Introduction

Fagraea Thunb., a genus of the gentian family (Gentianaceae), consists of shrubs, trees, hemi-epiphytes and lithophytes occurring in tropical and subtropical regions from the Indian subcontinent to mainland Indochina, Taiwan, the Malay Archipelago, through to Australia and into the Pacific Islands (Struwe et al., 2002). Its centre of species diversity lies in Malesia, a well-known biogeographical region located within Southeast Asia (van Steenis, 1950; Leenhouts, 1962a; Struwe et al., 2002) and also one of the important biodiversity hotspots in the world (Myers et al., 2000). Thunberg dedicated the genus to the Swedish naturalist Jonas Theodor Fagraeus (1729–1797), and it is typified by F. ceilanica Thunb., a shrub known for its long trumpet-like, showy and scented flowers (Thunberg, 1782; Philcox, 1996). Fagraea ceilanica was described by Thunberg based on a flowering shrub he discovered when botanising between Colombo and the fort city of ‘Gale’ [Galle], ‘Ceilon’ [Ceylon; now Sri Lanka] (Thunberg, 1782). Sri Lanka was Thunberg’s final stopover before he returned to Europe after his maiden voyage to Japan (Thunberg, 1791; Svedelius, 1944). Accompanying Thunberg’s protologue of F. ceilanica there was an illustration, Tab. IV, featuring a flowering branch and a cross-section of a fruit of the new genus and new species (Fig. 1A).

When Fagraea was revised for Flora Malesiana (Leenhouts, 1962a), F. ceilanica was interpreted in the broadest sense as a highly polymorphic species under the concept Rassenkreis, a German term for ring-species or species with several geographical subspecies (Goldschmidt, 1933; Meikle, 1957; Leenhouts, 1962b; Mayr, 1963). Based on morphological evidence gathered from herbarium specimens, multiple interbreeding neighbouring populations or local forms of F. ceilanica were interpreted to occur from the Indian subcontinent to mainland Southeast Asia, Taiwan, across the Malay Archipelago and into the Pacific Islands (Leenhouts, 1962a, 1962b). In addition, several interspecies hybridisations were also believed to have taken place between F. ceilanica and other Fagraea species, such as in New Guinea with F. berteroana A. Gray ex Benth., which likely resulted in F. annulata Hiern (Leenhouts, 1962b). Fagraea berteroana, as defined at present, is yet another polymorphic species complex that occurs in New Guinea, Australia, to the archipelagos of the Pacific Ocean (Leenhouts, 1962a, b; Conn, 1995; Motley, 2004). Under the broad species concept adopted for F. ceilanica, many distinct species across the Malesian region and its vicinities were synonymised (Leenhouts, 1962a, b). When F. ceilanica s.l. was re-evaluated for regional taxonomic studies of Fagraea conducted for Borneo (Wong & Sugau, 1996) and the Malay Peninsula (Wong & Sugumaran, 2016), its species concept was found to be unsustainable, and subsequently reinterpreted in the strict sense. The outcome of these studies resulted in the recognition of distinct taxa previously synonymised (see Leenhouts, 1962a, b) being reinstated and new species described.

Fagraea ceilanica sensu Leenhouts in New Guinea

The ring-species hypothesis proposed for a species complex of Fagraea ceilanica recognised 14 forms or ‘races’, four of which occur in New Guinea: ‘angiensis’, ‘gardeniaeflora’, ‘litoralis’ and ‘ternatana’ (Leenhouts, 1962a, b). The ‘litoralis’ form is the most widespread throughout the distribution range of F. ceilanica s.l. It is defined in having (i) small- to medium-sized elliptical leaves, (ii) a pair of small bracteoles inserted in the middle along the length of the pedicel, and (iii) a cup-shaped calyx not constricted in the mouth (Leenhouts, 1962a, b).

The ‘angiensis' and ‘ternatana' forms share the following characteristics: (i) calyx constricted in the mouth, and (ii) a pair of small bracteoles inserted slightly above the middle along the length of the pedicel (Leenhouts, 1962a, b). The former is distinguishable from the ‘ternatana’ form in having slightly obovate leaves (Leenhouts, 1962a, b). On the other hand, the ‘gardeniaeflora’ form is distinct from the other three forms in New Guinea by having a pair of large bracteoles inserted just below the calyx and forming a cupule or a cup tightly clasping around the base of the calyx (Leenhouts, 1962a, b). Later, in the taxonomic treatment of Fagraea for Papua New Guinea, Leenhouts’ forms recognised in New Guinea were grouped into two primary forms, the ‘F. gardeniaeflora’/‘F. archboldiana’ form and the ‘F. litoralis’/‘F. ternatana’ form (Conn, 1995).

Morphologically, these four New Guinean forms proposed in Leenhouts (1962a, b) are distinct from Fagraea ceilanica s.s., and we believe they are worthy of taxonomic recognition. A similar conclusion was also made by Conn (1995). Nevertheless, we do not attempt to solve this taxonomic puzzle here as it requires a thorough and systematic evaluation of all the 14 F. ceilanica forms or ‘races’ proposed by Leenhouts across its distribution range, including all the synonyms.

Our taxononomic perspective

The first comprehensive taxonomic account of New Guinean Fagraea s.l. was by Leenhouts (1962a), who prepared it for the Flora Malesiana and recorded 13 species. In the flora of mountainous regions of New Guinea, F. salticola Leenh. is the only species enumerated to occur above 3,000 m elevation (Conn, 1983). Later, when a review of the F. gracilipes A. Gray complex sensu Leenhouts was carried out for Australia and New Guinea, five distinct taxa previously synonymised were reinstated (Conn & Brown, 1993). A flora account for Papua New Guinea, a sovereign state occupying the island's eastern half, enumerated 16 species of Fagraea s.l. (Conn, 1995). A recent checklist of New Guinean vascular flora (Cámaraleret et al., 2020) listed 11 species of Fagraea s.s., with one species transferred to Cyrtophyllum Reinw. and the other to Picrophloeus Blume following an approach based on molecular phylogenetic studies conducted on West Malesian Fagraea s.l. (Sugumaran & Wong, 2012; Wong & Sugumaran, 2012). Utania G. Don was unfortunately overlooked in Cámaraleret et al. (2020) but was accounted for in Trees of Papua New Guinea (Conn & Damas, 2019) and Trees of New Guinea (Utteridge & Jennings, 2021).

Fig. 1.

Fagraea ceilanica Thunb. A. Reproduction of the illustration, tab. IV (Thunberg, 1782); B. Lectotype (Thunberg s.n. [UPS-THUNB4308]); C. Flowering plant from Rahathangala, Sri Lanka. [Photos: A: Biodiversity Heritage Library (contributed by the Natural History Museum Library, London); B: reproduced with kind permission from the Museum of Evolution, Uppsala University, Sweden; C: Himesh D. Jayasinghe]

In 2019, a flowering and fruiting Fagraea specimen was collected around Ayamaru Lakes in the central region of the Bird's Head Peninsula (comprising the current Indonesian provinces of Southwest Papua and West Papua). Based on the Fagraea s.l. identification key in the Flora Malesiana account (Leenhouts, 1962a), we managed to key out the Ayamaru taxon as F. ceilanica s.l., but with much dissatisfaction considering it is a species with a broad species concept as discussed in Conn (1995) and Wong & Sugau (1996). While reviewing images of herbarium specimens of Fagraea s.l. from New Guinea at the Naturalis BioPortal website [ https://bioportal.naturalis.nl], we came across two collections from Ayamaru, Vink BW15226 and Vink BW15361, that were a perfect match to our own Ayamaru collection. These Vink collections bear determination slips of Leenhouts, with Vink BW15226 determined as F. ceilanica ‘angiensis’ form and Vink BW15361 as F. ceilanica ‘ternatana’ form. As we disagree with these determinations for the two Vink collections, we decided to examine this western New Guinean taxon more carefully.

Firstly, we affirmed that the three Ayamaru collections, Wanma et al. MAN-SING45, Vink BW15226 and Vink BW15361 are identical and belong to the same taxon. Separately, a collection from the vicinity of Ayamaru (Polak 923) that represents the unnamed Ayamaru taxon was brought to our attention by Timothy Utteridge (pers. comm.). We are convinced that the Ayamaru taxon is distinct from all 14 Fagraea ceilanica forms or ‘races’ recognised by Leenhouts, as the former has distinctly crenate bracteoles and calyx lobes. In contrast, the 14 F. ceilanica forms or ‘races’ of Leenhouts have entire bracteoles and calyx lobes margin (Leenhouts, 1962a, b). We further believe that the Ayamaru taxon is morphologically most similar to taxa that share the diagnostic morphological character of having stamens inserted upon a fleshy ring around the mid-portion of the inner surface of the corolla tube. In contrast, the fleshy ring is absent in F. ceilanica s.s. (Philcox, 1996). As we are unable to match the Ayamaru taxon to any named New Guinean, Australian or Malesian Fagraea species (Leenhouts, 1962a, b; Conn, 1995; Conn et al., 1996; Wong & Sugau, 1996; Wong & Sugumaran, 2016; Cámara-Leret et al., 2020), we hereby describe it as a new species, F. christinae Y.W. Low & V.A. Albert, endemic to the Bird’s Head Peninsula.

Material and methods

A survey of Fagraea ceilanica s.l., with an emphasis on Sri Lankan and New Guinean materials, was conducted based on herbarium specimens, protologues and monographs (Leenhouts, 1962a, b; Cámara-Leret et al., 2020). All measurements were taken from dried herbarium specimens held at BO, K, L, SING and UPS. Photographs of flowering and fruiting Fagraea species were taken in the field, from which herbarium materials were then gathered.Type images of Australian and Malesian Fagraea s.s. available on the JSTOR Global Plants website [ http://plants.jstor.org], as well as images of herbarium specimens of L, U and WAG on the Naturalis BioPortal website were accessed and examined.

The extent of occurrence (EOO) and area of occupancy (AOO) were calculated using GeoCAT (Bachman et al., 2011; Bachman & Moat, 2012) to assess the risk of extinction according to the IUCN Red List Categories and Criteria (IUCN, 2012). Data used for the assessment are based on herbarium records kept at BO, BRI, K, L, MAN, SING and WAG.

Taxonomic treatment

Fig. 2.

Fagraea christinae Y.W. Low & V.A. Albert. A. Habit of a flowering and fruiting branch; B. Close-up of calyx lobes outer surface showing warty or verrucose texture; C. Half of a longitudinal section of an open flower; note the stamens are inserted upon a thickened wall (forming a thickened ring in complete flower). [A–C: Wanma et al. MAN-SING45] [Drawings: X.Y. Loh]

Fig. 3.

Fagraea christinae Y.W. Low & V.A. Albert. A. Habit; B. Close-up of flower buds enclosed by calyx lobes and a pair of bracteoles with distinctive crenate margins; C. Side view of an open flower showing curled corolla lobe margins; D. Top view of an open flower with stamens clearly seen inserted upon a fleshy ring; E. Close-up of pink, mature fruits; fruit on the right artificially cut opened to reveal numerous small and black kidney-shaped seeds. [A–E: Wanma et al. MAN-SING45] [Photos: Y.W. Low]

Fig. 4.

Distribution of Fagraea christinae Y.W. Low & V.A. Albert in the Bird's Head Peninsula, New Guinea. Solid coloured area of the map reflects the West Papua Province (before the 2022 split into two separate provinces, Southwest Papua and West Papua), Indonesia.

Implications for future work

As demonstrated in previous studies (Wong & Sugau, 1996; Wong & Sugumaran, 2016; Venugopal et al., 2018) and the present work, the current broad circumscription of Fagraea ceilanica is unsustainable for species delineation, and it requires reappraisal throughout its distribution range. The first essential step is to interpret and define F. ceilanica in a strict sense based on the original protologue and the type specimen from Sri Lanka (see Fig. 1). Subsequently, a review of F. ceilanica sensu Leenhouts (1962a, b) based on morphological characteristics is to be carried out to (i) re-establish taxa with distinct morphological characters that were synonymised under it, and (ii) examine the 14 forms or ‘races’ proposed and describe novel taxa hidden under the broad species concept. Any population-level phylogenomics studies involving F. ceilanica s.l. should include all the 14 forms or ‘races’ proposed in Leenhouts (1962a, b), across its distribution range with involvement from local botanists, ecologists and naturalists to encourage diverse and inclusive scientific collaboration.

Species diversity of Fagraea is the highest in the Malesia region (POWO, 2023), and considering Borneo and the Malay Peninsula have been revised within the last 30 years (see Wong & Sugau, 1996; Wong & Sugumaran, 2016). Immediate attention is required for a taxonomic reassessment of F. ceilanica s.l. for Java, the Lesser Sunda Islands, the Maluku Islands, New Guinea, the Philippines, Sulawesi and Sumatra. Without taxonomic clarity, conservation assessment carried out for F. ceilanica, a “widespread” artificial entity, will be misleading as real biological entities were concealed under the broad species concept. In contrast, an accurate taxonomic assessment of a taxon will highlight the actual conservation requirements of the species, if necessary, to ensure adequate protection and management strategies provided.