Twenty-six additional new combinations in the Magnolia (Magnoliaceae) of China and Vietnam

Abstract In accordance with the previous reduction of the remaining genera of subfamily Magnolioideae (Magnoliaceae) into the genus Magnolia, twenty-six new nomenclatural combinations are formally made by transferring to Magnolia some additional Chinese and Vietnamese taxa from the segregate genera of Manglietia, Michelia and Yulania. The following nine new combinations are created from Manglietia, namely Magnolia admirabilis, M. albistaminea, M. guangnanica, M. jinggangshanensis, M. maguanica, M. pubipedunculata, M. pubipetala, M. rufisyncarpa and M. sinoconifera. Also, twelve new combinations are created from Michelia, namely Magnolia caloptila, M. caudata, M. fallax, M. gelida, M. hunanensis, M. maudiae var. rubicunda, M. multitepala, M. platypetala, M. rubriflora, M. septipetala, M. sonlaensis, M. xinningia. Finally, five new combinations are created from Yulania, namely Magnolia baotaina, M. pendula, M. pilocarpa var. ellipticifolia, M. puberula and M. urceolata.


Introduction
Richard B. Figlar (2012), a past president and present scientific advisor of Magnolia Society International, provides a concise but thorough background to the complex generic history of subfamily Magnolioideae of family Magnoliaceae, starting with J.E. Dandy in the early part of the previous century. This pre-eminent British plant taxonomist, specialising in Magnoliaceae, recognised the family as consisting of 2 tribes, the Liriodendreae representing the single distinct genus Liriodendron, with the remainder of the family, about which Dandy (1927) acknowledges there had never been uniformity of opinion, forming the Magnolieae, comprising 9 genera, which he subsequently increased by 2.
Revisions were to follow Dandy's death in late 1976, including the classification of the leading Chinese Magnoliaceae researcher, Liu Yu-hu (aka Law Yuh-wu). His proposed Taxonomic System of Magnoliaceae (Law 1984), republished in Magnolias of China in the year he died (Liu et al. 2004), basically added a further 4 genera to those of Dandy. Representatives of 10 of the 15 genera included in subfamily Magnolioideae in Liu's classification occur in China.
H.P. Nooteboom, who was to succeed Dandy at the forefront of Magnoliaceae research in Europe, realised that his predecessor had been mistaken in his interpretation of certain morphological characters and thus commenced his reduction of Magnolioideae (Nooteboom 1985), to just 6 genera. Ultimately, with the advent of molecular DNA sequencing data (Azuma et al. 1999, 2000, 2001, Kim et al. 2001, Nie et al. 2008, Wang et al. 2006, Kim and Suh 2013, combined with comparative morphological research (Figlar 2000, Figlar andNooteboom 2004), showing the remaining genera, including Manglietia Blume and Michelia Linnaeus, residing among the other sections of Magnolia, Figlar and Nooteboom proposed a new classification system in their 2004 paper. Their new system includes Magnolia at the head of a now monogeneric Magnolioideae subfamily comprising subgenus Magnolia with 8 sections and 7 subsections, subgenus Yulania with 2 sections and 6 subsections, and subgenus Gynopodium with 2 sections.
This system was not followed in the Flora of China treatment of Magnoliaceae (Xia et al. 2008), where previously recognised genera such as Manglietia and Michelia were retained, two former sections of Magnolia were given generic status as Houpoea N.H. Xia & C.Y. Wu and Oyama (Nakai) (Wu et al. 2015) and Yulania dabieshanensis T.B. Zhao, Z.X. Chen & H.T. Dai (Dai et al. 2012).
However, Figlar and Nooteboom's (2004) classification system is now widely accepted by the scientific community, with many authors following this broad view of Magnolia, such as Arroyo et al. (2013), Ninh et al. (2020), Pérez et al. (2016) and Zou et al. (2020). Figlar (2012) advised against the alternative classification system now operating: In a one genus system only Manglietia, Michelia and 3 minor genera require new names in Magnolia. In a 13 genera system, it would be necessary to dismantle the largest and most well-known genus, Magnolia, and rename the constituents into 10 new genera. That would be enormously destructive to the long-established Magnolia-centric nomenclature and literature, causing unnecessary and undesirable consequences to science, conservation and horticulture.
With this in mind, 26 new combinations are created here, representing nine species of Manglietia, eleven species and one variety of Michelia, plus four species and one variety of Yulania. Most of these taxa were named and described over the past decade, but include some older previously synonymised, now reinstated taxa, that are herein transferred to Magnolia, as will be numerous other taxa in a sequel to this paper Png 2019a, 2020).

Materials and methods
The new combinations proposed in this paper are made in compliance with the rules and recommendations of the 2018 International Code of Nomenclature for algae, fungi and plants (ICN), known as The Shenzhen Code (Turland et al. 2018), in particular ICN Article 41 and Recommendation 41A in respect of new combinations.
Where available, digital images of type specimens of newly named taxa posted to the internet at the websites of various herbaria have been sighted and these are indicated in the text by 'online image!' appearing after the herbarium acronyms whose representative names are listed in the appendix following the references. Additional sighted specimens are indicated by '!' after the herbarium acronym.
Consultation of the relevant literature was made to determine whether a number of taxa previously determined as synonyms of earlier named taxa were, in fact, genuine independent species or varieties as they had been originally described. Differences in numerous morphological features, natural distributions and/or elevations and where appropriate, the incompatible phenology of flowering and/or fruiting periods, are tabulated and referred to in the notes under the relevant taxa to fully substantiate their independent status.
Floras and other literature dealing with the Magnoliaceae of China that have been consulted during this study are cited in the text, with some of the more important sources of information including a number of papers by Dandy (1928Dandy ( a-c, 1930, The Magnoliaceae of China (Chen and Nooteboom 1993), Magnoliaceae in Flora Reipublicae Popularis Sinicae (Law et al. 1996), Magnolias of China (Liu et al. 2004), Magnoliaceae in Flora of China Vol. 7 (Xia et al. 2008), A Taxonomic Revision of the Magnoliaceae from China (Sima 2011) and the recent Ex Situ Cultivated Flora of China : Magnoliaceae (Yang et al. 2016), which documents the diversity of Magnoliaceae plants in Chinese botanical gardens.

Discussion
The transfer of the above twenty-six taxa to Magnolia is necessary following the present near universal acceptance by the scientific community and horticultural industry that the Magnolioideae is one of two monogeneric subfamilies within Magnoliaceae and the fact that the majority of resulting new combinations and names arising from the relegation of Manglietia and Michelia into Magnolia have previously been made by various authors such as Figlar (2000) for the majority of the Michelia species, with Sima (2001) transferring some additional Michelia species, Kumar (2006) transferring the majority of Manglieta species, Nooteboom transferring a number of species from both the previous genera plus Yulania in Flora of China Vol. 7 (Xia et al. 2008: 49-50) and most recently Callaghan and Png (2013) transferring species from these three genera that were mainly described and named subsequent to the publication of Flora of China.

Conclusions
To maintain these twenty-six predominantly recently described taxa in limbo in segregate genera will contribute to further instability and inevitable confusion in the scientific and popular literature, as well as within the botanical world and the horticultural industry, which has resulted from having two diverse systems operating simultaneously.
The authors would like to take this opportunity to suggest that to further substantiate their now reaffirmed species or varietal status, comparative DNA barcoding (Caddy-Retalic and Lowe 2012), should be undertaken of these and other taxa, often with small remnant populations and/or disjunct geographic distributions, that have been previously subsumed in synonymy under earlier-named species having much larger populations of widespread occurrence. As a result of becoming virtual non-entities, this can be detrimental to their conservation and ultimate survival in nature. Consequently their potential benefits to mankind, such as the medicinal properties that some Magnoliaceae species are known to possess, including present and prospective production of anti-cancer drugs and treatments (He et al. 2017, Huang et al. 2017, Lu et al. 2017, Ma et al. 2020, Prasad and Katiyar 2018, Zhang et al. 2020, are never assessed or realised.  Chen andNooteboom (1993: 1088), in which it was noted that specimens had not been seen. It was subsequently reduced to a synonym of Michelia fujianensis as noted above. It is recognised as a genuine species by Law et al. (1996: 189), Liu et al. (2004: 228), Deng and Yang (2015: 167), Yang et al. (2016: 237) and Sima (2011: 234), wherein M. caloptila is in Michelia subsection Micheliopsis, series Micheliopsis and M. fujianensis is in Michelia subsection Velutinae. Differences between the abaxial indumentum of the 9-16 cm long leaves of M. caloptila and of the 6-11 cm long leaves of M. fujianensis are illustrated in Plate 3-2E (M. caloptila) and Plate 3-3E (M. fujianensis) of Sima (2011: 325;326). Further substantiation of the specific status of M. caloptila is evident from a comparison of its morphological features with those of M. fujianensis, as shown in Table 1 on the following page.

Taxonomic section
Note 2. As a consequence of the above substantiation of the species status of Michelia caloptila, plus the past reduction to Magnolia of the remaining genera of subfamily Magnolioideae, Michelia caloptila is here transferred to Magnolia.  fig. 1 (2015).
Chinese name. 尾叶含笑 meaning "caudate-lobed michelia", referring to shape of leaf apex.  The differentiating features of Michelia fallax, whose flower is unknown, are cited from Dandy (1928c) and CVH (2017)   Note. The holotype and isotype specimens of Manglietia guangnanica could not be found by herbarium staff at IBSC, nor could the paratype specimen at MO be located (Jim Solomon, pers. comm., July 2019). However, the paratype that was received from IBSC, R.Z. Zhou (Zhou Ren-zhang) 9304 collected at 1600 m, inexplicably has the locality and collection date as for the holotype / isotype above and not Mount Houshan on the 4 October 1993 as is noted in the 2014 paper for this paratype. Note 1. The holotype specimen was irretrievably damaged during repeated relocations of the HFBG herbarium (Yan Lihong, pers. comm.). Photographs were sent in its place.
Note 2. The numerous known differentiating features compiled in Table 3 below confirm Michelia hunanensis as an independent species and not a variety of Magnolia maudiae, nor a synonym of Michelia cavaleriei var. cavaleriei as noted above. The distinguishing features of Michelia hunanensis are cited from Peng et al. (1995). Those of M. maudiae are cited from Dunn (1908), supplemented by Chen andNooteboom (1993:1072) (1) Note. Manglietia maguanica is listed as a synonym of M. insignis in Chen & Nooteboom (1993) and subsequently by the authors noted above. However, both are recognised as independent species in the majority of the more recent Chinese publications, including Liu et al. (2004: 164, 156), Xing et al. (2009: 198, 196), Lu (2009), Sima (2011: 98, 102), Deng and Yang (2015: 48, 54) and Yang et al. (2016: 192, 181). Note 3. Michelia multitepala is sufficiently distinct from M. doltsopa (Candolle 1818), to justify its species status, as shown by their known differentiating features compiled in Table 4 below. Additionally, M. multitepala is known only to occur at 1300-1500 m on Fadou Mountain in the southeast of Yunnan Province, whereas M. doltsopa occurs between 1500-2300 m throughout its widely dispersed geographical area from Yunnan to N Myanmar, NE India, Bhutan and SE Xiyang (Liu et al. 2004: 242), or 2100-2500 m from central Nepal and Burma (Myanmar) to Sichuan and Yunnan (Polunin and Stainton 1999: 19). As a consequence of the substantiation of its specific status, Michelia multitepala is here transferred to Magnolia in accordance with the past reduction of the remaining genera of subfamily Magnolioideae to the genus Magnolia.   (1818) Note 2. Yulania pilocarpa var. ellipticifolia is sufficiently distinguished from Y. pilocarpa to maintain its varietal status by the following features: indumentum of the branchlets (densely pubescent, later glabrous vs. glabrous [Law et al. 2004: 93]); the leaf shape (elliptical, rarely inverted-triangular vs. obovate to broadly obovate [Law et al. 2004]) and the shape and size of the inner 6 tepals (petaloid, 5-7 × 2-3.2 cm vs. nearly spathulate, 7-10 × 3-5 cm [Law et al. 2004]). Additionally, the two taxa are geographically isolated (central Henan Province vs. SE Hubei Province). The illustration of the leaves accompanying the original description of Yulania pilocarpa var. ellipticifolia : fig.1D) shows them to be in stark contrast to the leaves of Magnolia pilocarpa illustrated in Liu et al. (2004: 93). Note 1. Dandy (1928c: 130) provides relevant background information concerning the type collections of Michelia platypetala and M. fallax from the same general locality in Hunan Province in consecutive years and how they were both initially confused as the former species.
Note 3. Grimshaw and Bayton (2009: 500) record a personal communication received from Richard Figlar in 2007 advising that "this taxon (Magnolia maudiae var. platypetala) probably ought to be recognised at the specific level, as Magnolia platypetala, as it differs considerably from M. maudiae both in its hairiness and its later bud-break". Sima (2011: 327), illustrates the contrasting difference between the indumentum of the undersurfaces of the leaves of M. platypetala (Plate 3-4H) and that of the leaves of M. maudiae (Plate 3-4C). Additionally, in a study by Zhang and Xia (2007) on leaf architecture and its taxonomic significance in respect of subtribe Micheliinae of Magnoliaceae, the pronounced contrast in the leaves of Michelia platypetala and M. cavaleriei as revealed by stereoscopic magnified imaging (shown at figs. 36 and 37 in their paper), resulted in these authors concluding that these two taxa should be recognised as independent species". It is apparent that there is now an almost unanimous consensus of the species status of Michelia platypetala, which is confirmed by the comparison of its morphological features with those of M. cavaleriei compiled in Table 5 below. In view of its distinctive characteristics and accepting the majority recognition by the above-mentioned Chinese authors of Michelia platypetala as a genuine species, it is here transferred to Magnolia as a consequence of the past reduction of the remaining genera of subfamily Magnolioideae to the genus Magnolia. Note. The type specimens of Yulania puberula cannot be located at the Beijing herbarium of CAF (Wang Hongbin, pers. comm., March 2020). The distinguishing features of Michelia platypetala are mainly cited from Law et al. (1996: 177), Liu et al. (2004: 306) and Yang et al. (2016: 306) Zeng et al., Pakistan J. Bot. (6): 1917, 1919+ 1918 fig. 1 (2007).
Note. Manglietia pubipetala Q.W. Zeng is considered as conspecific with M. rufibarbata Dandy by the above authors. However, M. pubipetala can be sufficiently differentiated from M. rufibarbata Dandy to justify its species status, as shown by the comparative morphological features included in Table 6 on the following page  (adapted from Table 1, Zeng et al. 2007). M. pubipetala is therefore transferred to Magnolia consistent with the past reduction of the remaining genera of subfamily Magnolioideae to the genus Magnolia.  The differentiating features of Manglietia pubipetala are from Zeng et al. (2007) and those of M. rufibarbata are from Dandy (1928), supplemented by Liu et al. (2004: 190)  Note. While Michelia rubriflora is noted as a synonym of M. mediocris in Flora of China (Xia et al. 2008), the present authors agree with Wang and co-authors that Michelia rubriflora can be more than sufficiently differentiated from M. mediocris by the diagnostic features of these two species included in Table 1 of their paper (Wang et al. 2005), to substantiate its species status. A more comprehensive analysis of their differentiating features is compiled in Table 7 below. Michelia rubriflora also does not key out with the original validating description for M. subulifera (Dandy 1930:212), with which it shares synonymy under M. mediocris in Flora of China. Evidently an independent species, Michelia rubriflora is transferred in the present paper to the genus Magnolia by reason of the past reduction of the remaining genera of subfamily Magnolioideae to that genus. greyish-white or pale brown appressed pilose appressed grey or yellowish-brown tomentose leaf shape ovate-elliptic elliptic or elliptic-oblong leaf dimensions 5-9 × 2.5-3.5 cm 6-13 × 3-5 cm § leaf indumentum abaxially greyish-white or pale brown appressed pilose initially appressed greyish pubescent (greyishwhite appressed puberulent) ¶ leaf texture leathery thinly leathery ¶ lateral leaf veins 9-11 either side of midrib 12-15 either side of midrib stipular scars 1-2 mm long none ¶ petiole length and indumentum 1-2.5 cm, greyish-white or pale brown appressed pilose 1.5-3 cm § , initially appressed grey tomentellous, then glabrescent tepal number /colour 9, red 9-10 # , white ¶ tepal size and shape 2.5-3.5 × 1.0-1.2 cm, lanceolate 1.8-2.2 × 0.5-0.8 cm, spathulate § stamen length /colour 1.5-1.7 cm, red 1.0-1.5 cm § , yellowish-green gynophore not exserted above androecium extended well above androecium The differentiating features of Michelia rubriflora are from F.G. Wang et al. (2005) and those of M. mediocris are from Dandy (1928a), supplemented by Chen andNooteboom (1993: 1073)  Note. Manglietia rufisyncarpa is listed as a synonym of M. insignis in Flora of China (Xia et al. 2008), by Sima and Lu (2009) and by Sima (2011). However, the present authors agree with Wang and co-authors that M. rufisyncarpa can be more than sufficiently differentiated from M. insignis (Wall.) Bl. by the diagnostic characters of these two species compiled by Wang et al. (2004: Table 1), to substantiate its independent species status. Additionally, M. rufisyncarpa flowers from April-May whereas M. insignis flowers from May-June (Liu et al. 2004: 156). Also, among the many Manglietia photos in Magnolias of China, the bright red gynoecium of this species, alluded to in its Chinese name, is particularly noticeable as one of only a few exhibiting this colour, with M. insignis displaying a green gynoecium. Michelia fujianensis Q.F. Zheng. In: Xia and Deng (2002: 130) and Xia et al. (2008: 83), both p.p. quoad syn. Michelia septipetala Z.L. Nong.
Note. The holotype specimen of Michelia septipetala cannot be found at IBK (Xu Wei-bin, pers. comm., July 2019). However, M. septipetala can be easily differentiated from both M. fujianensis and M. caloptila, the 2 species under which it is noted in synonymy above, by the comparison of their morphological and phenological characteristics summarised in Table 8 below.  Law and Wu (1984) and Liu et al. (2004: 228) (1993). Chinese name. 那坡木莲 meaning "Napo manglietia" Type. CHINA. Guangxi Zhuang Autonomous Region: Guilin Botanical Garden (cultivated; introduced from Napo County, W Guangxi), 3 June 1991, Wei Fa-nan 1910.
Note 2. Some of the features distinguishing Manglietia sinoconifera from M. dandyi, under which it is questionably placed as conspecific in Flora of China due to uncertainty over its status (because the holotype could not be sighted), are listed in Table 9 below. M. sinoconifera (to 10 m) also does not key out with the description for the large-leaved M. megaphylla Hu & W.C. Cheng (1951), a tree to 40m (Liu et al 2004), with which it shares synonymy under M. dandyi in Flora of China. Manglietia sinoconifera is recognised as a genuine species in Yang et al. (2016: 213-214), wherein its introduction to Guilin Botanical Garden from Napo County is recorded as 1973 (18 years earlier than stated in the protologue).  (Xia et al. 2008), the 9-tepalled Michelia xinningia from Hunan, with a published height by the naming authors of 20 m, appears incongruously as a synonym of the ca. 12-tepalled M. cavaleriei var. cavaleriei with a height to 10 m (Liu et al. 2004: 229;Xia et al. 2008: 8;Deng and Yang 2015: 148). This would indicate that this remains about the maximum height of M. cavaleriei since being described as a small tree of 4-7 metres more than a century earlier (Finet and Gagnepain 1906: 573), based on a collection from Guizhou ca. 400 km distance from the type locality of Michelia xinningia in Hunan. This discrepancy in their heights indicates that M. xinningia was evidently meant to appear in Flora of China as a synonym of the  Xing et al. (2009: 212) and Yang et al. (2016: 331). Also, Michelia xinningia can easily be differentiated from M. foveolata, under which it is made a synonym by Sima and Lu (2009) and included as such in Sima (2011: 216), by the comparative features compiled in Table 10. Note 3. Bearing in mind the above discussion and comparative features, Michelia xinningia is an obviously distinct species. Therefore it is here transferred to Magnolia due to the past reduction of the previous segregate genera of subfamily Magnolioideae to the genus Magnolia.
Note 4. A search of the literature has found that Michelia xinningia is in cultivation at 4 Chinese botanical gardens, each in which M. platypetala and M. foveolata are also cultivated (Callaghan and Png 2019b

Herbaria references
The above herbarium acronyms and their institutes were located in the following publications: