Melichrus

History of taxonomic study of Melichrus View in CoL

Luis Née made the earliest known herbarium specimen collection of a Melichrus species in Port Jackson, New South Wales, in 1793 ( Orchard 1999; JSTOR Global Plants, see http:// plants.jstor.org/). The specimen, housed at Herbario Real Jardín Botánico (MA 476478), was later described by Antonio José Cavanilles, as Ventenatia procumbens Cav. ( Cavanilles 1797, p. 28 & tab. 349). The first use of the name Melichrus was by Robert Brown in Prodr. Fl. Nov. Holland. ( Brown 1810) in which two species were described; Melichrus urceolatus R.Br. and M. rotatus R.Br. Brown described Melichrus urceolatus from material collected by botanical illustrator Ferdinand Lukas Bauer in 1804, at ‘(J.) [Newcastle; New South Wales]’ ( Brown 1810, p. 539; JSTOR Global Plants, see http://plants.jstor.org/; BM 000797778) and Melichrus rotatus from personal collections made at ‘(J.) [Sydney; New South Wales]’ and ‘(T.) [Fraser Island; Queensland]’ between 1802 and 1804 ( Brown 1810, p. 539; Orchard 1999). Brown (1810) acknowledged that Melichrus rotatus was synonymous with Ventenatia procumbens but did not make the recombination using the epithet ‘ procumbens ’ within the new genus, as nomenclatural precedence was not yet a formal protocol at that time. When Brown (1810) transferred the only other congeneric Ventenatia humifusa to Astroloma R.Br. , the name Ventenatia fell out of use. Later, nomenclatural precedence was acknowledged by English botanist George Claridge Druce, the name Melichrus procumbens (Cav.) Druce was formalised ( Druce 1917) and M. rotatus R.Br. was synonymised as a superfluous name.

Brown (1810) also placed Melichrus in a hierarchical classification for the first time when erecting the family Epacridaceae and established two infrafamilial sections based on fruit type. Melichrus was included in section 1, characterised by indehiscent fruits with one ovule per locule. This section was subsequently formalised as the tribe Styphelieae ( Bartling 1830) and remains currently recognised ( Watson 1967; Kron et al. 2002; Puente-Lelièvre et al. 2016; Crayn et al. 2020).

Allan Cunningham collected and recognised three new species of Melichrus . The first was collected from the ‘Plains at Bathurst’ ( Field 1825, p. 323) and called M. medius A.Cunn. due to appearing to be ‘intermediate between the two already described species [presumably referring to M. rotatus and M. urceolatus ]’ ( Field 1825, p. 323). Augustin Pyramis De Candolle described M. erubescens A.Cunn. ex DC. and M. adpressus A.Cunn. ex DC. from dried specimens collected by Cunningham in ‘plains … around Liverpool’ and ‘sterile land near Wellington Valley’ respectively ( De Candolle 1839, p. 740).

In the late 1860s, both Ferdinand von Mueller (1868) and George Bentham (1868) revised Melichrus in respective treatises on the Australian flora. Bentham retained a great deal of the generic structure of Styphelieae established by Brown, whereas Mueller had a broader concept of the genus Styphelia and expanded this to incorporate many genera, including Melichrus . Generic delimitation within Styphelieae remained uncertain, with early authors on the family divided (see Bentham and Hooker 1876; Drude 1887; Watson 1967) until recent molecular work decisively delimited Melichrus as a monophyletic genus and Styphelia was recircumscribed ( Puente-Lelièvre et al. 2016; Crayn et al. 2020).

Bentham and Mueller also differed regarding species delimitation within the group. Bentham, working entirely from dried specimens, could not detect notable morphological differences among most of the described species of Melichrus . Accordingly, M. adpressus , M. erubescens and M. medius were subsumed into M. urceolatus , and rotatus was retained ( Bentham 1868). Mueller, like Bentham, synonymised M. erubescens and M. medius with M. urceolatus . Melichrus urceolatus was subsequently transferred to Styphelia as S. urceolata (R.Br.) F.Muell. and M. rotatus was retained as S. rotata F.Muell. A third species, Styphelia cunninghamii F.Muell. was named from two collections – Hermann Beckler: Hastings River (New South Wales) and Walter Hill from southern Queensland ( von Mueller 1868, p. 39). Mueller chose to synonymise M. adpressus that was originally described from collections made in central western New South Wales with Styphelia cunninghamii presumably based on De Candolle’s description of M. adpressus as ‘a very distinctive species… [with] adpressedly imbricate leaves’ ( De Candolle 1839, p. 740). Bentham also cited the Beckler and Hill specimens but did not find these to be morphologically distinct from M. urceolatus .

The species-level taxonomy of Melichrus was not studied again until 1958. On the basis of much wider sampling and more detailed morphological observations than earlier authors, Paterson (1958) recognised four species; M. urceolatus , M. erubescens , M. procumbens and M. adpressus .

Since Paterson’s revision, two narrowly endemic species of Melichrus were collected for the first time in northern New South Wales. These were recently described as M. hirsutus J.B.Williams ex H.T.Kenn. & I.Telford and M. gibberagee J.B.Williams ex H.T.Kenn. & J.J.Bruhl ( Kennedy et al. 2020) based on strong morphological evidence, therefore six described species are currently recognised.

Taxonomic uncertainty, nomenclatural confusion and putative new species

In this study, Paterson’s (1958) revision (i.e. the status quo taxonomy) was used as a framework for explicitly testing species limits. Table 1 provides a full list of published and phrase-named species for Melichrus , accompanied by geographic distribution and a textual description of the symbol by which these are represented in figures.