Artabotrys pachypetalus sp. nov. is described from Guangdong, Guangxi, Guizhou, Hunan and Jiangxi in China. A detailed description, distribution data, along with a color plate and a line drawing are provided. In China, specimens representing this species were formerly misidentified as A. multiflorus or A. hongkongensis (= A. blumei). Artabotrys blumei typically has a single flower per inflorescence, whereas both Artabotrys pachypetalus and A. multiflorus have multiple flowers per inflorescence. In addition, A. pachypetalus is readily distinguished from A. multiflorus in having thicker and shorter petals, and connivent and somewhat trigonal or terete inner petal blades. Artabotrys pachypetalus is most similar to A. punctulatus because both have multi-flowered inflorescences and similar petal length, but A. pachypetalus differs in having cream petals in vivo, connivent inner petal blades, and a short, raised rim above the inner petal claw. Artabotrys multiflorus should be excluded from the flora of China because none of the Chinese specimens of Artabotrys collected so far fall within the variation of A. multiflorus.

Annonaceae, Artabotrys, morphology, South China, taxonomy

Artabotrys R.Br. is one of the largest genera in the Annonaceae, with over 100 species of woody climbers distributed in Africa (including Madagascar) and Asia (Chen et al. 2019; Rainer and Chatrou 2021). Among the lianescent genera of the Annonaceae, Artabotrys is distinctive in possessing hooked inflorescence axes that facilitate climbing (Posluszny and Fisher 2000). Although the persistent inflorescence hook is a synapomorphy for the genus, identification at the species level is often complex. In general, the flowers of Artabotrys possess a uniform structure characterised by a tightly enclosed floral chamber (Chen et al. 2019, 2020). Both the outer and inner petals are concave at the base, with an expanded, generally flattened (sometimes terete or triquetrous) blade above the concave base. In addition, the inner petals have a projecting rim between its blade and the concave base (see fig. 9d in Chen et al. 2020); the elaborate rims of the three inner petals are tightly connivent, forming a dome around the reproductive organs. The concave base of the petals has been traditionally referred to as the ‘claw’ by Sinclair (1955) and widely followed by recent workers (e.g., Turner 2012; Prabhu et al. 2015; Turner and Utteridge 2015; Chen et al. 2018; Chen and Eiadthong 2020), although the term ‘claw’ is also used (in a strict sense) to refer to the narrowed, stalk-like, basal portion of the inner petals of other genera, e.g. Mitrephora (Blume) Hook.f. & Thomson, Monodora Dunal and Pseuduvaria Miq. (Sinclair 1955; Su and Saunders 2006; Couvreur 2009; Weerasooriya and Saunders 2010). The terms ‘claw’ and ‘blade’ have been explained in detail by Chen et al. (2020) in the context of the pollination biology of Artabotrys. To date, eight Artabotrys species have been recorded in China, viz. A. fragrans Jovet-Ast, A. hainanensis R.E.Fr., A. hexapetalus (L.f.) Bhandari, A. blumei Hook.f. & Thomson (recorded as A. hongkongensis Hance), A. multiflorus C.E.C. Fisch., A. pilosus Merr. & Chun, A. punctulatus C.Y.Wu ex S.H.Yuan, A. rhynchocarpus C.Y.Wu ex S.H.Yuan (Li and Gilbert 2011).

During our field exploration in Yangchun City in Guangdong Province, we collected an Artabotrys species with multi-flowered inflorescences. Morphological comparison with herbarium specimens from China revealed that the newly collected specimen is conspecific with many specimens that were previously identified as A. multiflorus or A. hongkongensis. However, our Yangchun collections (and the above-mentioned specimens) do not match type specimens or descriptions of A. multiflorus and A. hongkongensis. Originally described from Myanmar (Fischer 1937; Kress et al. 2003; Turner 2015), Artabotrys multiflorus also occurs in adjacent forests in Thailand (Chen et al. 2018; Chen and Eiadthong 2020) and is distinct in having multiple flowers per inflorescence and lanceolate petals (Fischer 1937; Chen and Eiadthong 2020). In China, A. multiflorus was first incorrectly recorded by Xu and Li (1985) based on three collections from Guizhou Province (Z.R. Xu. L1224, S827, L437, SYS). Subsequently, more specimens in China were misidentified as A. multiflorus. Thus, Li and Gilbert (2011) recorded this species in the Flora of China with a quite wide distribution in Guangdong, Guangxi, Guizhou and Yunnan. Originally described from Hong Kong, A. hongkongensis is also reportedly widespread in China, where it is known from Hunan, Guangdong, Guangxi, Yunnan, Guizhou and Hainan provinces (Li and Gilbert 2011); it is also recorded in Vietnam (Bân 2000). This species usually bears a single-flowered inflorescence (rarely two-flowered). However, this name has now been synonymized with A. blumei Hook.f. & Thomson (Turner 2018).

Apart from Yangchun, flowering individuals are also found in Ruyuan in Guangdong Province, as well as Mulun National Nature Reserve in Guangxi. Based on detailed comparisons of available living material as well as herbarium specimens, we confirm that our collections and many specimens misidentified as A. multiflorus and A. hongkongensis in China represent a new species described here as Artabotrys pachypetalus. In addition, A. multiflorus is not present in China, as none of the Chinese specimens of Artabotrys collected so far fall within the variation of A. multiflorus.

Physical and scanned herbarium specimens of Artabotrys from 15 Chinese herbaria (BNU, CCNU, GF, GFS, GZAC, GXMG, GXMI, GZTM, HGAS, HITBC, IBK, IBSC, KUN, PE and SYS; acronyms according to Thiers 2021) were studied in detail and compared with the type specimens and descriptions of A. multiflorus, A. blumei and other similar species from China and Indochina. Fresh material was collected from Chun-wan Town, Yangchun City of Guangdong, China. The morphological study was based on fresh material and herbarium specimens, but the measurements in the description below are based on dried material. The distribution, habitat and phenology of the new species are based on field observations and specimen records.

Artabotrys pachypetalus is similar to A. multiflorus in having multiple flowers per inflorescence (Figs 1, 2), but the two species can readily be distinguished by their disparate petal morphology. The petals of A. multiflorus are 18–30 mm long, and 2–5 mm wide (Fischer 1937; Chen and Eiadthong 2020) whereas the petals of A. pachypetalus are much shorter (9–17 mm long, 3–8 mm wide). The petals of A. pachypetalus are also considerably thicker (1–2 mm thick when dry) than those of A. multiflorus (less than 1 mm thick when dry). In addition, the inner petal blades of A. pachypetalus are connivent and somewhat trigonal or terete whereas the inner petal blades of A. multiflorus are spreading and planar. Xu and Li (1985) recorded A. multiflorus in China based on a few collections from Li-bo Hsien, Guizhou Province (Z.R. Xu L1224, S827, L437, SYS). We failed to locate these specimens in SYS, but found a duplicate of Z.R. Xu L1224 in IBSC as well as many other collections from the same locality, i.e., Mao-lan National Nature Reserve, Guizhou Province, and hence were able to confirm that they are conspecific with the new species described that we collected in Guangdong and Guangxi. The misidentification probably arose in the absence of careful comparison against the type specimens of A. multiflorus.

Some specimens representing A. pachypetalus are also misidentified with the name A. hongkongensis in China, which is now a synonym of A. blumei (Turner 2018). The distinction between these two species is quite straightforward: the inflorescence of A. blumei usually consists of a single flower or rarely two (Li and Gilbert 2011), whereas that of A. pachypetalus has 6–10 co-occurring flowers (Figs 1C, F, 2A). Additionally, A. blumei can be distinguished by having planar (vs. somewhat trigonal or terete), spreading (vs. connivent) inner petal blades, subglabrous (vs. sparsely to densely pubescent) pedicels and abaxial surface of sepals, and thin pericarp (< 1 mm thick vs. 2–3 mm thick).

Although A. pachypetalus is most similar to A. punctulatus, it has not been confused with the latter. This is probably because A. punctulatus is restricted to the montane forests of Yunnan in China and is characterized by densely punctate abaxial leaf lamina. The similarities and differences between the two species are stated in the diagnosis.

Besides the above-mentioned species, A. pachypetalus is also similar to Artabotrys hienianus Bân from Vietnam, both with multiple small hairy flowers per inflorescence (Bân 2000). The two species differ, however, in the number of the flowers per inflorescence, petiole length, leaf shape, and thickness of the pericarp. The inflorescence of A. hienianus has more flowers borne on shorter pedicels (5–6 mm), leading to densely clustered flowers on the inflorescence (Bân 2000), whereas that of A. pachypetalus has fewer flowers borne on longer pedicels (6–15 mm, Figs 1C–F, 2A), leading to loosely clustered flowers on the inflorescence (Figs 1C, 2A). Artabotrys hienianus has longer and broader leaves (12–17 × 5–6 cm), whereas A. pachypetalus has smaller and narrower leaves (10–16 × 3.5–5 cm). In addition, A. hienianus has a thin pericarp (Bân 2000), whereas that of A. pachypetalus is quite thick (2–3 mm thick) (Fig. 1L).

This research was supported by a grant from National Natural Science Foundation of China (Grant no. 31872646) awarded to Bine Xue and Forestry Scientific Technology Innovation Project of Guangdong Province (No.2020KJCX010). We are grateful to the curators of BNU, CCNU, GF, GFS, GZAC, GXMG, GXMI, GZTM, HGAS, HITBC, IBK, IBSC, KUN, PE and SYS for permission to access their collections. Thanks are also given to Ya-Nan Guo and Wei Guo for field work assistance, Yun-Xiao Liu for the excellent line drawing, Yun-Yun Shao for preparing Fig. 1, Bao-You Huang for sharing his photographs of Artabotrys collection in Guangxi, Jun-Yan Xu for sharing his photographs of Artabotrys collection in Jiangxi, and Lars Chatrou and Thomas Couvreur for constructive comments on an earlier version of our manuscript.