 ± 0.307c 2.287 ± 0.383b
Leaf lateral vein pairs 10–15 12–19 12–15 11–14
Leaf shape narrowly elliptic- oblanceolate narrowly elliptic- lanceolate narrowly ovate narrowly ovate
Leaf margin strongly revolute flat to revolute revolute revolute
Leaf base obtuse-acute obtuse-acute obtuse-acute obtuse-cordate
Leaf trichome hairless tufted hair on abaxial midrib glandular hair on abaxial midrib branched hair on leaf abaxial
Flower bud length 3.615 ± 0.372a 2.170 ± 0.288b 2.019 ± 0.141b 1.928 ± 0.178b
Flower bud width 1.554 ± 0.240a 1.260 ± 0.083b 1.165 ± 0.096c 1.215 ± 0.069b
Flower bud length/width 2.367 ± 0.331a 1.722 ± 0.190bc 1.742 ± 0.162b 1.592 ± 0.173c
Style hair type glandular hairs glandular or puberulent glandular or puberulent glandular or puberulent
Ovary hair type glandular hairs glandular or puberulent glandular or puberulent glandular or puberulent
calyx lobes 5, triangular, with glandular hairs lobes 5, triangular, with glandular hairs lobes 5, triangular, with glandular hairs lobes 5, triangular, with glandular hairs
Pedicel hair type glandular hairs glandular or puberulent glandular or puberulent glandular or puberulent
Locule number 6–10 6–8 6–7 6–7
Stamen number 10–12 10–12 10–12 10–12
Open in a new tab

Mean ± SE (n = 5); Unit = centimeter (cm). Different letters indicate statistically significant differences (p < 0.05).

Figure 4.

Figure 4.

Open in a new tab

Boxplot comparing the leaf and flower bud characteristics of Rhododendron pseudochrysanthum species complex. Different letters in each boxplot indicate statistically significant differences (p < 0.05).

Within the R. pseudochrysanthum species complex, traits such as the indumentum on the style, ovary, calyx, pedicel, and stamens, as well as the number of ovary locules, exhibit variation among individuals within each taxon. Floral color also varies depending on environmental conditions. However, R. pseudochrysanthum ssp. morii var. taitunense consistently possesses a higher number of ovary locules compared to the other taxa. (Fig. 3, Table 2). Therefore, some of these characters are not suitable as diagnostic features for species delimitation within the species complex. However, in terms of floral bud size, R. pseudochrysanthum ssp. morii var. taitunense displays significantly greater bud length, width, and length-to-width ratio (p < 0.05) compared to the other taxa (Figs 2, 4). In contrast, R. morii, R. pseudochrysanthum, and R. hyperythrum show no significant differences in floral bud dimensions (Fig. 4, Table 2).

Figure 3.

Figure 3.

Open in a new tab

Flower morphology of Rhododendron pseudochrysanthum species complex. A. R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki; B. R. morii Hayata; C. R. pseudochrysanthum Hayata; D. R. hyperythrum Hayata: 1, flower; 2, style; 3, ovary; 4, ovary locules; 5, calyx; 6, pedicel.

Figure 2.

Figure 2.

Open in a new tab

Flower bud morphology of Rhododendron pseudochrysanthum species complex. A. R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki; B. R. morii Hayata; C. R. pseudochrysanthum Hayata; D. R. hyperythrum Hayata.

Pollen morphology

The pollen grains of the R. pseudochrysanthum species complex are all tetrads, arranged in a tetrahedral configuration, and exhibit heteropolarity (Fig. 5). The average diameter of the tetrads ranges from 40.243 ± 0.848 µm to 45.090 ± 1.096 µm; the average polar axis length ranges from 19.525 ± 0.465 µm to 21.804 ± 0.638 µm; the average equatorial diameter ranges from 29.469 ± 1.464 µm to 32.462 ± 1.622 µm; and the average colpus length ranges from 14.035 ± 1.362 µm to 19.723 ± 0.764 µm (Table 3). The P/d ratio ranges from 0.666 ± 0.038 to 0.687 ± 0.020; the D/d ratio ranges from 1.363 ± 0.042 to 1.393 ± 0.062; and the 2f/D ratio ranges from 0.350 ± 0.037 to 0.439 ± 0.019 (Table 3). According to Erdtman (1952), based on the classification of pollen grain size and shape, these pollen grains are categorized as medium-sized (Mediae, 25–50 µm) and oblate in shape. They are 3-colporate, with a circular outline in polar view.

Figure 5.

Figure 5.

Open in a new tab

Pollen morphology of the Rhododendron pseudochrysanthum species complex. A. R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki; B. R. morii Hayata; C. R. pseudochrysanthum Hayata; D. R. hyperythrum Hayata: 1, tetrads polar plane; 2, tetrads equatorial plane; 3, tetrads polar plane; 4, apocolpial exine sculpture.

Table 3.

Pollen morphological differences of the R. pseudochrysanthum species complex.

taxon R. pseudochrysanthum ssp. morii var. taitunense R. morii R. pseudochrysanthum R. hyperythrum
traits
Tetrad diameter (D) 45.090 ± 1.096a 40.988 ± 0.896b 40.369 ± 0.882bd 40.243 ± 0.848d
Polar length (P) 21.804 ± 0.638a 20.050 ± 0.569b 19.525 ± 0.465c 20.325 ± 0.403b
Equatorial diameter (d) 32.462 ± 1.622a 29.894 ± 1.139b 29.469 ± 1.464b 29.621 ± 0.675b
Colpus length (2f) 19.723 ± 0.764a 16.122 ± 0.746b 15.991 ± 1.332b 14.035 ± 1.362c
P/d ratio 0.675 ± 0.048ab 0.673 ± 0.034ab 0.666 ± 0.038a 0.687 ± 0.020b
D/d ratio 1.393 ± 0.062a 1.374 ± 0.045a 1.376 ± 0.050a 1.363 ± 0.042a
2f/D ratio 0.439 ± 0.019a 0.394 ± 0.016b 0.397 ± 0.032b 0.350 ± 0.037c
apocolpial exine sculpture gemmate gemmate gemmate gemmate
Open in a new tab

Mean ± SE (n = 10); Unit = micrometer (μm). Different letters indicate statistically significant differences (p < 0.05).

Overall, R. pseudochrysanthum ssp. morii var. taitunense exhibits significant differences (p < 0.05) in all quantitative pollen traits (except for the P/d and D/d ratios) compared to the other three taxa within the R. pseudochrysanthum species complex (Fig. 6). According to the study by Sarwar and Takahashi (2013), the apocolpial exine sculpture of pollen in the R. pseudochrysanthum species complex is consistently of the gemmate type (Fig. 5).

Figure 6.

Figure 6.

Open in a new tab

Boxplot comparing the pollen grain characteristics of the R. pseudochrysanthum species complex. Different letters in each boxplot indicate statistically significant differences (p < 0.05).

Seed morphology

The fruit of the R. pseudochrysanthum species complex is a septicidal capsule, and the seeds are elongate or narrowly elliptic in shape (Fig. 7). The average seed length ranges from 1.062 ± 0.095 mm to 1.371 ± 0.146 mm, and the average seed width ranges from 0.552 ± 0.121 mm to 0.662 ± 0.058 mm (Table 4). Within the R. pseudochrysanthum species complex, R. pseudochrysanthum ssp. morii var. taitunense shows significant differences (p < 0.05) in both seed length and width when compared to R. morii, R. pseudochrysanthum, and R. hyperythrum (Fig. 8). However, no significant differences in seed width were observed among R. morii, R. pseudochrysanthum, and R. hyperythrum.

Figure 7.

Figure 7.

Open in a new tab

Seed morphology of the Rhododendron pseudochrysanthum species complex. A. R. pseudochrysanthum Hayata ssp. morii (Hayata) Yamazaki var. taitunense Yamazaki; B. R. morii Hayata; C. R. pseudochrysanthum Hayata; D. R. hyperythrum Hayata.

Table 4.

Seed morphological differences of the R. pseudochrysanthum species complex.

taxon R. pseudochrysanthum ssp. morii var. taitunense R. morii R. pseudochrysanthum R. hyperythrum
traits
Seed length 1.303 ± 0.070a 1.371 ± 0.146b 1.062 ± 0.095c 0.996 ± 0.102d
Seed width 0.662 ± 0.058a 0.574 ± 0.092b 0.552 ± 0.121b 0.552 ± 0.094b
Seed shape narrowly elliptic narrowly elliptic elongate elongate
Open in a new tab

Mean ± SE (n = 30); Unit = millimeter (mm). Different letters indicate statistically significant differences (p < 0.05).

Figure 8.

Figure 8.

Open in a new tab

Boxplot comparing the seed characteristics of the R. pseudochrysanthum species complex. Different letters in each boxplot indicate statistically significant differences (p < 0.05).

As elevation increases, wind becomes stronger, facilitating more effective seed dispersal. Therefore, Rhododendron species at higher elevations often possess narrower seeds and seed wings. Since stronger winds at high altitudes can effectively disperse seeds, there is less selective pressure to develop elaborate dispersal structures (Howe and Smallwood 1982). Nonetheless, R. pseudochrysanthum and R. hyperythrum, both distributed at high elevations, still exhibit prominent wing-like seed appendages (Fig. 7).

Cluster analysis

To determine the optimal number of clusters, we employed the Elbow Method in conjunction with average silhouette width as an evaluation metric, using hierarchical clustering (HC) analysis. The results indicated that when the number of clusters was set to two, the average silhouette width reached its maximum value (approximately 0.8; Fig. 9), suggesting the highest intra-cluster similarity and inter-cluster dissimilarity at this point. Therefore, based on this analysis, dividing the samples into two clusters was considered the optimal solution, further supporting a clear distinction between R. pseudochrysanthum ssp. morii var. taitunense and the other taxa in terms of pollen and seed characteristics (Fig. 10). This result indicates that R. pseudochrysanthum ssp. morii var. taitunense is distinct from the other three taxa within the R. pseudochrysanthum species complex in terms of reproductive traits.

Figure 9.

Figure 9.

Open in a new tab

Elbow Method for Determining Optimal Number of Clusters in Hierarchical Clustering. Number of clusters k = 2.

Figure 10.

Figure 10.

Open in a new tab

Cluster analysis of reproductive traits. The yellow lines are R. pseudochrysanthum ssp. morii var. taitunense, and the orange lines are R. morii, R. pseudochrysanthum, and R. hyperythrum.

Distribution and habitat

Within the Rhododendron pseudochrysanthum species complex, R. pseudochrysanthum ssp. morii var. taitunense shows a distinct geographic distribution compared with other taxa, occurring in the Datun Volcanic Group, Keelung Volcanic Group, Shiding District, and Wulai District at elevations of 440–1,200 m (Fig. 11). Currently, it is known from Mt. Qixing and Mt. Wuzhi in Taipei City, as well as Mt. Zhuzi, Mt. Caigongkeng, Mt. Xin, Mt. Canguangliao, Mt. Banping, Mt. Bijia, and Mt. Hongludi in New Taipei City. This taxon primarily grows along ridges, cliff edges, or in open forest habitats, whereas the other three taxa are mainly distributed in mid- to high-elevation mountainous regions of Taiwan.

Figure 11.

Figure 11.

Open in a new tab

Distribution map of Rhododendron pseudochrysanthum species complex in Taiwan.

Conclusions

In summary, based on the literature review, the geographical distribution, and morphological analyses, we found that R. pseudochrysanthum ssp. morii var. taitunense is clearly distinguishable from other taxa within the R. pseudochrysanthum species complex by its strongly revolute mature leaves, larger flower buds with a long conical shape (over 3 cm), and larger pollen grains. Furthermore, the restricted and localized distribution of this taxon solely to the low-elevation mountainous areas of Northern Taiwan further supports its distinctness.

In addition, we found that R. morii, R. pseudochrysanthum, and R. hyperythrum are difficult to delineate taxonomically based on certain morphological characteristics, such as floral morphology, pollen, and seed. This complexity is compounded by the occurrence of intermediate individuals between R. pseudochrysanthum and R. hyperythrum in high-elevation areas, making species delineation challenging. Therefore, more in-depth studies on these remaining taxa are planned for the future, including expanding sampling efforts and incorporating molecular evidence.

Acknowledgements

We are deeply grateful to Prof. Diego Tavares Vasques (The University of Tokyo) for his invaluable assistance in providing specimens of the subgenus Hymenanthes from Taiwan. We appreciate Prof. Chang-Fu Hsieh (謝長富) for his suggestions on classification of Rhododendron. We thank Mr Qi-Xiang Lian (連啟翔), Ms Yun-Ju Fang (方韻茹), Ms Chia-Hsiu Liu (劉家秀), Ms Shu-Ting Chiu (邱舒莛), and all the members of the Laboratory of Forest Plant Taxonomy and Ecology for field assistance.

Citation

Zheng Y-R, Tseng Y-H, Tzeng H-Y (2026) Morphological variation in the Rhododendron pseudochrysanthum Hayata (Ericales, Ericaceae) species complex from Taiwan. PhytoKeys 271: 187–204. https://doi.org/10.3897/phytokeys.271.175682

Funding Statement

This study was supported by the grant of National Science and Technology Council no. 112-2313-B-005-016-MY3 to Hsy-Yu Tzeng (曾喜育).

Additional information

Conflict of interest

The authors have declared that no competing interests exist.

Ethical statement

No ethical statement was reported.

Use of AI

AI was used for academic English editing.

Funding

This study was supported by the grant of National Science and Technology Council no. 112-2313-B-005-016-MY3 to Hsy-Yu Tzeng (曾喜育).

Author contributions

Conceptualization: YHT, HYT, YRZ. Data curation: YRZ. Formal analysis: YRZ. Funding acquisition: HYT. Investigation: YRZ. Methodology: HYT, YRZ. Project administration: HYT, YRZ. Resources: YRZ, HYT. Software: HYT. Supervision: YHT, HYT. Validation: YHT, HYT. Visualization: YRZ. Writing – original draft: YRZ. Writing – review and editing: YHT, HYT, YRZ.

Author ORCIDs

Ya-Rong Zheng https://orcid.org/0009-0007-7991-5260

Yen-Hsueh Tseng https://orcid.org/0000-0003-3644-1863

Hsy-Yu Tzeng https://orcid.org/0000-0003-0952-6255

Data availability

All of the data that support the findings of this study are available in the main text.

References

  1. Barthlott W (1990) Scanning electron microscopy of the epidermal surface in plants. In: Claugher D (Ed.) Scanning Electron Microscopy in Taxonomy and Functional Morphology, Clarendon Press, Oxford, 69–94.
  2. Barthlott W (1981) Epidermal and seed surface characters of plants: Systematic applicability and some evolutionary aspects. Nordic Journal of Botany 1: 345–355. 10.1111/j.1756-1051.1981.tb00704.x [DOI]
  3. Cao JJ, Li YS, Chang CT, Chung JD, Hwang SY (2022) Adaptive divergence without distinct species relationships indicate early stage ecological speciation in species of the Rhododendron pseudochrysanthum complex endemic to Taiwan. Plants 11: 1226. 10.3390/plants11091226 [DOI] [PMC free article] [PubMed]
  4. Chamberlain DF (1982) A revision of Rhododendron II subgenus Hymenanthes. Notes from the Royal Botanic Garden Edinburgh 39: 270–486. 10.24823/nrbge.1982.3252 [DOI]
  5. Chen CY, Liang BK, Chung JD, Chang CT, Hsieh YC, Lin TC, Hwang SY (2014) Demography of the upward-shifting temperate woody species of the Rhododendron pseudochrysanthum complex and ecologically relevant adaptive divergence in its trailing edge populations. Tree Genetics & Genomes 10: 111–126. 10.1007/s11295-013-0669-x [DOI]
  6. Chung JD, Lin TP, Chen YL, Cheng YP, Hwang SY (2007) Phylogeographic study reveals the origin and evolutionary history of a Rhododendron species complex in Taiwan. Molecular Phylogenetics and Evolution 42: 14–24. 10.1016/j.ympev.2006.06.027 [DOI] [PubMed]
  7. Ding BY, Wu HX, Zhang HM, Fang YY (1995) Seed morphology of Rhododendron (Ericaceae) from Zhejian and its taxonomic significance. Acta Botanica Boreali-Occidentalia Sinica 15(6): 36–42.
  8. Erdtman G (1952) Pollen Morphology and Plant Taxonomy-Angiosperms. Almqvist and Wiksell, Stockholm, 539 pp. 10.1080/11035895209453507 [DOI]
  9. Fang MY, Fang RZ, He MY, Hu LZ, Yang HB, Chamberlain DF (2005) Rhododendron. Flora of China. http://www.efloras.org/flora_page.aspx?flora_id=2
  10. Fuhsiung W, Nanfen C, Yulong Z, Huiqiu Y (1995) Pollen Flora of China. 2nd ed. Academia Sinica, Inst. of Botany, Beijing, 461 pp.
  11. Gao LM, Zhang CQ, Li DZ, Wei ZX (2002) Pollen morphology of Rhododendron subgenus Azaleastrum. Plant Science Journal 20(3): 177–181.
  12. Geng YY (2014) The Genus Rhododendron of China. China Science Publishing & Media Ltd, 612 pp.
  13. Halbritter H (1998) Preparing living pollen material for scanning electron microscopy using 2,2-dimethoxypropane (DMP) and critical point drying. Biotechnic & Histochemistry 73: 137–143. 10.3109/10520299809140519 [DOI] [PubMed]
  14. Hayata B (1913) Rhododendron. Icones Plantarum Formosanarum 3: 133, 140, 141.
  15. Howe HF, Smallwood J (1982) Ecology of seed dispersal. Annual Review of Ecology and Systematics 13: 201–228. 10.1146/annurev.es.13.110182.001221 [DOI]
  16. Hsieh CY (2007) The Origin and Evolution History of Rhododendron Species Complex in Taiwan Based on the Intron Sequences of LEAFY Gene. Master Thesis, Graduate Institute of Biotechnology, Chinese Culture University, 78 pp.
  17. Hsu CC (1973) Biosystematics investigation on the Rhododendron of Taiwan. Proceedings of the National Science Council 6: 13–50.
  18. Huang TC (1972) Pollen Flora of Taiwan. Department of Botany, National Taiwan University, Taipei, Taiwan, 297 pp.
  19. Kanehira R (1917) Formosan Trees. Bureau of Productive Industry, Government of Formosa Press, Taihoku (Taipei), 651 pp.
  20. Kanehira R (1936) Formosan Trees: Indigenous to the Island (revised). Department of Forestry, Government Research Institute, Formosa, 574 pp.
  21. Kron K (1997) Phylogenetic relationships of Rhododendroideae (Ericaceae). American Journal of Botany 84(7): 973–980. 10.2307/2446288 [DOI] [PubMed]
  22. Kron KA, Judd WS, Stevens PF, Crayn DM, Anderberg AA, Gadek PA, Quinn CJ, Luteyn JL (2002) Phylogenetic classification of Ericaceae: Molecular and morphological evidence. Botanical Review 68: 335–423. 10.1663/0006-8101(2002)068[0335:PCOEMA]2.0.CO;2 [DOI]
  23. Kurashige Y, Mine M, Kobayashi N, Handa T, Takayanagi K, Yukawa T (1998) Investigation of sectional relationships in the genus Rhododendron (Ericaceae) based on matK sequences. Journal of Japanese Botany 73(3): 143–154.
  24. Kurashige Y, Etoh JI, Handa T, Takayanagi K, Yukawa T (2001) Sectional relationships in the genus Rhododendron (Ericaceae): Evidence from matK and trnK intron sequences. Plant Systematics and Evolution 228: 1–14. 10.1007/s006060170033 [DOI]
  25. Léveillé H, Vaniot E (1911) Rhododendron rubropunctatum. Repertorium Specierum Novarum Regni Vegetabilis 9: 448.
  26. Li HL (1963) Woody Flora of Taiwan. The Livingston Publishing Company, Narberth, Pennsylvania, USA, 974 pp.
  27. Li HL (1978) Ericaceae. In: Li HL, Liu TS, Huang TS, Koyama T, DeVol CE (Eds) Flora of Taiwan 1st ed., Vol. 4. Epoch Publishing Company, Taipei, Taiwan, 21–38.
  28. Li HL, Lu SY, Yang YP, Tseng YH (1998) Ericaceae. In: Hsieh CF, Huang TC, Li YZ, Lo HC, Ohashi H, Shen CF, and Wang JC (Eds) Flora of Taiwan 2nd ed., Vol. 4. Editorial Committee of the Flora of Taiwan, Taipei, Taiwan, 17–39.
  29. Liang BK (2011) Postglacial Population Bottleneck and Species Divergence in Rhododendron pseudochrysanthum Species Complex. Master Thesis. Department of Life Science, National Taiwan Normal University, 85 pp.
  30. Lin CT (2018) QGIS Template for Displaying Species Distribution by Horizontal and Vertical View in Taiwan. https://github.com/mutolisp/distrmap_tw.qgis
  31. Lu SY, Yang TP (1989) A revision of Rhododendron (Ericaceae) of Taiwan. Bulletin of Taiwan Forestry Research Institute 4(4): 155–166.
  32. Mao ZJ, Yang YF, Hou LJ (2000) A study on the pollen morphology of Rhododendron in northeast of China. Bulletin of Botanical Research 20(1): 58–62.
  33. Masamune G (1936) Short Flora of Formosa. The Editorial Department of “Kudoa”, Taihoku (Taipei), Taiwan, 410 pp.
  34. Masamune G (1954) A List of Vascular Plants of Taiwan. Hokuriku no Shaka, Kanazawa, Japan, 192 pp.
  35. Milne RI (2004) Phylogeny and biogeography of Rhododendron subsection Pontica, a group with a Tertiary relict distribution. Molecular Phylogenetics and Evolution 33(2): 389–401. 10.1016/j.ympev.2004.06.009 [DOI] [PubMed]
  36. Miyoshi N, Fujiki T, Kimura H (2011) Pollen Flora of Japan. Hokkaido University Press, Sapporo. [in Japanese]
  37. R Core Team (2024) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org
  38. Sarwar AKMG (2007) Pollen Morphology and Its Systematic Significance in the Ericaceae. PhD Thesis. Division of Bioresources and Product Science Graduate School of Agriculture, Hokkaido University Sapporo, Japan, 302 pp.
  39. Sarwar AKMG, Takahashi H (2013) Pollen morphology of Rhododendron L. and related genera and its taxonomic significance. Bangladesh Journal of Plant Taxonomy 20(2): 185–199. 10.3329/bjpt.v20i2.17393 [DOI]
  40. Sasaki S (1928) List of Plants of Formosa. The Natural History Society of Formosa, Taihoku (Taipei), 564 pp.
  41. Su HJ (1985) Studies on the climate and vegetation types of the natural forests in Taiwan. (III). A scheme of geographical climatic regions. Quarterly Journal of Chinese Forestry 18(3): 33–44.
  42. Thiers B (2026) Index Herbariorum: A Global Directory of Public Herbaria and Associated Staff. New York Botanical Garden’s Virtual Herbarium. http://sweetgum.nybg.org/science/ih/ [Accessed on 12 January 2026]
  43. Vasanthy G, Pocock SAJ (1987) On the pollen tetrads of the south Indian Ericaceae, Gaultheria, Rhododendron and Vaccinium with special reference to R. nilagiricum Zenk. Bulletin du Jardin Botanique National de Belgique 57: 213–245. 10.2307/3668319 [DOI]
  44. Wang YG, Li GZ, Qi XX, Ou ZL (2006) Pollen morphology of Rhododendron and its taxonomic implication. Guihaia 26(2): 113–119.
  45. Wang YG, Li GZ, Zhang WJ, You J, Chen JK (2007) A systematic study of the genus Rhododendron (Ericaceae) using micromorphological characters of fruit surface and seed coat. Acta Phytotaxonomica Sinica 45(1): 21–38.
  46. Wickham H (2016) ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. https://ggplot2.tidyverse.org
  47. Wilson EH (1925) The Rhododendron of Eastern China, the Bonin and Liukiu Islands and of Formosa. Journal of the Arnold Arboretum 6: 156–185. 10.5962/p.317996 [DOI]
  48. Yamazaki T (1981) Some new taxa of Rhododendron from Japan and Taiwan. Journal of Japanese Botany 56(11): 357–366.
  49. Ying SS (1976) The Ericales of Taiwan. Quarterly Journal of Chinese Forestry 9(4): 107–137.
  50. Zhang YJ, Jin XF, Ding BY, Zhu JP (2009) Pollen morphology of Rhododendron subgen. Tsutsusi and its systematic implication. Journal of Systematics and Evolution 47: 123–138. 10.1111/j.1759-6831.2009.00011.x [DOI]
  51. Zhou LY, Wang YQ, Zhang L (2008) Pollen morphology and its taxonomy of twenty-six species in Rhododendron. Scientia Silvae Sinicae 44(2): 55–63.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

All of the data that support the findings of this study are available in the main text.

Articles from PhytoKeys are provided here courtesy of Pensoft Publishers

RESOURCES