New combinations and synonyms in Rehderodendron (Styracaceae)

Abstract We demonstrate with morphological characters that the species Pterostyrax burmanicus W.W.Sm. & Farrer and Parastyrax macrophyllus C.Y.Wu & K.M.Feng (Styracaceae) are best placed in the genus Rehderodendron Hu. Rehderodendron burmanicum (W.W.Sm. & Farrer) W.Y.Zhao, P.W.Fritsch & W.B.Liao, comb. nov. and Rehderodendron macrophyllum (C.Y.Wu & K.M.Feng) W.Y.Zhao, P.W.Fritsch & W.B.Liao, comb. nov., are created. We also provide a lectotype for R. macrophyllum. These revisions result in the reduction of Pterostyrax Siebold & Zucc. to three species and this genus is no longer considered to be documented from Myanmar. Further, Parastyrax W.W.Sm. becomes a monotypic genus comprising only P. lacei (W.W.Sm.) W.W.Sm., distributed in Kachin State, northeast Myanmar and Yunnan Province, south-western China.


Introduction
Rehderodendron Hu (Styracaceae) is a genus of 6 to 10 species of trees native to China, Vietnam and Myanmar (Ming 1983;Hwang and Grimes 1996;Fritsch et al. 2001;Fritsch 2004;Zhao et al. 2019). It can be easily distinguished from the other genera of the Styracaceae by its large cylindrical fruit harbouring an endocarp with many irregular rays intruding into the mesocarp (Zhao et al. 2019). Species number and boundaries in the genus are poorly understood. To address this problem, we have been conducting phylogenetic and taxonomic research on Rehderodendron since 2017 with the goal of creating a comprehensive taxonomic revision of its species.
In the process of carrying out this research, we have discovered that two species of Rehderodendron have been mistakenly assigned to the genera Pterostyrax Siebold & Zucc. (Wu and Li 1965) and Parastyrax W.W.Sm. (Smith 1920). Parastyrax, Pterostyrax and Rehderodendron share such features as fertile shoots strictly lateral, hypanthium adnate to the ovary wall through the entire length of the ovary wall and upper ovules apotropous and lower ovules epitropous (Fritsch et al. 2001). Pterostyrax reportedly differs from the other two genera by, for example, flowers arranged on one side of ultimate inflorescence branches (versus not strictly on one side) and small ribbed or winged fruit with a prominent rostrum and mesocarp absent (versus larger fruit with rostrum not prominent and mesocarp present; Hwang and Grimes 1996;Fritsch et al. 2001). Parastyrax reportedly differs from Rehderodendron by, for example, petals connate at their bases only (coherent; versus connate distinctly beyond their bases), stamens all subequal to equal in length (versus distinctly unequal), style solid (versus hollow), smaller fruit size, fruit surface with conspicuous lenticels and endocarp without rays intruding into the mesocarp (Hwang and Grimes 1996;Fritsch et al. 2001

Materials and methods
This contribution is primarily based on a review of literature and examination of herbarium specimens. Herbarium specimens examined were both the physical specimens at SYS and the digital images of specimens from nine other herbaria (BRIT, E, HITBC, K, KUN, L, P, PE and WUK; herbarium acronyms as in Thiers (2020)). Herbarium numbers are barcode numbers unless otherwise indicated. Where available, physically checked specimens are indicated by "!" after the herbarium acronym and digital images that have been examined are indicated by "image!" appearing after the herbarium accession number or barcode number. Additionally, fieldwork to observe living plants of Parastyrax macrophyllus was undertaken in Hekou County by the first author in March 2020. The new combinations proposed herein accord with the rules and recommendations of the International Code of Nomenclature Article 41 and Recommendation 41A (Turland et al. 2018 Discussion. Pterostyrax burmanicus was described on the basis of a single collection, R. Farrer 803, from Langyang, eastern Upper Burma (Smith 1920;Fig. 1). From then until the present work, it had been accepted as a species of Pterostyrax Siebold & Zucc. endemic to Myanmar (Wu and Li 1965; Chen and Chen 1996; Kress et al. 2003). The species was not included in the phylogenetic analysis of Fritsch et al. (2001). Although the fruit of P. burmanicus is unknown, its vegetative and floral characters all agree with those of Rehderodendron (Table 1) and, in combination, match no other genus of the Styracaceae (Hwang and Grimes 1996;Fritsch et al. 2001). Four critical characters differ from those of Pterostyrax, but agree with those of Rehderodendron: P. burmanicus has faintly serrulate leaf blade margins (versus dentate in Pterostyrax), fertile shoots comprising only inflorescences (Figs 1, 2A; versus leaves (proximally) and inflorescences (distally); Fig. 2B), flowers that are not strictly arranged on one side of the inflorescence branches (Figs 1, 2A; versus arranged on one side of ultimate inflorescence branches; Fig. 2B) and petals that are distinctly connate beyond their bases (versus connate at their bases only (coherent)). On this basis, we conclude that P. burmanicus is best placed in Rehderodendron (Table 1).
Rehderodendron burmanicum is very similar to R. microcarpum K.M.Feng ex T.L.Ming, a species of far north-western Yunnan Province, China and northern Kachin State, Myanmar (Ming 1983;Zhao et al. 2020). Both species are deciduous trees with a paniculate and densely stellate-pubescent inflorescence, filaments connate well below the middle and a pubescent style. The young leaves of R. burmanicum have sparsely-scattered stellate pubescence and the stamens are longer than the petals (Figs 1, 2A), whereas the young leaves of R. microcarpum are densely covered with stellate pubescence and the stamens are shorter than the petals (Fig. 2C). However, the only known locality of R. burmanicum is in Chipwi Township, east-central Kachin State, Myanmar, which is well outside of the Two new combinations Rehderodendron burmanicum and Rehderodendron macrophyllum 83 known geographic range of R. microcarpum. Properly assessing the species boundaries or whether such exist between these two species must await more specimen collections and the generation of morphological and molecular data at the population level.
"Parastyrax burmanicus W.W.Sm." is a name listed in Kress et al. (2003). We could not find any original published literature for this species or digitised specimens identified as such and conclude that it is a nomen nudum, likely resulting from a publication error meant as Pterostyrax burmanicus.   (Ming 1983;Hwang 1987;Hwang and Grimes 1996). In the original description, Wu and Feng (in Wu and Li 1965) indicated that its inflorescence is very short and its fruit is drupaceous with a fleshy "exocarp" (likely the mesocarp instead), which would be consistent with the characters of Parastyrax. We re-examined the original material, including the paratype collection Cai Ke-Hua 533 (KUN [acc. # KUN0026239!]), which possesses the only fruit gathered of the material cited in the protologue (Wu and Li 1965). We found that the fruit of this collection is clearly immature, so its internal structure, which would otherwise help to confirm the placement of the species to genus, could not be determined. In March 2020, we conducted fieldwork in Binglangzhai, the type locality of P. macrophyllus. There, we discovered plants that possess the same vegetative features as those in the type material of the species, yet with fruit in which the endocarp has many irregular rays intruding into the mesocarp, as in Rehderodendron (Figs 4D, E) and unlike in P. lacei (W.W.Sm.) W.W.Sm., the type species of Parastyrax. On this basis, we conclude that P. macrophyllus is best placed in Rehderodendron. Parastyrax must be redefined on characters that exclude those of R. macrophyllum; we include diagnostic characters for this purpose in Table 1, based on several previous studies (Smith 1920;Dickison 1993;Hwang and Grimes 1996;Fritsch et al. 2001).
In the protologue, Yu Ping-hua s.n., a flower gathering deposited in KUN, was indicated as the type. Of the three sheets comprising this collection, KUN0026238