Distribution and genetic diversity of Gueldenstaedtia verna in Korea

Occurrence and phenology survey of G. verna

To assess the global distribution of G. verna, georeferenced occurrence data were retrieved from the Global Biodiversity Information Facility (GBIF.org; April 12, 2024; GBIF Occurrence Download, https://doi.org/10.15468/dl.m78bef). For the current distribution and flowering–fruiting period of G. verna within Korea, additional records were reviewed from the Korea National Arboretum (KH) herbarium database and the citizen science platform Naturing (https://www.naturing.net). Observational and herbarium records from these sources were compiled to supplement field-collected data.

Distribution maps were created using QGIS 3.32.3 (Quantum Geographic Information System, 2023). Three maps were produced to illustrate: (1) the global distribution of the species based on georeferenced occurrence data from published sources and public databases (e.g., GBIF.org); (2) the known populations in Korea, including newly confirmed sites; and (3) spatial mapping of G. verna within the newly discovered locality at Hannam University, Daejeon. All geographic data were visualized using the EPSG:4326 coordinate reference system (WGS 84), and final maps were exported at high resolution for figure presentation.

Field survey, plant material sampling, and DNA extraction

Field surveys were conducted at representative localities across South Korea, encompassing sites in the Daegu region as well as other geographically distant areas. Specimens of G. verna were collected specifically from five selected localities within Daegu, Gyeongsangnam-do, and Daejeon during their flowering and fruiting period (Fig. 1B, Table 1). These specimens were collected for phylogenetic analysis and assessment of genetic diversity within South Korea. The voucher specimens have been deposited in the Hannam University Herbarium (HNHM). Fresh leaf samples were collected and preserved in silica gel for DNA extraction. Total genomic DNA was extracted using the NucleoSpin Plant II kit (Macherey-Nagel, Düren, Germany), following the sodium dodecyl sulfate method.

Fig. 1

Geographic distribution of Gueldenstaedtia verna. A. Worldwide distribution of G. verna. B. Previously reported localities and confirmed sites of G. verna in South Korea. C. Detailed mapping of G. verna within the newly discovered population at Hannam University, Daejeon.

Table 1

List of field survey sites of Gueldenstaedtia verna in South Korea.

Polymerase chain reaction and Sanger sequencing

Nuclear (nrITS) and plastid (psbA-trnH and trnL-F) DNA markers were selected because they have been previously used in a molecular phylogenetic study of Gueldenstaedtia, allowing for direct comparison with earlier datasets (Xie et al., 2016). In contrast, commonly used barcoding regions such as matK and rbcL were not employed due to their limited discriminatory power at the interspecific level within this genus.

For the ITS region, primers ITS4 (TCCTCCGCTTATTGATATGC) and ITS5 (GGAAGTAAAAGTCGTAACAAGG) were used, based on the standard primer set provided by Macrogen Inc. (Seoul, Korea). The psbA-trnH region was amplified using psbA-trnHF (CGCGCATGGTGGATTCAACATC) and psbA-trnHR (GTTATGCATGAACGTAATGCTC), following the CBOL Plant Working Group (2009). For the trnL-F region, trnL_F_GF (CGCGCATGGTGGATTCAACTC) and trnL_F_GR (GTTATGCATGAACGTAATGCTC) were used, based on Zhu et al. (2013) and Taberlet et al. (1991). PCR amplifications were conducted in 20 μL volumes using 2× Taq PCR MasterMix (BIOFACT, Daejeon, Korea). The cycling conditions consisted of an initial denaturation at 95°C for 5 min; 35 cycles of denaturation at 95°C for 1 min, annealing at 55°C for 1 min, and extension at 72°C for 1 min; followed by a final extension at 72°C for 10 min. All PCR products were sequenced using the Sanger method at Macrogen Inc., and forward and reverse reads were assembled and edited using Geneious Prime 2025.0.1 (https://www.geneious.com/).

Phylogenetic analysis

To determine the phylogenetic position of Korean G. verna within a broader context, we included additional sequences from related taxa representing the genera Gueldenstaedtia, Tibetia, Chesneya, and Caragana (On-line Supplemental Data Table S1). This analysis incorporated 53 published sequences per marker (nrITS, psbA-trnH, and trnL-F) retrieved from GenBank (On-line Supplemental Data Table S1), along with a representative Korean G. verna sequence generated in this study. The representative Korean nrITS sequence was deposited in GenBank under accession number PQ897958.

Sequence alignments were conducted using MAFFT v7.520 (Katoh and Standley, 2013) with default parameters. A maximum likelihood phylogenetic analysis was performed using IQ-TREE v2.2.2.6 (Minh et al., 2020), employing 100,000 bootstrap replicates to assess tree robustness. The optimal nucleotide substitution model was selected based on the Akaike Information Criterion, and branches with negligible length were collapsed. The resulting phylogenetic tree was visualized using the Interactive Tree Of Life (iTOL) online tool (Letunic and Bork, 2021).

Next-generation sequencing, assembly, and annotation

Genomic DNA extracted from the Daejeon individual BG2404001 was sent to Macrogen Inc., for next-generation sequencing. Genomic library was prepared using the TruSeq Nano DNA kit (Illumina, San Diego, CA, USA), and sequencing was performed to generate 151 bp paired-end reads on an Illumina platform. The sequencing reads were processed to assemble the plastome of G. verna using Geneious Prime, following the assembly protocol outlined by Choi et al. (2019). The assembled genomes were annotated using GeSeq, and they were manually refined using G. verna (NC_085784) as the reference (Tillich et al., 2017). Some tRNA regions were not annotated in the reference genome and were manually reannotated for comparative analysis in Geneious Prime. Subsequently, sequence variation between the newly assembled plastome and the reference genome (G. verna, NC_085784) was examined to identify divergent regions, which were manually inspected in Geneious Prime through comparative alignment. The annotated plastome of G. verna was visualized as a genome map using Organellar Genome DRAW (OGDRAW) v1.3.1 (Greiner et al., 2019).

Updated distribution range and new localities of G. verna in Korea

A global survey based on occurrence data from GBIF (1,268 records) and Xie et al. (2016) (15 records) (On-line Supplemental Data Table S1) highlighted the geographic isolation of South Korean populations of G. verna from the species’ main range (Fig. 1A). Within South Korea, distribution was also somewhat discontinuous, with localities around Daegu and more distant regions (Fig. 1B), as determined through field surveys (8 sites) and supporting occurrence data from Korea National Arboretum (16 records) and Naturing (9 records) (Table 1, On-line Supplemental Data Table S2).

In Daegu, G. verna was commonly found around the Bullodong ancient tombs. Additionally, we also found individuals within typical campus lawns at Kyungpook National University and Keimyung University. The newly documented locality at Hannam University in Daejeon, located over 165 km north of Daegu, similarly featured individuals in general university lawn habitats (Fig. 1B, C).

Field surveys conducted at previously recorded localities in Busan and Cheongsong (Gyeongsangbuk-do) failed to detect any individuals. The Busan site corresponded to a landscaped lawn, and the Cheongsong site was located near a roadside graveyard.

Habitat characteristics and distribution of G. verna at Hannam University, Daejeon

Individuals of G. verna were identified at Hannam University in Daejeon, South Korea (specific locality: 1646 Yuseong-daero, Yuseong-gu, Daejeon; elevation approximately 60 m). Observations were made at two closely situated sites on the campus, both characterized by turf-covered, landscaped flower beds. Approximately 60 mature individuals were distributed across a grassy area of about 10 m² (Figs. 1C, 2).

Fig. 2

Morphological characteristics of Gueldenstaedtia verna at Hannam University. A. Whole plant. B. Stem. C. Leaf (C-1: adaxial; C-2: abaxial). D. Leaflet (D-1: adaxial; D-2: abaxial). E. Inflorescence before flowering. F. Inflorescence after flowering. G. Flower (G-1: front view; G-2: side view). H. Fruit. I. Seed.

The habitat conditions at these sites, including open grassy slopes with well-drained soils, did not exhibit noticeable differences from previously known localities near Daegu. Moreover, no distinct morphological differences were observed between the Daejeon and Daegu individuals.

Phenology of G. verna across South Korean localities

Data from herbarium records (Korea National Arboretum), citizen science platform (Naturing), and our field surveys indicate that the flowering period of G. verna typically begins in early spring (Table 1, On-line Supplemental Data Table S2). In the surveyed localities, initial flowering was observed in mid-March around Daegu, continuing until early June in northern regions such as Gyeongsangbuk-do. Specifically, flowering occurred from March to April in Gyeongsangnamdo, March to May in Daegu, late April to early June in Gyeongsangbuk-do, during April in Busan (based on a single record), and from April to May in Daejeon.

Fruit development generally commenced approximately one month after flowering. Due to the extended flowering period and the presence of multiple inflorescences per plant, simultaneous flowering and fruiting were commonly observed. Additionally, fruiting was recorded as late as November in Daegu, suggesting potential continuation of flowering into the summer months in certain localities.

Genetic diversity of G. verna in Korea

To assess the genetic diversity of G. verna within Korea, we collected a total of 18 individuals: six from Daejeon, nine from Daegu, and three from Changnyeong-gun, Gyeongsangnam-do (Table 1). All collected individuals were successfully sequenced for the nrITS region as well as two plastid intergenic regions (psbA-trnH and trnL-F). The aligned nrITS sequences measured 614 bp, while the combined plastid sequences (psbA-trnH and trnL-F) totaled 1,372 bp. No sequence variation was observed among the sampled individuals for either the nrITS or the plastid markers.

The phylogenetic position of Korean G. verna

To clarify the phylogenetic placement of Korean G. verna, we conducted a combined analysis using nuclear (nrITS) and plastid (psbA-trnH and trnL-F) sequence data. Initially, our analysis focused on variations within the genus Gueldenstaedtia, comprising three recognized species—G. verna, G. henryi, and G. guangxiensis—with particular emphasis on G. verna accessions. A Korean sample from Yuseong-gu, Daejeon (BG2404001) represented the Korean locality in our study.

Several nucleotide substitutions and indels were detected among G. verna individuals, especially within the psbA-trnH (6 sites) and trnL-F (4 sites) plastid regions (Table 2). The Korean sample exhibited identical sequences for all three markers when compared to six Chinese G. verna samples: X238 and X256 (Shandong), X018 and X142 (Shaanxi), X182 (Hebei), and X150 (Gansu). Conversely, the most divergent accession, G. verna X172 (Hebei), differed from the Korean sample at six nucleotide positions across the three markers.

Table 2

Sites of variability among aligned sequences for ITS, psbA-trnH, and trnL-F from Gueldenstaedtia verna and related species.

The resulting phylogenetic tree, constructed using the combined dataset of the three markers (Fig. 3), grouped the Korean G. verna sample (BG2404001) within a moderately supported clade (bootstrap support of 75) containing Chinese G. verna accessions. However, this clade also included individuals of G. guangxiensis and G. henryi, and the internal relationships among species and individuals remained unresolved due to limited sequence variability.

Fig. 3

Phylogenetic tree of Gueldenstaedtia verna and related species inferred by a maximum-likelihood using concatenated ITS, psbAtrnH, and trnL-F sequences. Bootstrap values (BS) are indicated on each branch, ranging from 62 to 100. The triangle indicates the phylogenetic position of G. verna.

Plastid genome level variation between Korean G. verna individuals

For comparative purposes, we assembled the complete plastid genome of G. verna from the Daejeon locality (GenBank accession number: PQ899157) (Fig. 4). A total of 83,954,952 reads were generated from the Illumina HiSeq platform. Of these, 14,413,147 reads were mapped to the plastome of G. verna, resulting in a mean coverage of 17,787X. The assembled plastome from Daejeon is 122,283 bp in length and comprises 110 unique genes, including 76 protein-coding genes, 30 tRNA genes, and four rRNA genes.

Fig. 4

The complete plastid genome map of Gueldenstaedtia verna. Genes belonging to different functional groups are color-coded, as indicated in the legend at the bottom left. Genes shown on the outside of the outer circle are transcribed clockwise, while those on the inside are transcribed counterclockwise. The inner circle indicates guanine-cytosine content (darker shading) and adenine-thymine content (lighter shading).

We compared it with a previously published plastome from an individual collected in Daegu (NCBI GenBank accession number: NC_085784). The Daegu plastome is slightly longer at 122,569 bp, but both genomes show conserved overall structure, gene content, and gene order, with identical numbers of unique genes: 76 protein-coding genes, 30 tRNA genes, and four rRNA genes. Between these two plastomes, we identified 25 single nucleotide polymorphisms and 29 insertions/deletions (indels), predominantly located within intergenic regions such as psbA-psbK and trnT(UGU)-trnL(UAA) (Table 3).

Table 3

Comparison of regional sequence variants between the plastid genomes of Gueldenstaedtia verna from Daejeon (PQ899157) and Daegu (NC_085784).