Ceradenia, a New Genus of Grammitidaceae

Abstract

William J. Hooker's generic concepts for ferns essentially dominated the taxonomic community for the latter half of the nineteenth century. We now see that his genera, based almost solely on soral and venational characters, were nearly as artificial in conception as those of Linnaeus' sexual system for flowering plants. With the recognition in this century that some superficially similar groups such as the thelypterids and dryopterids were not at all closely related, the Hookerian dream of easy and obvious fern genera was permanently undermined. Today, even in complex groups, there is still resistance to the elucidation of genera based on nontraditional characters. Although conservatism is generally a cardinal virtue in taxonomy, the maintenance of very large, complex genera seems cumbersome for cases in which natural and discrete groupings of more manageable size can be demonstrated and defined by multiple, correlating characters. In my paper on Cochlidium (Bishop, 1978), I suggested that the only rational alternative to recognizing radically revised generic concepts in the Grammitidaceae is the inclusion of all species of the family into a single genus. Workers in at least two subsequent publications have adopted this approach for New World species (Tryon & Tryon, 1982; Proctor, 1985). The necessarily detailed studies to support more workable and hierarchic generic concepts in the family have proceeded rather slowly. But both Parris, from her studies of Indo-Malaysian species, and I, in my investigation of Neotropical groups, have become convinced that well defined, natural genera can be delimited among these ferns (Parris, 1984, 1986). As to the ease of generic recognition, those who wish to constitute the entire family Grammitidaceae as a single genus are recognizing a genus based on spore and sporangial characters, whereas the natural groupings within the family are to be founded on the more easily observed features of trichomes and general morphology. Among the larger grammitid ferns, the presence or absence of hydathodes seems to be a conservative character. In the Neotropics, the great majority of anhydathodous species probably constitute an allied group. The only Neotropical anhydathodous grammitids not part of this greater alliance are Grammitis graminea, G. turquina, and those few species of Grammitis sensu stricto in which the hydathodes are reduced to the point of extinction. In addition to lacking hydathodes, ferns of the alliance in question have concolorous scales, generally coriaceous fronds, and rather weak laminar setae. Another interesting character found among these ferns is the tendency for the base of the rachis or the distal stipe to be geniculate. This character is found among the hydathodous grammitids only in Grammitis asplenifolia and its immediate relatives.