Typification of 23 names in Eriobotrya (Maleae, Rosaceae)

Abstract As part of a comprehensive systematic study on the genus Eriobotrya and its close relatives from the E & SE Asia, new typifications of 23 names are presented here, along with some nomenclatural notes of the names involved. We lectotypified 22 names including accepted names and synonyms. They are: E.acuminatissima, E.bengalensisvar.angustifolia; E.bengalensisf.intermedia, E.brackloi, E.brackloivar.atrichophylla, E.ellipticavar.petelotii, E.fragransvar.furfuracea, E.glabrescens, E.grandiflora, E.henryi, E.oblongifolia, E.petiolata, E.platyphylla, E.poilanei, E.prinoides, E.prinoidesvar.laotica, E.salwinensis, E.serrata, E.stipularis, Hiptagecavaleriei, Photinialongifolia, Symplocosseguinii. One neotype of Photiniadubia was also proposed in this study, and E.pseudoraphiolepis and Mespiluscuila were identified as superfluous names. In addition, we also summarized the typification of 18 names for taxonomic reference: E.angustissima, E.balgooyi, E.condaoensis, E.×daduheensis, E.elliptica, E.fulvicoma, E.fragrans, E.glabrescensvar.victoriensis, E.hookeriana, E.latifolia, E.obovata, E.malipoensis, E.merguiensis, E.tengyuehensis, E.wardii, Mespilusbengalensis, Photiniadeflexa, and M.japonica.


Introduction
The application of names of taxa at the rank of family and below is determined by means of nomenclatural types (Shenzhen Code, Turland et al. 2018: Art. 7.1), and it is advisable to lectotypify those names with no type or sometimes even no original materials at all (Turland 2019). The type, usually a specimen or illustration, is the only entity permanently linked with the name. As one of the six principles in the International Code of Nomenclature for algae, fungi, and plants (Shenzhen Code, Turland et al. 2018), typification will pave the way for taxonomic stabilization of the names. The role of nomenclature in taxonomy was rightly expressed by Davis and Heywood (1963): "biologists must know what organisms they are working with before they can pass on information about them to other people -a function of taxonomy which makes stability of nomenclature an important consideration".
Several groups of the tribe Maleae (Rosaceae) are well-known because of the numerous north temperate fruits and ornamentals, such as apples (Malus Mill. spp.), pears (Pyrus L. spp.), serviceberries (Amelanchier Medik. spp.), chokeberries (Aronia Medik. spp.), loquats (Eriobotrya japonica (Thunb.) Lindl.), and photinias (Photinia Lindl. spp.). Due to the polyphyly of Sorbus L. and frequent hybridization within and between genera in Maleae, Christenhusz et al. (2018) argued for merging all members of Maleae into Pyrus L., returning to the Linnaean concept, and 849 new names and new combinations in Maleae were proposed. This extreme taxonomic strategy will hardly be appropriate for this ecologically and economically important lineage. Although we listed all new combinations proposed by Christenhusz et al. (2018) in this study, we prefer to recognize the genera in a traditional sense for better communication between plant scientists and the genera publics (Crane et al. 2017;Liu et al. 2019). Due to the rapid radiation and frequent hybridization Campbell et al. 2007), limited markers from plastome and/or nuclear based on Sanger sequencing could not provide enough resolutions for resolving the phylogenetic relationships among the genera in Maleae. With the development of next-generation sequencing, large datasets could be obtained at an affordable cost. Resolving the phylogenetic and evolutionary questions in Maleae is getting much easier and the Pyrus s.l. concept might be premature. For example, Liu et al. (2019) used the whole chloroplast genome and entire nuclear ribosomal DNA obtained via genome skimming approach to clarify the generic delimitation between Photinia Lindl. and its morphological allies in Rosaceae (Liu et al. 2019); additionally, this method has also been successfully employed to resolve the systematic problems in other angiosperm lineages (Wen et al. 2018;Valcárcel and Wen 2019). Therefore, phylogenomics based on next/thirdgeneration sequencing will provide an opportunity for clarifying the taxonomic and evolutionary problems in Maleae, especially Eriobotrya Lindl.
Loquats have been cultivated widely in the world as fruits and/or ornamentals. Eriobotrya consists of ca. 15-20 species, and is widely distributed from the Himalayas throughout continental southeast Asia to Japan and the islands of western Malesia (Kalkman 2004). The generic delimitation among the members of Maleae has been notoriously controversial for a long time, and efforts to resolve the intergeneric relationships have not been successful Campbell et al. 2007;Li et al. 2012;Lo and Donoghue 2012;Verbylaitė et al. 2006;Sun et al. 2018). As part of our integrative systematic studies of Eriobotrya and close relatives Liu et al. 2020), it will be necessary to typify the taxa described under Eriobotrya.

Materials and methods
We summarized the names listed in Tropicos, IPNI (International Plant Names Index), and the Plant List as the first step, and relevant literature has been consulted for the names treated in the present study. For the typification of each name summarized, all the relevant monographs and floras (e.g., Vidal 1965Vidal , 1968Vidal , 1970Gu and Spongberg 2003) have been checked as well as the additional original literature publishing the names and combinations (listed below). The rules governing holotype recognization and lectotypification followed McNeill (2014), and Turland (2019), and Turland et al. (2018). Thanks to the world's largest database of digitized plant specimens (JS-TOR: Global Plant), we checked 158 images of specimens in the herbaria all around the world, including A, B, BK, BM, C, E, HBG, K, L, M, MO, MSC, NY, P, TCD, UPS, VNMN, and WU. The authors consulted the potential type material in the following herbaria: CDBI, HITBC, IBK, IBSC, KUN, PE, SN, SYS, SZ, US, and WUK (Index Herbariorum 2019).

Typifications
Eriobotrya Lindl., Trans. Linn. Soc. London 13: 96, 102 (1821 in the protologue, however, he did not indicate the herbarium where the type has been deposited. We located three duplicates in A, BM, and L, all of them from the locality given in the protologue. According to Stafleu and Cowan (1981), most of Nakai's specimens have been deposited at TI; a part of type material kept at A and P. However, we have not found any duplicate in TI. From 1923 to 1925, Nakai visited the principal botanical institutions in Europe and North America, includ-ing herbarium A (Hara, 1953), and Eriobotrya acuminatissima was published in 1924. We concluded that E. acuminatissima was described based on the specimens in herbarium A when he visited the herbarium A at Harvard University. We therefore designated the specimen at A as the lectotype, since it was annotated by Nakai: "Type; Eriobotrya acuminatissima Nakai". Nouvelles D'extreme-Orient". According to Stafleu and Mennega (1995), Cardot's specimens are kept at P. We have located two sheets in P which represent duplicates from a homogeneous collection. The red tag on the sheet barcoded with P02143256 was obviously not made by Cardot, this sheet was not holotype (Art. 9.1, Turland et al. 2018). A further lectotypification thus is necessary. Furthermore, we have not found it to be published anywhere for the lectotypification (cf. Nakai 1924;Vidal 1965Vidal , 1968Vidal , 1970Kuan and Yu 1974;Gu and Spongberg 2003). We followed the informal typification designated by the staff in herbarium P, and herein formally designate the sheet (P02143256) as the lectotype. (P)". We located two syntypes kept at P, from which the lectotype could be chosen. These two specimens both have complete information in accordance with the protologue. We designated the sheet P[barcode P02143261] as the lectotype, as it has been labeled with a red printed tag "TYPE".  Stafleu and Cowan (1976), Chevalier's specimens were deposited in P and PC, although there is original material in other herbaria too. We located eight original specimens in A, C, K, L, and P, and all of them are in accordance with the locality in the protologue "Vietnam (Sud), prov. Nha Trang, massif du Hon Ba, 1000-1500 m". The author (Vidal) designated the specimens in P as type, however, three sheets were located in herbarium P. We designate the specimen (P[barcode P02143263]) with red printed tag "TYPE" as the lectotype.  Stafleu and Cowan (1979), the main set of Hooker's type and material has been deposited in K, although some of them are in MANCH and E too. We located one specimen in K with two labels, "Sikkim mts; Jayl? 9000 ft" and " [...] Hat. Sikkim; [...]; alt. 9000 feet; coll. J. D. H.". This locality fitted with the locality provided by Hooker: "Sikkim, alt. 5-9000 ft.", therefore, we designate this sheet K[barcode K000758394] as the lectotype of the name E. petiolata. We also located two gatherings from herbarium P collected by Griffith in the East Himalaya, Griffith 2085 (P02143222) and Griffith 2086 (P02143223, P02143224). These three specimens partially matched with the locality in the protologue: "Eastern Himalaya; [...] Griffith". So, these three specimens deposited in P are syntypes.  Vidal (1965) cited the gathering (E. Poilane 22591) deposited in P as the type. We located five sheets of this gathering in C, L, and P, three of them kept in the herbarium P. We designate the sheet (P02143226) as the lectotype, since it has a printed red tag "TYPE".  Vidal (1965) mentioned the following typification in the protologue: "Poilane 2345 (P)". We located three syntypes in P, from which the lectotype could be chosen, and two other duplicates were also found in A and L. Vidal's type material is commonly considered to be kept in P. We designate the sheet (P02143235) with a red printed tag "TYPE" as the lectotype. Craib, Bull. Misc. Inform. Kew 1929(4): 109. 1929 Craib (1929) mentioned the following locality in the protologue: "Satul, Adang, 500 m, on the rocky ridge, Kerr 14125". Vidal (1970) listed three herbaria, ABD, BM, and K, in which the type has been deposited, however, we located only four original specimens in BK, K, and TCD and did not find any duplicates in ABD and BM. According to Stafleu and Cowan (1976), Craib's type material has been deposited in K and WRSL. We designate the sheet (K000758408) as the lectotype, since it has a complete inflorescence.  Léveillé (1912) did not provide any typification information in the protologue, but mentioned it in his "Flore du Kouy-Tchéou" with the following information: "Pin-fa, montagne en pente, mai 1907, (Cavalerie 3220)". According to Stafleu and Cowan (1979), all of Léveillé's type specimens have been purchased by E in 1919, including the collections by J. Cavalerie. We located six duplicates among the available collections in A, E, K, and P, two of which have been deposited in E. Although all of Léveillé's names have been comprehensively reviewed by Lauener (1983), he did not provide the typification information. Since the sheet (E00011330) has been annotated by Léveillé "Hiptage cavaleriei Levl.." and is in a good condition, we designate this sheet as the lectotype, the other five ones as the isolectotypes.  Stafleu and Cowan (1983), Roxburgh's main collection should be at K. We have not found any potential original material among the available collections at K, and this was also indicated by Vidal (1965Vidal ( , 1968. Although some of Roxburgh's names have been validated and typified (Robinson 1912;Forman 1997;Turner 2013), Mespilus bengalensis has never been typified. Vidal, therefore, selected one specimen, Wallich 668.2, from K as the neotype.

Mespilus bengalensis
It should be noted that Vidal (1965) designated it as the lectotype, but later corrected this to neotype in his Flore du Cambodge, du Laos et du Vietnam (Vidal 1968 ". But we have not found any potential original material which was in accordance with the locality in the protologue among the available collections. According to Cowan (1976, 1979), Bank's type material is kept at CGE and K, and Lambert's herbarium was sold to many herbaria including K and BM after his death (Miller 1970