﻿Inflorescences of Fargesiaangustissima T.P. Yi and Yushaniapauciramificans T.P. Yi (Poaceae, Bambusoideae) shed light on the taxonomy of the Sino-Himalayan alpine bamboos

﻿Abstract The taxonomy of the Sino-Himalayan alpine bamboos is controversial due to their complex evolutionary history and further complicated by the scarcity of inflorescence. Here, we supplement the description of the inflorescence of Fargesiaangustissima T.P. Yi and Yushaniapauciramificans T.P. Yi, which shed light on the taxonomy of Fargesia Franchet, Borinda Stapleton and Yushania Keng. F.angustissima has compressed inflorescence unilateral stretching out from reduced spathe, showing a transitional state between species with condensed inflorescence embraced by spathe-like bracts and species with open inflorescence without bracts. Considering that extensive gene flow existed between several clades of Fargesia found in recent studies, a broadly-defined Fargesia s. l. should be adopted. Meanwhile, the inflorescence of Y.pauciramificans has typical characteristics of Yushania, such as axilla with tuberculate glands, rachilla internodes ciliate and cylindrical florets, supporting the delimitation of Yushania.


Introduction
Although the taxonomy of bamboos has entered a new stage since the proliferation of molecular phylogenetic and phylogenomic studies (e.g., Bamboo Phylogeny Group 2012; Attigala et al. 2016;Zhang et al. 2020;Ye et al. 2021b), morphological characters play an important role in the naming and identification of the species. Reproductive characters are traditionally assumed to be critical in bamboo evolution and taxonomy, especially at the generic level (Li et al. 2006). However, due to variable blooming intervals, ranging from a couple of decades up to 120 years (Janzen 1976), many bamboo species were described without inflorescence information (Yi 1986(Yi , 1988aStapleton 1994;Keng and Wang 1996;Ohrnberger 1999;Li et al. 2006). This has caused the confusion in the definition of some genera, including some Sino-Himalayan alpine bamboos, e.g., Fargesia Franchet, Yushania Keng f. and Borinda Stapleton (Guo et al. 2002;Stapleton 2021).
Fargesia was delimited as having short-necked pachymorph rhizome with unicaespitose culms and compressed inflorescence subtended by several small or large spathes, while Yushania has long-necked rhizome with diffuse culms and open inflorescence without bracts (Li et al. 2006;Yi et al. 2008;Shi et al. 2022). Borinda was described as clumping temperate bamboos, similar to Yushania for its inflorescence with reduced bracts, and Fargesia for its short-necked pachymorph rhizome (Stapleton 1994). Moreover, most species of Borinda were transferred from Fargesia (Stapleton 1998(Stapleton , 2006(Stapleton , 2021. Although morphological differentiation of rhizome and inflorescence has been used to distinguish these genera, the intermediate state of rhizome and inflorescence between them makes the genus delimitation very ambiguous. As a result, the bamboo accounts of the "Flora of China" recognize two genera, i.e., Fargesia and Yushania (Li et al. 2006).
In our recent molecular analyses of this taxonomically difficult group based on the double digest-restriction-site associated DNA sequencing (ddRAD) analyses (Ye et al. 2019), Yushania was resolved as a well-supported monophyletic lineage, demonstrating the phylogenetic importance of the rhizome type. If considering the rhizome type alone, Fargesia yunnanensis Hsueh & T.P. Yi needs to be transferred into Yushania for its generally long rhizome neck (12-35 cm). Actually, this species was nested in the 'Fargesia1' clade in the analysis of Ye et al. (2019) with high support. Therefore, more information on inflorescence knowledge should be provided for the delimitation of Yushania.
Fargesia was resolved as a polyphyly (Wang et al. 2017;Zhang et al. 2019;Zhou et al. 2019;Zhou et al. 2020;Ye et al. 2021a) and divided into several clades with high support in the recently ddRAD analyses (Ye et al. 2019). Stapleton (2021) transferred species in the 'Fargesia3' + 'Fargesia4' + F. angustissima clade of Ye et al. (2019) and several species sampled by Zhang et al. (2019) into Borinda based on the molecular phylogenetics and some floral and vegetative characteristics. Concurrently, Stapleton (2021) considered that Fargesia s. s. possesses tightly unilateral racemes and only distributed along the Qinling Mountains. In this case, most of the species originally described in Fargesia could not be retained in this genus. Nevertheless, several species which were transferred into Borinda shared the floral characteristics of Fargesia s. s., with raceme enclosing by spathe-like sheaths and protruding from unilateral side, such as F. edulis Hsueh & T.P. Yi and F. adpressa T.P. Yi (Li et al. 2006;Shi et al. 2022). This indicates that limited reproductive characters cannot distinguish Borinda from Fargesia appropriately. Therefore, more knowledge of reproductive features should be provided to improve our understanding of the relationship of Borinda, Fargesia and Yushania.
In recent field surveys, we collected the floral and vegetative specimens for two bamboo species. A supplementary description of the inflorescence of these two species is presented here, providing new information on the delimitation of alpine bamboos.

Materials and methods
We collected two specimens with both floral and vegetative organs during our field work in Yunnan (YXY2020023) and Sichuan (WL2021001), China. Morphological studies were based on the living individuals in the field, specimens, and literature (Yi 1985(Yi , 1988aKeng and Wang 1996;Li et al. 2006;Shi et al. 2022

Results
According to our observation and comparison of the type specimens and original literature (Yi 1985(Yi , 1988b and bamboo accounts of "Flora Reipublicae Popularis Sinicae" (Keng and Wang 1996), "Flora of China" (Li et al. 2006) and "Illustrated Flora of Bambusoideae in China" (Shi et al. 2022), we identified specimen YXY2020023 to be Yushania pauciramificans T.P. Yi based on paniculate inflorescence on terminating leafy branches without spathes subtending, pachymorph rhizomes with long neck (20-50 cm in length), culms 2-3.2 m, internodes terete, branches 1-3 at lower nodes, 5-6 at upper, culm sheaths cartilaginous, with erect gray setae and absent auricles. WL2021001 was identified to be Fargesia angustissima T.P. Yi according to panicles on terminating leafy branches subtended by slightly expanded bracts, pachymorph rhizomes with short neck (2-5 cm in length), unicaespitose culms with fine ridged internodes, culm node less prominent than sheath scar, culm sheaths persistent, which were longer than internodes, narrowly triangular, apically papery, linear, and narrowed for distal 1/3-1/2 of length, sparsely brown setulose, leaf blade abaxially proximally white-gray pubescent. All voucher specimens were deposited at Kunming Institute of Botany, Chinese Academy of Sciences (KUN), and epitype of these two species are also designated here (Turland et al. 2018).
Phenology. New shoots August. Flowering April to June. Distribution and habitat. Yushania pauciramificans is known from Xinping, south-central Yunnan, mainly distributed in the evergreen broadleaved forest at an elevation of 2250-2500 m.

Discussion
Reproductive features play an important role in the delimitation of the genera of alpine bamboos, and can improve our understanding of the relationship of Borinda, Fargesia and Yushania. Although Fargesia angustissima was recombined into Borinda in the analysis of Stapleton (2006), its inflorescence characters provide new insight in its delimitation. According to the description, the leaf sheath underneath the inflorescence of F. angustissima is inflated but smaller than the spathe of some species of Fargesia s. s., such as F. funiushanensis T.P. Yi, F. qinlingensis T.P. Yi & J. X. Shao, while more similar to those with spikelets stretching out from one side of sheaths and arranging relatively loosely (Zhang and Ren 2016). These characteristics indicate that the inflorescence of F. angustissima is in a transitional state between compressed and open ones (Fig. 3). Moreover, F. angustissima possesses an independent position on the molecular phylogenetic trees and has a special habitat (Ye et al. 2019;Shi et al. 2022). The florets, phylogenetic position, and habitat of F. angustissima imply that it is different from other species of Fargesia and could be treated as a new genus, albeit following a narrower genus concept. Additionally, as currently circumscribed, Borinda is a paraphyletic group without distinct synapomorphies despite some hair-like vegetative morphology and gene flow occurring frequently between it and Fargesia1 clade (Ye et al. 2019;Stapleton 2021;Ye et al. 2021a). Considering the variable vegetative characters and insufficient reproductive features, any new combination of these alpine bamboos should be made cautiously, especially when extensive gene flow exists. Thus, we support the "Flora of China" in adopting a broadly-defined Fargesia s. l., rather than Borinda to minimize nomenclatural change.
The flowering branches and flowers of Yushania pauciramificans are similar to those of species of Yushania, both have similar paniculate inflorescence, axilla with tuberculate glands, ciliate rachilla internodes and florets. Combined with the reduced gene flow between Yushania and Fargesia revealed by D-statistic tests, the inflorescence state of Y. pauciramificans supports the monophyly of Yushania further (Ye et al. 2019).