Typifications and combinations in the Ebenaceae of Peninsular Malaysia and Singapore

Many of the current and synonymous names applied to the species of Diospyros L. (Ebenaceae) from Peninsular Malaysia and Singapore lack a type. Here we designate lectotypes and two neotypes and establish the correct combinations for the taxa belonging to the species hitherto known as Diospyros pilosanthera.


Introduction
Diospyros is being revised for the Flora of Peninsular Malaysia and the Flora of Singapore. In Peninsular Malaysia we recognise 65 species of Diospyros, 13 of which are endemic, whereas in Singapore there are 15 native species, none of them endemic to the island. Full descriptions, keys and notes on distribution and habitat will be given in the forthcoming accounts of the Ebenaceae in the Flora of Peninsular Malaysia and in the Flora of Singapore.
During the preparation of these accounts, it became clear that many of the names adopted in the region needed to be lectotypified, and that the lack of formal types resulted in species being incorrectly recognised and inappropriately named. The aims of this paper, therefore, are to designate the type specimens of accepted and synonymous names, and to establish correct denominations as relevant. We chose to publish these amendments independently from the two Flora accounts under way in order to benefit from a stricter nomenclatural peer-review, a freer format that allowed us to justify our choices and provide additional details, and an earlier publication, so that new taxonomic investigations can start using established names and types consistently.
Almost all species of Diospyros are dioecious. The majority of the fertile collections that we have seen thus far are in fruit, and some species are in fact only known from fruiting material. Of the flowering specimens, most are male, and the female flowers are frequently poorly known or unknown. A likely reason for this gap in the collections is perhaps a bias towards the phase that lasts longer, which in this case is the fruiting season. Furthermore, in many species the female inflorescences are shorter and fewer-flowered than their male equivalents, and thus less conspicuous among the foliage. The profusion of fruiting material shows that the female plants are not uncommon, therefore they simply are under-collected when in flower. Considering the prevalence of fruiting material in herbaria and in literature, whenever possible, we have given preference to fruiting specimens when designating a lectotype.
This work is based on herbarium collections from Peninsular Malaysia and Singapore currently housed at the Natural History Museum, London (BM, herbarium codes from Thiers, continuously updated), Herbarium Bogoriense (BO), the Royal Botanic Garden Edinburgh (E), the Royal Botanic Gardens, Kew (K and K-W), the Forest Research Institute of Malaysia (KEP) and the Singapore Botanic Gardens (SING). Type material from other institutions was seen via JSTOR Global Plants (https://plants.jstor.org/plants) and accessed between December 2018 and April 2019. Types that are cited in literature but were not available online for verification, are listed as "not seen".

Synonymisations and nomenclatural adjustments
There has been some uncertainty over the status of Diospyros pilosanthera Blanco var. oblonga (Wall. ex G.Don) Ng and its basionym D. oblonga Wall. ex G.Don. Ever since Ng (1977) reduced Diospyros oblonga to a variety of D. pilosanthera, the taxon has been inconsistently treated as a species or a variety by subsequent authors (Chong et al. 2009;Keng 1990;Phengklai 1981;Turner 1993Turner , 1995Wong et al. 2013). The taxonomic investigation carried out during the preparation of the Flora accounts for Peninsular Malaysia and Singapore has brought up the issue once again, and we concluded that the taxon named Diospyros oblonga or D. pilosanthera var. oblonga is synonymous with D. pilosanthera.
According to the International Code of Nomenclature (Turland et al. 2018), the name that takes priority is the oldest, however, Diospyros oblonga and D. pilosanthera were published validly in the same year. Diospyros oblonga was validly published in the fourth volume of Don's General History of the Dichlamydeous plants, which was issued in 1837 in parts that cannot be dated individually (Stafleu & Cowan 1976). Given that D. oblonga is presented at page 40, it could be assumed that the issue in which the name was included was released early in the year. The name D. pilosanthera was also established in 1837 but, according to Stafleu & Cowan (1976), Blanco's Flora de Filipinas reached other herbaria only at the end of the year. We do not know if the publication was circulated in the region earlier, potentially before Don's publication. Based on the current knowledge, and admittedly making several assumptions, we believe that the name Diospyros oblonga was published first, or at least should be accepted as the oldest name. Accordingly, a series of new combinations need to be made, and these changes are summarised here. New typifications are reported in the following section.
Diospyros oblonga Wall. ex G. Don Don (1837: 40); Diospyros pilosanthera Blanco var. oblonga (Wall. ex G.Don) Ng (1977: 237 In the original description of this species, the only gathering mentioned was Maingay 1514 from Penang, in Peninsular Malaysia. A first step lectotypification was made by Ng (1977), who designated the herbarium. Here, we establish the individual lectotype among the two sheets found at K.
Diospyros areolata King & Gamble (1905: 228 Blume (1826) did not give any information about the original material upon which the description of this taxon was based. Later, Ng (1977) made an initial lectotypification by stating that the type specimen (Blume s.n. from Java) was housed at L. We found several specimens at L which could potentially be designated as lectotype: [L0005899], [L0005900], [ L 0 0 0 5 9 0 1 ] , [ L 0 0 0 5 9 0 2 ] , [ L 0 0 0 5 9 0 3 ] a n d [L0005904]. Of these, only three sheets (L0005902, L0005903 and L0005899) have original collection data (Java, G. Parang). Here we designate the nicest specimen as the lectotype and the others as possible isolectotypes. The original description of the species provided information on the locality ("Sumatra occid. in prov. Priaman") and what is presumably the initial of the collector ("D." for Diepenhorst). The collection cited above appears to match Miquel's material and the herbarium designation is here proposed.  Ng (1977), and here we complete the typification by designating the unique specimen.

Diospyros dumosa
Diospyros ellipsoidea King & Gamble (1906: 220 Here we designate the lectotype among the four specimens of the type gathering. Diospyros juppii Merr. (Merrill 1923: 25 A primary lectotypification was made by Ng (1977), who cited the herbarium where the type is deposited (SING). However, as two duplicates were found at SING, we designate here a unique sheet in a second step lectotypification.
Diospyros kunstleri King & Gamble (1905: 226 Only one collection was cited in the original publication of Diospyros kurzii. A lectotypification was made by Ng (1977), as he designated the herbarium (K). However, there are three specimens at K, and the nicest sheet [K000792478] is designated here in a second step lectotypification. In the original description of this taxon, the gatherings mentioned by the author were Foxworthy from Selangor (Kopis and Sungei Buloh), and Kunstler from Perak. The latter was identified with less uncertainty than the former, and additionally duplicates were found in three main herbaria, therefore it was preferred as the lectotype collection. The three duplicates all have the same label with the same information, likely handwritten by the same hand, however it must be noted that the L and SING specimens are attributed to "Dr King's Collector" rather than Kunstler. The original description of Diospyros paniculata provided the locality, which is in today's Western Ghats of India. Three specimens potentially belonging to the original material were found at K and the one that best represents the key characters of the species has been designated as the lectotype.

Diospyros lucida
Diospyros penangiana King & Gamble (1906: 227 The type collection of Diospyros pubicalyx was designated in the original publication. Here we finalise the typification by designating an individual specimen. Here we also correct the epithet's spelling from "pubicalix" to "pubicalyx", following article 60.6 of the International Code of Nomenclature for algae, fungi, and plants (Turland et al. 2018).
Only one specimen with matching label data, apart from the collector name, was found, and is here designated as the lectotype.
Diospyros pyrifera Ridl. (Ridley 1908: 117, as 'pyriferus') SING52860]). We also found a specimen of Diospyros singaporensis collected on 13 Jan. 1920 by Deshmukh. The label used on this sheet suggests the provenance to be Penang Botanic Gardens. Since Deshmukh was based in Singapore and the species, date and type of collection (from cultivated) all matched with Bakhuizen van den Brink's data, we believe this to be the actual flowering specimen vouchered at Singapore Botanic Gardens from the same female tree as the 1920 collection. As this is also the only flowering specimen of the three retrieved, we designate it here as the lectotype.
Diospyros styraciformis King & Gamble (1905: 216 In the original description of this species, the author cited only one gathering: Hochreutiner 17, from Sarawak, Borneo. There are two specimens of this gathering at G and K. As Hochreutiner's types are housed at G (Stafleu & Cowan 1976), we prioritised this herbarium for our choice of a lectotype. The specimen deposited at G with barcode [G00341345] and with a (not original) "holotype" label, is in fact Hochreutiner 17a, and therefore not a type.  Ng (1977) in believing that Yeop's collection is in fact the material that was examined by Ridley, and here we complete his typification.