Integrative Taxonomic Revision of Melichrus R.Br. (Ericaceae Subfam. Epacridoidae)

Abstract

<p>A taxonomic classification that captures both the evolutionary history and the morphological diversity of a group of organisms lays an essential foundation for studies on their biology, and for better management and conservation of the named species. The urgent importance of science-informed conservation management cannot be overstated amidst the global species extinction crisis, especially in Australia, which is both megadiverse and losing that diversity through species extinctions at a high rate. A comprehensive, integrative taxonomic revision of understudied groups, such as <i>Melichrus</i> R.Br. is a potent means of assuring species are discovered, described, named and ultimately conserved.</p> <p><i>Melichrus</i> R.Br. (Ericaceae: Epacridoideae) is a genus of shrubs endemic to Australia, with a widespread distribution in mainland eastern Australia (four species) and a more geographically constrained presence in south-western Western Australia (three phrase-named taxa). Prior to this study, <i>Melichrus</i> had received very little taxonomic study, and its few treating authors largely disagreed on species delimitation. The eastern Australian clade of <i>Melichrus</i> was last revised in 1958.</p> <p>This thesis investigated species boundaries in the eastern Australian <i>Melichrus</i> (Chapter 3) and the phylogenetic relationships amongst all species of <i>Melichrus</i> (Chapter 4), leading to a new revised taxonomy based on in-depth morphological and molecular evidence (Chapters 2 & 5).</p> <p>A thorough morphological dataset was constructed (90 individuals from 68 populations, for 26 characters) for analysis by NMDS ordination with Bayesian Inference cluster modelling (Mclust) and UPGMA hierarchical clustering to identify morphological discontinuities in the genus that could indicate species boundaries. Discontinuities in the morphological analyses were compared to those apparent in ordination and clustering (Principal Component Analysis, SplitsTree Neighbour-Net, STRUCTURE and ConStruct) and statistical analyses (<i>H_E, H_O, F_IS</i>and <i>F_ST</i>) of a DArTseq SNP dataset of 548 samples from 110 populations of <i>Melichrus</i>. This new evidence was interpreted to provide detailed recommendations for a revised species taxonomy of <i>Melichrus</i> (Chapter 3, Table 5).</p> <p>The first species-level phylogenetic trees for <i>Melichrus</i> were inferred from two independent genomic datasets: a DArTseq SNP dataset of 557 samples and 4,100 high quality SNP markers and a target capture dataset of 34 samples of <i>Melichrus</i> based on the Angiosperms353 bait set. The phylogenetic trees inferred from these complementary datasets were used to test the sister relationship between the eastern and western groups of <i>Melichrus</i>, the monophyly of species limits proposed in Chapter 3, and to gain a deeper understanding of the evolutionary relationships among species of <i>Melichrus</i>. HybPhaser was used to identify the parental lineages of a putative hybrid <i>Melichrus</i> sp. Mareeba (Chapter 4).</p> <p>I considered all available evidence to revise <i>Melichrus</i>, which included the description of eleven new species, clarification of the boundaries of previously described species, and the rectification of a longstanding nomenclatural misapplication (Chapters 2 & 5). In Chapter 2, two long-standing, APC accepted, phrase-named species were formally named and described as, <i>Melichrus hirsutus</i> J.B.Williams ex H.T.Kenn. & I.Telford and <i>Melichrus gibberagee</i> J.B.Williams ex H.T.Kenn. & J.J.Bruhl. Both species are gazetted as Endangered at the State and Federal levels and were therefore prioritised for description early in this study. Chapter 5 brings together the morphological and molecular evidence produced in Chapters 3 and 4 to describe nine new species. <i>Melichrus urceolatus</i> R.Br. was recircumscribed so that it is no longer paraphyletic and is reliably recognisable by its morphology. The long-standing misapplication of the name <i>Melichrus adpressus</i> A. Cunn. ex DC. was rectified. A dichotomous key to species is included in Chapter 5 with additional resources for identification, including an online key and curated photographs that will be deployed in the <i>Flora of Australia</i> (see Appendix 1).</p> <p>A revised taxonomic classification for <i>Melichrus</i> is produced in this study on the basis of strong new evidence, rectifying many of the long-standing taxonomic and nomenclatural problems inherent in the <i>status quo</i> taxonomy. The improved understanding of species limits and their phylogenetic relationships establishes a strong foundation for future biological research, including further systematics studies. Through the development of identification aids, the results of this research are now accessible and usable for a wide range of stakeholders, in turn facilitating better conservation outcomes for species of <i>Melichrus</i> in eastern Australia.</p>