Systematics of Desmos (Annonaceae) in Malesia

Abstract

Desmos is a woody climbing genus of Annonaceae which has been widely used as traditional medicine in several countries such as China, Indo-China, Thailand, and Peninsular Malaysia to treat dysentery, vertigo, fever, and postpartum. Phylogenetic studies on Desmos showed a close relationship between Desmos and Dasymaschalon. They share similar moniliform monocarpic fruits, but their flowers are different. It has been debated among Annonaceae researchers whether both genera should be treated as the same or distinct genus. Therefore, investigation on the taxonomic delimitation of Desmos species and Dasymaschalon will be necessary. Plastid genome (plastome) identification, leaf epidermal section of Desmos species and Dasymaschalon species, a taxonomic revision of Desmos in Malesia using morphological data, the spatial distribution of Desmos species in Malesia, and morphological and genetic variation in Desmos chinensis population using ISSR marker were used to investigate taxonomic delimitation of Desmos. Plastome study of Desmos chinensis and Dasymaschalon dasymaschalum was carried out to determine the origin, evolution, phylogenomics, and identify marker candidates for Desmos and Dasymaschalon or with other genera in Annonaceae. We discovered that the plastome of Desmos chinensis was circular with the size of 167,677 bp (smaller than Dasymaschalon dasymaschalum), consist of 111 genes, including 78 protein-coding genes, 29 transfer RNAs, and 4 ribosomal RNAs. The plastome showed a typical quadripartite structure, i.e., large single copy (LSC), small single copy (SSC), and two inverted repeats (IRs). The plastome of Dasymaschalon dasymaschalum was 191,488 bp in length, which also exhibited a typical quadripartite structure (LSC, SSC, and two IRs). Desmos chinensis and Dasymaschalon dasymaschalum showed different copies of accD, ndhD, ndhF, rps18, and ycf2. Protein coding genes of rpoB, rpoC1, rbcL, matK, rpl32, infA, and rpl36 were identified based on nucleotide diversity, sequence variability, and parsimony-informative sites. The genes are defined as molecular marker candidates of plastome for phylogenetic, barcoding, and population genetic research. We predicted that Desmos chinensis and Dasymaschalon dasymaschalum separated at ca. 64 million years ago during the Paleocene. Leaf anatomical feature study was carried to support the taxonomic delimitation of interspecific Desmos and intergeneric of Desmos and Dasymaschalon. Leaf anatomy in the paradermal section of 20 taxa (12 species of Desmos, eight species of Dasymaschalon) have been carried out. According to the leaf epidermal section, Desmos and Dasymaschalon are two distinct genera mainly distinguished based on the anticlinal wall undulation of epidermal cells and supported by variation of the crystal type and size. The anticlinal wall undulation of Desmos was almost straight to slightly wavy and never sinuous; meanwhile, Dasymaschalon varies from almost straight to deeply sinusoid. The crystals of Desmos consist of rhombohedral, druse type A, and druse type B crystals, whereas Dasymaschalon has prism, druse type A, druse type B, and druse type C crystals. The recent taxonomic revision using the morphological data was mainly based on the examination of about 1374 specimens from BO, L, WAN, ANDA, SING, SAN, K, P, HUH herbaria. The Desmos species in Malesia consists of eight species: D. acutus, D. chinensis, D. chryseus, D. dunalii, D. pedunculosus, D. subbiglandulosus, two new records: D. lawii and D. macrocarpus, and one reestablished: D. pachypetalus. Important features used to distinguish Desmos species in Malesia include leaf shape, leaf texture, indumentum density, indumentum type, tertiary venation type, flower position, pedicel length, pedicel bract position, the shape of outer and inner petals, outer and inner petals thickness, the undulation of outer petals, ovary indument, and monocarp length and width. Herbarium databases of the spatial distribution of species were used as important information for the conservation of Desmos species. The objective of the study was to provide information on environmental variables affecting the geographical distribution of Desmos species under different climate change scenarios. Nineteen bioclimatic variables were obtained from the WorldClim database. The importance of bioclimatic variables was identified using MaxEnt 3.4.1 software. The study results showed that the suitable areas for the current distribution of Desmos species were the lowland areas of less than 1000 m asl and 1000 m asl to around 2000 m asl for D. chinensis. The variable that most influences the distribution of D. acutus were temperature seasonality (BIO4), with a percentage of 63.1%. The variable that most influenced the distribution of D. chinensis, D. chryseus, and D. lawii was isothermality (BIO3) at 35.5%, 80.4%, 73.9%, respectively. The variable that most influenced the distribution of D. dunalii, D. pedunculosus, and subbiglandulosus was the precipitation of the coldest quarter (BIO19) with 35.5%, 36.8%, 49.2%, respectively. A study of the genetic variability of D. chinensis as the most widespread species was necessary for genetic conservation management. We analyzed variation in morphological characteristics and genetic diversity of D. chinensis using the ISSR marker. The morphological characteristics of D. chinensis in Java vary in many characters, such as habit; leaf shape, leaf size, leaf color; flower position, flower reflexed, pedicel bract position, number of petals, petals color, petals shape, petals apex, the undulation of petals; the number of monocarps per fruit. Fifty-three samples collected from seven populations were screened by 25 ISSR primers, wherein 11 primers produced 47 clear and reproducible bands that included 85.1% of polymorphic bands. Desmos chinensis in Java were divided into five natural populations using the STRUCTURE program. The mean value of the Shannon index (I) and genetic diversity index (h) were 0.223 and 0.151, respectively. Variations within and among populations examined using AMOVA showed the occurrence of 38% of total genetic variations among the different populations; therefore, the remaining 62% was due to variations within the population.