Typification of Notholaena R. Br. (Pteridaceae)

Abstract

Notholaena has been accepted widely as the name for a of ferns inhabiting arid and seasonally dry habitats. Prior to 1950, a few authors, notably Mettenius (1859), Domin (1914-1915), and Copeland (1947), treated Notholaena as a synonym of a broadly circumscribed Cheilanthes Sw. Since 1950, a few authors have followed Mickel (1979) this practice, and have done so the context of various floristic projects for the pragmatic reason of avoiding having to deal with some aspects of the complex morphological variation that can confound generic recognition the cheilanthoid ferns as a whole. In most floristic and taxonomic accounts of the group from the last two centuries, authors have accepted the Notholaena. However, typification of the name and resulting taxonomic circumscription of the have remained controversial, spite of attention from highly respected pteridologists. The typification of Notholaena, variously treated as a member of Pteridaceae or Sinopteridaceae, was discussed detail by Pichi Sermolli (1983, 1989) and is summarized briefly here. When he first described Notholaena a treatise on plants of Holland [Australia], Robert Brown (1810) named three Australian species, N. distans R. Br., N. vellea R. Br., and N. pumilio R. Br. He also admitted two additional taxa from outside the region to Notholaena, namely Acrostichum marantae L. [N. marantae (L.) R. Br.] and Pteris trichomanoides L. [N. trichomanoides (L.) R. Br.]. In a cryptic statement, he implied that his concept of the extended beyond those taxa specifically mentioned [...et aliae nonnullae ineditae, (and several others, unpublished)]. Brown's brief description was sufficiently concise that it applies to all five of the original species. In fact, it is sufficiently imprecise that it could be applied to many other species of cheilanthoid ferns: Sori marginales, continui v. interrupti. Involucrum nullum (nisi setae interstinctae v. squamulae lanave frondis). Three different attempts have been made to lectotypify Notholaena, but none has yet been universally accepted. John Smith (1875) published the first of these, selecting the Antillean Pteris trichomanoides as the type, though supplying no rationale for choosing this species. Underwood (1899) chose as the lectotype Notholaena distans, which occurs Australia, New Zealand, and a few other Pacific islands. This typification has the appeal that Brown was writing on Australian plants, and N. distans is the only type suggested thus far that is referable to an Australian taxon. However, Underwood selected it only because of the stated reason that it was the first taxon listed by Brown, a method of selection that has been considered largely mechanical by the Code (Art. 10.5; Art. 10 Ex. 7, a voted example). Few authors have adopted it, although it was independently redesignated by Copeland (1947). The third attempt to typify the was Index Filicum by Christensen (1905-1906), who designated Acrostichum marantae L. as the type. Like Smith (1875), Christensen did not provide any reason for his choice. Many later authors accepted Christensen's (1905-1906) lectotypification without further investigation. For example, the Index Nominum Genericorum project (Farr & al., 1979) cited N. marantae as the generic type, referencing Index Filicum without further comment. Pichi Sermolli (1977) adopted Christensen's lectotypification his comprehensive classification of fern families and genera and later (1983, 1989) provided arguments support of this choice. He also argued to overturn the initial lectotypification by John Smith since it was based upon a misinterpretation of its original description and was established without sufficient understanding of Brown's genus (Pichi Sermolli, 1983: 112-117). This part reflects the wording of editions of the ICBN prior to the Sydney Congress 1981 (e.g., Stafleu & al., 1978: Art. 8.1), before the option of superseding had the current more stringent requirement of the first selected type being in serious conflict with the protologue (Art. 10.5 of the current edition of the ICBN Greuter & al., 2000). Pichi Sermolli took this view mainly because Brown did not mention the possibility of fronds with farinose indument, such as possessed by N. trichomanoides. We point out that the ICBN has never required that a rationale be given for the choice of lectotype, nor does it suggest that one may exclude a species (like N. trichomanoides) from consideration that possesses a trait (such as farina) differing from the condition found other species included the original circum-