Updated description of Diospyrosdussaudii Lecomte (Ebenaceae), with lectotypification and notes on its distribution

Abstract Diospyrosdussaudii is a poorly known species and previous descriptions lacked details about its female flowers and fruits. The species had not been recorded since type collections were made in 1913. As a result of our Diospyros research in Thailand, new specimens and data are now available for this species. In this study, we provide an updated morphological description, illustrations, lectotypification and a distribution map. The species was previously reported to be endemic to Laos; as such, the occurrences in Thailand greatly expand the distribution of the species. In addition, we analysed the phylogenetic relationships between D.dussaudii and other Diospyros species from Southeast Asia and other regions, using DNA sequence data from eight plastid regions. Our phylogenetic analyses indicate that D.dussaudii is closely related to D.castanea, D.dasyphylla and D.insidiosa. Their taxonomic affinities are discussed.


Introduction
Although the taxonomy of the genus Diospyros L. in Thailand is fairly well documented, some specimens do not fit species in the known flora of Thailand (Phengklai 1981(Phengklai , 2005Duangjai et al. 2018Duangjai et al. , 2020Eiadthong 2020). This was the case for specimens collected from Phu Langka National Park in Nakhon Phanom Province, north-eastern Thailand in 2013 and from Chumphon Province, peninsular Thailand in 2014. The leaves and buds of these specimens resemble those of D. dasyphylla Kurz, which also occurs in Thailand. However, their fruits are globose or depressed-globose, ca. 4.0-5.0 cm in diameter and densely covered with orange hairs, similar to those of Sandoricum koetjape (Burm. f.) Merr. (Meliaceae), but different from those of D. dasyphylla, which has glabrous fruits. When we compared these specimens with species known to occur in Indochina (Lecomte 1928(Lecomte , 1930, as well as specimens from the Muséum national d'Histoire naturelle (P), we observed similarities in the leaves and buds to a poorly known species from Laos, D. dussaudii Lecomte. However, we could not positively identify the specimens due to lack of detail on the female plants of D. dussaudii in the published descriptions and due to the absence of male specimens amongst our collections.
Diospyros dussaudii is one of thirty Indochinese Diospyros species described by the French botanist Paul Henri Lecomte (1856Lecomte ( -1934 in 1928 (Lecomte 1928). Lecomte's description was based only on male specimens, which were collected by M. Dussaud from what is now the Lao People's Democratic Republic (PDR) on 4 October 1913. Until recently, the species was known only from one collection with three type specimens (P00721485, P00721486 and P02141495) held at P. The species is poorly known and was previously reported to be endemic to Laos (Lecomte 1928(Lecomte , 1930. It was not thought to occur in Thailand (Phengklai 1981) or China (Lee et al. 1996) and is not included in a checklist of vascular plants of the Lao PDR (Newman et al. 2007a). No information was available about the female flowers and fruits, which are important in the systematics of Diospyros and no specific locality was mentioned in the protologue (Lecomte 1928).
In 2019, Kwanjai Khammongkol collected additional specimens of the same unknown Diospyros species from Tat Pho Waterfall, Phu Langka National Park, Thailand. Later, in 2020, additional populations of this species were found in peninsular Thailand during floristic surveys conducted by teams from the Protected Area Regional Office 4 (Surat Thani) and the Surat Thani National Park and Protected Area Innovation Centre. Female and male flowers were collected from Surat Thani Province and Chumphon Province, respectively. Based on examination of type specimens at P, we identified a male specimen of the unknown Diospyros from peninsular Thailand as D. dussaudii.
In late December 2020, Sukid Rueangruea and Somran Suddee found a sapling of D. dussaudii during a field trip with Japanese botanists on the Bolaven plateau in southern Laos, but did not collect it. However, after searching for Laotian specimens in various herbaria, another Laotian collection (Newman et al. LAO 833) of D. dussaudii was found in E and L (L0409075). These specimens were collected in 2005 from Khammouan and were initially identified as Diospyros sp.
The objectives of this study were to report the rediscovery of D. dussaudii in Laos and provide an updated description of the species, as well as photographs, illustrations and notes on its distribution. In addition, we typified the species name and selected a lectotype. We also determined the phylogenetic placement of the species using DNA sequence data. Finally, to facilitate the distinction of D. dussaudii from closely allied species, a comparison of morphological characters is presented.

Morphological investigation, description and geographical distribution
Examination of D. dussaudii was based on specimens and preserved spirit collections obtained from north-eastern and peninsular Thailand. These voucher specimens, representing both male and female plants, were deposited in the Bangkok Herbarium (BK) and Bangkok Forest Herbarium (BKF). We also examined digital images of specimens held at BM, E, K, L and P (abbreviations follow Thiers 2020). We further compared these specimens with all published records of Diospyros species in Thailand and adjacent regions (Lecomte 1930;Bakhuizen van den Brink 1936-1955Ng 1978Ng , 2001Phengklai 1981Phengklai , 2005Lee et al. 1996;Singh 2005). Material collected from north-eastern and peninsular Thailand was photographed in the field. The habit, habitat, coordinates and elevation were documented in the field. Floral morphology was studied with dissecting microscopes at the Department of Forest Biology, Faculty of Forestry, Kasetsart University. An updated description of the species was developed from digital images of type specimens from P, the protologue, digital images of Laotian specimens held at L, specimens collected from Thailand and field observations. A distribution map, based on specimens and field observations, was created with Sim-pleMappr (Shorthouse 2010). The conservation status of the species was evaluated with IUCN Red List Categories and Criteria (IUCN 2019).

Phylogenetic analysis
One accession of D. dussaudii from north-eastern Thailand and three accessions from peninsular Thailand were compared with DNA sequences of eight plastid regions (rbcL, atpB, matK, ndhF, trnK intron, trnL intron, trnL-trnF spacer and trnS-trnG spacer). Total DNA was extracted from silica-dried leaf samples with a modified 2× cetrimonium bromide procedure (Doyle and Doyle 1987). The primers and polymerase chain reaction (PCR) protocol used for amplification are as described in Duangjai et al. (2009), except that we used 2× DreamTaq Green PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA), following manufacturer's protocols. Successfully amplified products were cleaned with FastAP Thermosensitive Alkaline Phosphatase and Exonuclease I (Thermo Fisher Scientific). The cleaned PCR products were sequenced with the same primers used in the initial amplifications. Sanger sequencing was performed at the Macrogen sequencing facility (Macrogen, Inc., Seoul, South Korea).
The DNA sequences of D. dussaudii were manually aligned to the dataset from Duangjai et al. (in prep.). Phylogenetic analyses were carried out with Maximum Parsimony (MP) and Bayesian Inference (BI; Rannala and Yang 1996;Yang and Rannala 1997). The MP analyses were conducted with equally weighted, unordered nucleotide substitutions (Fitch 1971) in PAUP* v.4.0b10 (Swofford 2002). The most parsimonious trees were searched heuristically with 1,000 replicates of random sequence addition; the settings included tree bisection and reconnection (TBR) swapping and MulTrees = on. TBR swapping was performed on a maximum of 200 trees (nchuck = 200) per replicate. Node support was evaluated with 1,000 bootstrap replicates of 1,000 random additions. BI was performed with MrBayes v.3.2 (Ronquist et al. 2012) on the CIPRES Science Gateway platform (Miller et al. 2010). Nucleotide substitution models were selected with the Akaike Information Criterion (AIC), implemented in MrModeltest v.2.3 (Nylander 2004). We performed two independent Markov chain Monte Carlo analyses, with four simultaneous chains of 10,000,000 generations, sampling one tree per 1,000 generations. The first 25% were discarded as burn-in and the remaining trees were used to construct a majority-rule consensus tree with Bayesian Posterior Probabilities (PPs). Euclea L., Lissocarpa Benth. and Royena L. (Ebenaceae) species were selected as the outgroup. Genetic distances between D. dussaudii and closely related species were generated with the Kimura 2-parameter model (Kimura 1980), with all gaps treated as missing (complete deletion option). DNA sequences of the eight plastid regions from four individuals of D. dussaudii were submitted to GenBank (accession numbers: MZ457089-MZ457112).
We compared the morphology of D. dussaudii with that of three species determined to be closely related on the basis of the results of the phylogenetic analyses. Data for D. castanea, D. dasyphylla and D. insidiosa Bakh. were obtained from previous studies (Lecomte 1930;Bakhuizen van den Brink 1936-1955Ng 1978;Phengklai 1981;Lee et al. 1996;Singh 2005) and supplemented by our own observations.

Results and discussion
After careful study of the protologue and type specimens ( Fig. 1), we determined that our collections from peninsular Thailand matched the description and type specimens of D. dussaudii. Diospyros dussaudii is a poorly known species and the protologue included a description of male plants only. The rediscovery of D. dussaudii in Laos and the recent collections from Thailand allowed us to complete the description of the species.

Phylogenetic analysis
We investigated the phylogenetic relationships of one individual of D. dussaudii from north-eastern Thailand and three individuals from peninsular Thailand using DNA sequence data from eight plastid regions. In addition, we clarified the phylogenetic relationships between D. dussaudii and other Diospyros species. When sequences of four individuals of D. dussaudii were included in the data matrix, the concatenated alignment of the 186-terminal dataset consisted of 8,293 characters, amongst which 1,991 sites were variable and 1,150 were MP-informative. MP analysis yielded 23,000 equally parsimonious trees with 3,484 steps (consistency index = 0.66; retention index = 0.84). The results of the MP and BI analyses were generally congruent; therefore, we present only the latter (Fig. 2). Phylogenetic analyses of eight plastid regions indicate with strong support (PP 1.0) that Diospyros and three other genera, Euclea, Lissocarpus and Royena, are monophyletic. The overall phylogenetic relationships of these four genera and of the clades recovered within Diospyros are congruent with previous reports (Duangjai et al. 2009(Duangjai et al. , 2018(Duangjai et al. , 2020. We identified 11 major clades of Diospyros, eight of which (I, III, V, VI, VII, IX, X and XI) include Asian species. The results of our phylogenetic analyses unambiguously place the four D. dussaudii individuals within clade XI (Fig. 2) with 43 other species from Asia, the Americas, New Caledonia and the Pacific Islands. Although relationships within the clade are not fully resolved, the four individuals of D. dussaudii group together with high support (PP = 1.00) and have a well-supported sister relationship with D. castanea (PP = 1.00) and this clade is sister to a clade of D. dasyphylla and D. insidiosa. The genetic distance between D. dussaudii and D. castanea, D. dasyphylla and D. insidiosa, based on data from eight plastid regions, ranges from 0.0041 to 0.0080, whereas intraspecific distances amongst the four individuals of D. dussaudii range from 0 to 0.0009 (Table 2) (Lecomte 1930;Bakhuizen van den Brink 1936-1955Ng 1978;Phengklai 1981).
Ecology. Scattered along streams in tropical rain forests and dry evergreen forests. The species occurs in the understorey at altitudes of 100-300 m.  As mentioned in Newman et al. (2007b), duplicates of Laotian specimens were deposited in three other herbaria in Laos, as well as at P. However, we were unable to study the specimens in Laos due to the COVID-19 pandemic. It is unsurprising that only a few specimens were collected from Laos because Laos has a low rate of botanical collection (Middleton et al. 2019).