Search and Access Archived Conference Presentations

2019 ASHS Annual Conference

The Deceptive Plumeria: The Quest to Find Meaningful Characters to Delimit Species within a Genus

Tuesday, July 23, 2019: 5:00 PM
Partagas 3 (Tropicana Las Vegas)
Kauahi Perez, University of Hawaii at Manoa, Honolulu, HI
Richard M. Manshardt, University of Hawaii at Manoa, Honolulu, HI
Teresita D. Amore, University of Hawaiʻi at Mānoa
Richard A Criley, University of Hawaii, Honolulu, HI, United States
Michael Kantar, Assistant Professor, University of Hawaii at Manoa, Honolulu, HI
Sterling Keeley, PhD, University of Hawaii at Manoa, Honolulu, HI
Kenneth W. Leonhardt, University of Hawaii at Manoa, Honolulu, HI
Species of the genus Plumeria in the plant family Apocynaceae are used for ornamental, cosmetic, and ethnomedicinal purposes. However, only a few of the species, specifically P. rubra, P. obtusa, P. pudica, and their cultivated varieties, find their way into commercial use, while other potentially useful species remain untapped sources of horticultural, medicinal, and cosmetic value. Disagreement among collectors and taxonomists alike, combined with misused or unconfirmed appellations and lack of unambiguous morphological descriptors and molecular markers, has thus made it difficult to identify species. Hence, a combination of 43 descriptive morphological characters and five molecular regions (ITS2, partial matK, psbJ-petA, trnH-psbA, rpl32-trnL) were evaluated to determine their efficacy of identifying groups of taxa as species within this genus. Iterative principal component analysis (PCA) was used to determine informative morphological characters that could distinguish among Plumeria spp. Maximum likelihood (ML) and Bayesian inference (BI) methods were employed on separate and combined molecular data to determine the phylogenetic utility and species discriminatory abilities of these markers. A total of 103 samples representing 16 putative species and botanical varieties were used for these studies. Iterative PCA revealed that some specimens formed distinct clusters, suggesting that certain taxa should be recognized as legitimate species, and verified that descriptive morphological characters will suffice for 8 out of the 11 Plumeria spp. examined in that particular data set. Molecular analyses revealed that rpl32-trnL provided the best phylogenetic signal and ability to discriminate 8 species clades, although other regions contained indels that were unique to different species. Concatenating rpl32-trnL with ITS2, trnH-psbA, and psbJ-petA enhanced tree resolution and phylogenetic signal. In addition, analyses confirmed that certain named species are in fact synonyms of P. obtusa or at least belong to the P. obtusa Complex. Further, analyses confirm that P. caracasana is not a synonym of P. pudica, and an accession known as ‘Isabella’ should be given consideration as a species. Although these individual molecular regions can be used as markers to distinguish Plumeria spp., there may be other molecular regions better suited for phylogenetic studies and species identification for this genus. Furthermore, the addition of quantitative, reproductive, and anatomical characters will likely enhance proper diagnosis of Plumeria species as a total evidence approach. In conclusion, the morphological characters and molecular markers used in this study can identify distinct morphological species in most cases.
See more of: Genetics & Germplasm 2
See more of: Oral Sessions