Introduction

Mitranthes O.Berg has a tortuous taxonomic history. It was described as a New World genus with five species, but no indication of a generic type, by Berg (1856) in his Revisio Myrtacearum Americae. Unfortunately, under modern circumscription, species that have been attributed to Mitranthes belong to distantly related genera, each pertaining to a different subtribe of Tribe Myrteae (Lucas et al. 2019): Eugenia L. (subtribe Eugeniinae), Myrcia DC. (subtribe Myrciinae), Myrceugenia O.Berg (subtribe Luminae) and Psidium L. (subtribe Pimentinae). These subtribes are difficult to recognise based only on flowering material, as embryo morphology is strongly diagnostic. Eugenia has a bean-like ‘eugenioid’ embryo in the classification originally established by de Candolle (McVaugh 1968), with a minute hypocotyl and well-developed plano-convex cotyledons that are either free or connate. Myrcia and Myrceugenia have similar ‘myrcioid’ embryos i.e., a C-shaped hypocotyl that encircles the well-developed, crumpled cotyledons, and Psidium has a ‘pimentoid’ embryo, i.e., a C-shaped hypocotyl with rudimentary, appendage-like cotyledons (McVaugh 1968). The distinction between these subtribes becomes particularly difficult if only flowering material is available, particularly if the flowers are isolated and the calyx closed in the floral bud. Isolated flowers occur in the four subtribes mentioned above, although they are exceptional in the Myrciinae (Lucas et al. 2019). Complete calyx closure in the floral bud is found sporadically in almost all Myrteae subtribes; its low value as a generic marker has only recently become clear (Landrum 1984; Proença 1990; Parra-O. & Bohórquez-Osorio 2016; Giaretta et al. 2018; Lucas et al. 2018).

Berg noted that his new genus Mitranthes included species previously assigned to Calyptranthes Sw. and Psidium by other authors, but believed Mitranthes to have eugenioid affinities, putting it after Myrcianthes O.Berg and before Calycorectes O.Berg (both eugenioid genera). The distinguishing characters given were a closed, calyptrate floral bud, solitary flowers and a pauci-ovulate, bilocular to tetralocular ovary; Mitranthes fruits were unknown (Berg 1855).

Mitranthes ottonis O.Berg was the only species to be described by Berg (1855) in the Revisio. Mitranthes browniana (DC.) O.Berg and M. eugenioides (Cambess.) O.Berg were transfers from Psidium, correctly citing the original descriptions. The fourth and fifth species, M. gardneriana O.Berg and M. ovalifolia O.Berg, were not described in the Revisio, but Berg alluded to his Flora Brasiliensis monograph as if it had already been published, although this did not actually appear until the following year, including these descriptions (McVaugh 1956a).

Due to this taxonomic confusion, lectotypification strongly affects Mitranthes’ placement within tribe Myrteae as its species are not closely related. There were three candidates for lectotypification, since Mitranthes gardneriana and M. ovalifolia were nomina nuda when the genus was described. Mitranthes has been lectotypified three times by three different authors. We will discuss these in reverse chronological order for reasons that will become clear.

Lectotypification of Mitranthes O.Berg

McVaugh (1956b) lectotypified Mitranthes (along with over 50 other Myrtaceae genera) choosing M. browniana (DC.) O.Berg = Psidium brownianum DC. He justified his choice believing Mitranthes browniana to have priority over M. ovalifolia; moreover, Berg in his writings had consciously assigned it to Mitranthes, whereas he had not seen the type of Mitranthes eugenioides. McVaugh (1956b) excluded Mitranthes ottonis from consideration because of the earlier publication of Psidium brownianum and because of Berg's statement on the transfer of the latter to Mitranthes. McVaugh’s (1956b) method was not mechanical but based on clear criteria. McVaugh (1968) later recanted on this, however, when he discovered that Burret (1941) had previously chosen M. ottonis as lectotype of Mitranthes. In a very influential publication, McVaugh (1968) also pointed out that Mitranthes ottonis O.Berg was closely allied to Calyptranthes Sw. (a ‘myrcioid’ genus) — thus establishing Mitranthes as a member of Subtribe Myrciinae. In fact, Mitranthes ottonis had long since been transferred to Calyptranthes by Charles Wright (1868). The third (chronologically the first) lectotypification of Mitranthes was proposed in Nomenclator Botanicus by Ludwig Pfeiffer (1873), unbeknownst to either Burret (1941) or McVaugh (1956b). Pfeiffer chose the same species that McVaugh (1956b) had chosen: M. browniana (DC.) O.Berg = Psidium brownianum DC. We quote verbatim the very clear explanation of how this happened given by Silva & Nelson (2008, p. 1352) who treated a similar case in Bangia Lyngb. (Bangiaceae): ‘Pfeiffer’s lectotypifications were not appreciated until contributors to the [20th Century] Index Nominum Genericorum (Plantarum), drew attention to them in the Praefatio to Nomenclator botanicus, wherein Pfeiffer stated that he was indicating type species for newly described genera or those with new circumscriptions’. The Pfeiffer (1873) lectotypifications have proved controversial in other groups (e.g., Asteraceae; Welker et al. 2021) as a result of their being overlooked for more than 100 years, during which time other taxonomic actions were taken and lectotypes proposed. Some of these steps now must be overturned as previous manoeuvres take priority, often to the dismay of those seeking taxonomic stability; it is likely that more such corrections will be seen in other taxa in the future.

All species assigned at some time to Mitranthes have now been transferred to other genera or have had their original basionyms re-established (Flickinger et al. 2020). Most Mitranthes names are currently assigned to Neomitranthes D.Legrand (13 names: Mattos 1966, 1981; Legrand & Klein 1977a, 1977b; Silveira 1981; Mattos 1989; Govaerts et al. 2008; Sobral et al. 2010). Others have been transferred either to Myrcia (five names: Flickinger et al. 2020), Psidium (three names: Niedenzu 1898; Govaerts et al. 2008), Siphoneugena O.Berg (three names: Mattos 1969; Proença 1990), Myrceugenia (two names: Landrum 1984), Blepharocalyx O.Berg (two names: Landrum 1986; Sobral et al. 2010) or Eugenia (one name: Giaretta et al. 2018).

Unfortunately, the forementioned lectotypification of Mitranthes by M. ottonis (Burret 1941), a Cuban species of recognised ‘myrcioid’ affinities, has been accepted by botanists for the past 50 years with several consequences. Legrand in Legrand & Klein (1977a, 1977b) erected the genus Neomitranthes to accommodate several Brazilian species that were similar to Mitranthes but had ‘eugenioid’ embryos. Since Legrand believed Mitranthes to be lectotypified by a ‘myrcioid’ species as stated by McVaugh (1968), a new genus to accommodate these ‘eugenioid’ species was necessary. Neomitranthes closely resembles Mitranthes in its closed floral buds, prolonged hypanthial tube and bilocular ovaries, hence Legrand’s choice of name. Neomitranthes was considered by Legrand to belong to Subtribe Eugeniinae but is now included in Subtribe Pliniinae due to molecular research (Lucas et al. 2019). Mitranthes has been cited in several recent papers as a synonym of Myrcia, also based on the belief that the lectotype of the genus is Mitranthes ottonis O.Berg = Myrcia ottonis (O.Berg) Flickinger, e.g., Lucas et al. (2018). The purpose of this paper is to correct this error. There is no reason to discredit the first lectotypification by Pfeiffer (1873), that is accepted by Index Nominum Genericorum and has been widely applied to other botanical families (Silva & Nelson 2008; Sennikov et al. 2011; Welker et al. 2021).

Therefore, we have applied article 10.5 of the International Code for the Nomenclature of Plants, Fungi and Algae, that states: ‘The author who first designates a type of a name of a genus or subdivision of a genus must be followed [unless] the author used a largely mechanical method of selection’. The genus Mitranthes should henceforward be treated as a synonym of Psidium, as we have recently done (Proença et al. 2022), where it has been recognised as Psidium section Mitranthes (O.Berg) Tuler & Proença.