Redefining the distributional boundaries of Indigofera kirilowii and Indigofera koreana in Korea

Distribution survey and taxon identification

From 2023 to 2024, field surveys were conducted at 32 locations across South Korea to document the geographic distribution of I. kirilowii and I. koreana (Table 1). During these surveys, unexpected occurrences of I. koreana were found in the northeastern region of South Korea, specifically in Sokcho-si and Yangyang-gun, Gangwon-do. Based on these findings, we conducted additional targeted surveys in nearby areas to further investigate the northern distribution of I. koreana. As part of this effort, focused field surveys and habitat assessments were carried out at 11 sites in Gangwondo (Table 2). These additional surveys aimed to clarify the northern distributional limit of I. koreana and to identify its potential range boundaries in the region. Selection of survey sites was based on proximity to the initial discovery sites and the presence of habitat conditions likely suitable for I. koreana. At each survey site, voucher specimens were collected in the field and deposited at the herbarium of Hannam University (HNHM). In areas where specimen collection was restricted (national parks and other protected areas), photographs were taken to document the distribution of the two Indigofera species. These photographs were uploaded to the citizen science platform, Naturing (http://www.naturing.net), as photographic specimens (Baskauf and Kirchoff, 2008). In addition to field surveys, approximately 1,700 specimens were observed in the herbaria of Hannam University (HNHM); Inha University (IUI); National Institute of Biological Resources, Korea (KB); Korea National Arboretum (KH); Honam National Institute of Biological Resources (HIBR) (On-line Supplemental Data Table S1).

Table 1.

List of field survey sites for Indigofera kirilowii and I. koreana in South Korea.

Table 2.

List of field survey sites for the northernmost distribution of Indigofera koreana in South Korea.

To accurately distinguish I. kirilowii and I. koreana, we followed the diagnostic characteristics proposed by Cho et al. (1997) and Kim et al. (2005). Specimens with trichomes on the abaxial surface of the leaflets and longer flowers (1.2–1.6 cm) were treated as I. kirilowii, while those lacking trichomes and having shorter flowers (0.8–1.2 cm) were identified as I. koreana (Fig. 1). These traits were checked on live plants in the field using a loupe, and on herbarium specimens with a digital microscope (Dino-Lite Edge AM7915, New Taipei City, Taiwan).

Fig. 1.

Diagnostic characteristics of two Korean Indigofera species in this study. A. I. kirilowii. B. I. koreana. 1, trichomes on the adaxial surface of the leaflet; 2, trichomes on the abaxial surface of the leaflet; 3, flower length.

Visualization of the distribution of Indigofera kirilowii and Indigofera koreana in Korea

Geographical data (GPS, Global Positioning System) were obtained from specimens to construct distribution maps of I. kirilowii and I. koreana. During the examination, the coordinates recorded on specimen labels were verified after accurate species identification. If a specimen lacked coordinates, we used coordinates of the nearest mountain or a specific peak near the collection site. Distribution maps of I. kirilowii and I. koreana in Korea were generated based on the obtained coordinates using QGIS 3.32.3 software (Quantum Geographic Information System, 2023).

Distribution of Indigofera kirilowii and Indigofera koreana in Korea

By examining approximately 1,700 specimens and conducting field surveys, we identified a total of 1,050 unique sites with confirmed occurrences of I. kirilowii or I. koreana. During the verification process, about 250 specimens that had been misidentified in the past were corrected to their proper species. As a result, I. kirilowii was confirmed to be present at 463 sites, and I. koreana at 587 sites across Korea. Using these verified occurrence data, we constructed a detailed distribution map for both species (Fig. 2).

Fig. 2.

Distribution map of Indigofera kirilowii and I. koreana in Korea. AS, Asan-si; DJ, Daejeon; SC, Sokcho-si; YY, Yangyang-gun.

Overall, Indigofera kirilowii is distributed in the central inland and northern regions of South Korea, ranging from 35° to 38°N and 124° to 129°E. It is primarily found in Gyeonggido, Gangwon-do, Chungcheongbuk-do, Chungcheongnam-do, and Gyeongsangbuk-do, with less frequent presences in Gyeongsangnam-do, Jeollabuk-do, and Jeollanam-do. In contrast, I. koreana is distributed in the southern and southwestern regions of South Korea, ranging from 33° to 38°N and 125° to 129°E. It is mainly found in Chungcheongnam-do, Gyeongsangbuk-do, Gyeongsangnam-do, Jeollabuk-do, and Jeollanam-do, with sporadic presences in Gyeonggi-do, Gangwon-do, and Chungcheongbuk-do. The overall distribution patterns confirm that I. kirilowii predominantly occupies the northeastern regions, while I. koreana is mainly found in the southwestern regions of South Korea.

Distribution boundaries and co-distribution of Indigofera kirilowii and Indigofera koreana

Where the ranges of I. kirilowii and I. koreana meet, there is a broad transitional zone rather than a sharp boundary. We found that the interface of their distributions extends across parts of central and southern regions of South Korea. Notably, this boundary zone includes Mt. Gayasan region (on the border of Gyeongsangbuk-do and Gyeongsangnam-do), and spans across Chungcheongbuk-do, Daejeon, and Chungcheongnam-do in the central region, as well as Gyeongsangbuk-do, Daegu and Gyeongsangnam-do further south (Fig. 2). Within these areas, the two species’ ranges gradually overlap and become delineated. In addition, I. koreana was identified at a few sites in Yeongwol and Jeongseon, Gangwon-do, which are areas north of its main southwestern distribution, indicating that the overlap zone extends into parts of southern Gangwon-do as well.

Within this distributional overlap zone, we documented instances of I. kirilowii and I. koreana co-occurring on the same mountain, indicating they grow in close proximity in the wild. Two notable co-occurrence sites were confirmed (Fig. 3). The first is Mt. Baebangsan in Asan-si and the second is Mt. Gyejoksan in Daejeon. At Mt. Baebangsan, the codistribution site was located at an elevation of 128 meters with coordinates 36°44′39.3″N, 127°03′13.9″E. At Mt. Gyejoksan, I. kirilowii and I. koreana inhabit the same mountain but exhibit an altitudinal distribution difference, with I. koreana occurring at a lower elevations near the mountain entrance at 155 meters (36°23′58.6″N, 127°26′07.7″E) and I. kirilowii at a higher elevation near the mountain summit at 212 meters (36°23′45.3″N, 127°26′11.5″E).

Fig. 3.

Co-distribution of Indigofera kirilowii and I. koreana. A. Mt. Baebangsan, Asan-si, Chungcheongnam-do. B. M t. G yejoksan, Daedeok-gu, Daejeon. C. Habitat of Mt. Baebangsan.

The northernmost distribution of Indigofera koreana

To pinpoint the northernmost limit of I. koreana’s distribution, we surveyed 11 candidate sites in northeastern Gangwon-do (Table 2). Among all these surveyed localities, I. koreana was confirmed at only two sites—Seorak-dong in Sokcho-si and Hyeonnam-myeon in Yangyang-gun. In all the other surveyed areas, no individuals of I. koreana were detected. The Sokcho and Yangyang occurrences lie far to the northeast of I. koreana’s previously known range, extending the species’ distribution into the Gangwon-do region (see Figs. 2, 4).

Fig. 4.

Putative northernmost distribution and habitat of Indigofera koreana in Gangwon-do. A. Distribution map of Sokcho-si. B. Habitat of Sokcho-si. C. Distribution map of Yangyang-gun. D. Habitat of Yangyang-gun.

In Sokcho-si (Gangwon-do), I. koreana was observed within Seorak-dong district of Seoraksan National Park. The plants were primarily found in flowerbeds near the Seorak-dong Visitor Center, the stone bridge Biryonggyo, around the Mt. Seoraksan Cable Car ticket office at a low elevation (219–227 meters) (Fig. 4A, B). The confirmed I. koreana ramets exceeded 300 at this locality. These individuals were growing mostly alongside ornamental or planted vegetation, including Pinus densiflora Siebold & Zucc., Juniperus chinensis L., Acer palmatum Thunb., Rhododendron indicum (L.) Sweet, and Albizia julibrissin Durazz.

In Yangyang-gun (Gangwon-do), I. koreana was found in a more natural setting. The population is located around Pomaeho (Lagoon) in Hyeonnam-myeon (Fig. 4C, D). The surveyed habitat here ranges from 19 to 86 meters in elevation, extending along mountain paths and around graveyards near the lagoon. Indigofera koreana was the dominant understory shrub in this area, and we observed over 3,000 ramets in total. In the habitat, the following species were observed alongside I. koreana: Pinus densiflora, Quercus acutissima, Toxicodendron trichocarpum (Miq.) Kuntze, Zanthoxylum piperitum (L.) DC., Lespedeza bicolor Turcz., Smilax china L., and Melampyrum setaceum (Maxim. ex Palib.) Nakai.

The distribution patterns of Indigofera kirilowii and Indigofera koreana in Korea

In this study, we examined approximately 1,700 specimens of I. kirilowii and I. koreana from Korea, providing new insights into the geographical range of these two species (Fig. 2). The results show that I. kirilowii is predominantly distributed in northeastern South Korea and is rare in the southwest, while I. koreana primarily occurs in the southwest but is also found in limited areas of Gyeonggi-do and Gangwon-do. Although only a small number of specimens from some areas of North Korea were examined, all were identified as I. kirilowii. This distribution pattern largely aligns with the findings of Cho et al. (1997) and Kim et al. (2005). However, our extensive survey discovered additional distribution sites that were not previously reported. In particular, we clarified that the boundary between the distributions of two species extends across the southern and central in South Korea, including the Mt. Gayasan region, rather than being confined to a small area. We also documented a few mountains where both species coexist (Figs. 2, 3), demonstrating that their ranges are not completely allopatric. Despite the differences in their regional concentrations, there appear to be no absolute geographic barriers that entirely prevent I. kirilowii and I. koreana from growing in the same general areas. This lack of strict barriers enables occasional overlap and contact at the margins of their ranges, as observed in this study. It is also likely that additional instances of occurrence between these two species exist beyond those we have confirmed in this study.

We also identified instances of abnormal distribution that deviated from the general patterns (Figs. 2–4). Such exceptions might be explained by historical or ecological factors. One possibility is recent long-distance dispersal events, where seeds or other propagules were transported (perhaps by animals or human activity) beyond the usual range. Another possibility is that they are relict populations resulting from historical expansions and contractions of the species’ ranges during past climate oscillations (Hewitt, 2000, 2004; Hampe and Petit, 2005). Climate fluctuations in the Quaternary could have allowed I. kirilowii or I. koreana to temporarily extend into new areas and later retreat, leaving small isolated populations. Additionally, given that both species are frequently found in low-elevation disturbed sites and near human settlements or graveyards, we cannot rule out anthropogenic influence on their distributions. Human-mediated activities, such as the intentional or accidental moving of soil and plants, may have introduced these species to locales outside their core ranges. Thus, some outlier populations might be partially influenced by humans, rather than purely by natural range dynamics.

The northernmost distribution of Indigofera koreana

Our thorough investigation into the distribution of the Korean endemic I. koreana has enabled us to propose a putative northernmost range limit for this species. As noted, I. koreana is mostly confined to the southwestern regions of South Korea, aligning with the South and South Coast floristic subregions (Lee and Yim, 2002). The newly discovered populations in Sokcho-si and Yangyang-gun, Gangwon-do, are exceptional in that they occur far to the northeast of the species’ main range. These Gangwon-do records represent the extreme margin of I. koreana’s distribution.

We propose the populations in Sokcho-si and Yangyang-gun, Gangwon-do, as candidates for the northernmost limit of I. koreana (Fig. 4). While some low-altitude mountain regions along the eastern coast were also surveyed, the species was observed only at these locations (Table 2). The northernmost occurrence was recorded in Seoraksan National Park; however, the habitat conditions appear somewhat artificial. The I. koreana individuals there were mostly in landscaped flowerbeds and along maintained paths, and the population size (ca. 300 ramets) is relatively modest. These observations raise the possibility that at least some individuals, particularly found in obviously planted areas, might have been introduced by human activity (intentionally or unintentionally).

In contrast, the population in Yangyang-gun is situated in a more natural environment and exhibits a large population size (over 3,000 ramets). This population, spread across a forested hillside near Pomaeho, shows no signs of deliberate planting or gardening. Its size and setting suggest that it is a persistent, naturally occurring population that has likely been present and reproducing in that area for some time. Therefore, the Yangyang population is a strong candidate for a true natural northern limit of I. koreana.

The occurrence of I. koreana in these far-northeastern localities is biogeographically intriguing. Some evergreen tree species reach their northernmost limits in South Korea’s higher latitudes, likely influenced by warm oceanic currents (Lee and Choi, 2010). However, the case of I. koreana is somewhat unusual. Future phylogeographic studies and additional distribution surveys in the eastern coastal region will be essential to determine whether these proposed northernmost limits represent natural populations.