Begoniayenyeniae (Begoniaceae), a new species from Endau Rompin National Park, Johor, Malaysia

Abstract Begoniayenyeniae is a new species of horticultural value known only from the Endau Rompin National Park, Peninsular Malaysia. It is similar to Begoniarajah with which it had previously been confused in the number of tepals and leaf characters. The new species is compared with three similar species, B.foxworthyi, B.rajah and B.reginula and photographs of all four species and descriptions of B.yenyeniae and B.rajah are provided. Molecular analysis using the ndhF-rpl132 chloroplast marker confirms the four species as distinct. Amongst native species, the three variegated species, B.yenyeniae, B.rajah and B.reginula, are some of the most popular Malaysian begonias in cultivation. Based on its restricted distribution, Begoniayenyeniae, under the IUCN Red List Categories and Criteria, is assessed as Critically Endangered.


Introduction
Asian begonias are prized for the eye-catching variety of colours and patterns of their variegated leaves and so are targets for horticultural interest as a genetic source for new hybrid cultivar development. However, a vast array of begonias is of conservation importance either because they have restricted distributions, being known from very few tury, it is remains popular in cultivation today (Thompson 1976, Kiew 1989b, Tebbitt 2005. However, its origin is still shrouded in mystery. Ridley (1892) recorded only that it was collected by a native collector from Tringganu (the Malaysian state of Terengganu). Despite continuous botanical exploration, it has never been re-found, although another species, Begonia reginula Kiew (Figure 4), similar in leaf characters, was discovered in Terengganu, but it proved to be a distinct species based on the different number of tepals (Kiew 2005, Tan 2016). With recent publications of higher level (sectional) molecular phylogenies of Begonia (Moonlight et al. 2018, Tebbitt et al. 2006, especially involving species from South East Asia (Thomas et al. , 2012, it is now timely to perform a molecular phylogenetic study that includes greater representation of Begonia species from Peninsular Malaysia.

Methods
The description and measurements of the new species are based on herbarium specimens deposited in the Kepong Herbarium, Forest Research Institute Malaysia (KEP) and Singapore Botanic Gardens Herbarium (SING) and fresh material cultivated in the KEP nursery. The diagnostic characters of the new species were compared with those of the type specimen of B. rajah in the Herbarium of Royal Botanic Gardens, Kew (K), the watercolour painting made from the original B. rajah plant ( Figure 1) and early botanical descriptions. As B. rajah was confused with the new species (Kiew 2005), a new description is provided here. Other similar species, B. reginula Kiew and B. foxworthyi Burkill ex Ridl. are shown by molecular data to share a close relationship and are compared with the new species based on herbarium specimens in KEP, protologues (Kiew 2005) and living material available in the KEP nursery. The conservation status of the new species is assessed following the standard IUCN Categories & Criteria (IUCN 2012). B. yenyeniae, B. foxworthyi, B. rajah and B. reginula Interspecific relationships amongst the proposed new species (B. yenyeniae), B. foxworthyi, B. rajah and B. reginula were inferred from our current, on-going phylogenetic study that focuses on Begonia species from Peninsular Malaysia. DNA sequences for the chloroplast ndhF-rpl32 intergenic spacer region were generated for 48 taxa of Begonia which included species mostly from Peninsular Malaysia (Suppl. material 1).

Species relationships amongst
DNA Extraction, PCR amplification and DNA sequencing DNA of Begonia species was extracted from living material or silica gel dried material using the NucleoSpin Plant II kit following the manufacturer's protocols (Macherey Nagel, Germany). Amplification of the chloroplast ndhF-rpl32 intergenic spacer was done using the forward and reverse primers ndhFBeg-F and trnLBeg-R, respectively, designed by Thomas et al. (2011). Each PCR reaction contained 12.50 µl of 2X Go-Taq Green Master Mix (Promega, U.S.A.), 1 µl of each forward and reverse primer (10 µM) (Integrated DNA Technologies, U.S.A.), 1 µl of bovine serum albumin (New England BioLabs Inc., U.S.A.), 7.50 µl of RNase free water (Bio Basic Inc., Canada) and 2 µl (~ 100 ng) of DNA template to make up 25 µl total volume. The following PCR profile was used: initial denaturation at 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 1 min, primer annealing at 55 °C for 1 min and extension at 72 °C for 2 min; followed by a final extension step at 72 °C for 5 min. Amplicons were visualised using 1.0% agarose gel electrophoresis stained with SYBR Safe DNA Gel Stain (Invitrogen, U.S.A.). Samples with the expected size (~ 750 bp) were purified using the Wizard SV Gel and PCR Clean-Up Kit (Promega, USA) before being sent for DNA sequencing using both the forward and reverse primers by First BASE Laboratories Sdn Bhd (Selangor, Malaysia). Both the forward and reverse sequences were manually edited (cleaned) and aligned to generate a consensus sequence for each sample using the programme BioEdit. Multiple sequence alignment of the generated sequences together with the ndhF-rpl32 sequences of 79 Asian Begonia species (from an earlier phylogenetic study by Thomas et al. 2011) downloaded from GenBank was performed using Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/) using default parameters. The resulting alignment file was then manually edited before analysis and is available from the corresponding author upon request.

Data analysis
The final alignment file, comprising of 126 taxa and 1244 characters, was subjected to Bayesian analysis performed using MrBayes 3.2.6 on XSEDE via the CIPRES portal (https://www.phylo.org/). Following Thomas et al. (2011), the GTR + G model was selected for the analysis with the following MCMC parameters: number of generations = 1,000,000; number of runs = 2; number of chains = 4; temperature parameter = 0.20; sample frequency = 1000; and burn-in fraction = 0.25. The resulting 50% majority rule consensus tree with node support expressed as posterior probabilities was viewed using FigTree (http://tree.bio.ed.ac.uk/software/figtree/). The programme MEGA 6 (Tamura et al. 2013) was used to compute values of genetic distance (gd) using the Kimura 2 parameter (K2P) model for interspecific comparisons between the three species, B. yenyeniae, B. rajah and B. reginula.

Results
Results of our on-going study from Bayesian analysis of 47 taxa of Begonia from Peninsular Malaysia, which included accessions of B. rajah, B. reginula and the new species B. yenyeniae, together with 79 other Asian species using the ndhF-rpl32 chloroplast marker, showed that all three species resolved together in a strongly supported polytomous clade (Bayesian posterior probability (BPP) support = 1) where B. yenyeniae clearly separated in its own subclade (BPP support = 0.98) with accessions of B. foxworthyi and B. nurii. The tree shows that the B. yenyeniae accession resolved in a relatively long branch which lends support to its separation as a proposed new species. Part of the tree highlighting relationships amongst B. rajah, B. reginula, B. foxworthyi and B. yenyeniae are presented in Figure  The ndhF-rpl32 DNA sequences of B. yenyeniae, B. rajah and B. reginula have been deposited to GenBank (accession numbers MH454102, MH454103 and MH454104, respectively). Thus, molecular data support the close relationships of these species and the genetic distinction of B. yenyeniae. It is noted that the phylogenetic tree shows that a single chloroplast marker has limited power to resolve interspecific relationships in this group and, clearly, additional molecular markers are required as certain accessions of the same species did not resolve in clearly differentiated clades.
Distribution. Endemic in Peninsular Malaysia, Johor, Mersing District, Endau-Rompin National Park, Sungai Selai. It is apparently a rare species as it is known only from the type locality in Endau Rompin National Park.
Etymology. Named after Dr Sam Yen-Yen, Malaysian botanist, specialist in Zingiberaceae who first discovered the species and recognised its potential as an ornamental plant.
Conservation status. Critically Endangered CR B2ab(iii, v), D1. Although the type locality is within a Totally Protected Area, it is known from only one population about 1.5 km 2 and its habitat is threatened by ecotourism activity and illegal collecting. The area of observed population covers about 1.5 km 2 (P.T. Ong pers. comm.).
Ecology. In primary lowland mixed dipterocarp forest, growing on moss-covered rocks, rarely epiphytic, near a waterfall in deep shade.
Note. The ovary of species in section Jackia have three locules, each with an unbranched placenta, but in this species two vestigial branches are present near the base of the placenta ( Figure 2G). This is also seen, but is less pronounced, in B. rajah.
Etymology. From a Sanskrit word raja = king, presumably referring to it as the most beautiful begonia species.
Conservation status. Extinct in the Wild (EW). However, it survives in cultivation since its first introduction into cultivation in 1894 (Anonymous 1894b, 1895).
Ecology. Its natural habitat is not known but judging from its growth requirements in cultivation, it probably grew in cool and shaded conditions. Other specimens examined. Peninsular Malaysia. Herbarium specimens made from cultivated plants in Royal Botanic Gardens, Kew Anonymous s.n. 1903 (K!); Sander F. et al., s.n. 1913 (K!); Anonymous s.n. August 1913 (K!). ing the living collection in the nursery; to C.M. Chen (Dr. Cecilia Koo Botanic Conservation Center) for advice on the cultivation of begonias in KEP nursery condition; to Y.Y. Sam and P.T. Ong for permission to use their photographs and to the Curators of the Kew and Singapore Herbaria for permission to examine specimens in their care.