THE GENUS DIELLIA AND THE VALUE OF CHARACTERS IN DETERMINING FERN AFFINITIES

Abstract

IN THE CURRENT renaissance of fern taxonomy it vould seem important to emphasize the nature and relative value of the characters which can be used as evidence of relationship. Certain allegedly conservative characters, especially features of the sori, the characteristic aggregations of sporangia-such as their position in relation to the veins and leaf margins, their outlines, and protective indusiahave frequently been used as absolute criteria of affinities in the higher leptosporangiate ferns. Another character which has frequently been employed is leaf form-the outline of the blades and leaflets, the venation pattern, and the degree of dissection. These are conspicuous characters, easily seen with the naked eye or under a hand lens. But, unfortunately for the superficial observer, the ferns show numerous and striking, undoubted examples of parallel or convergent evolution in regard to these gross features. In the case of leaf form, the well-known reniform for instance, all have leaves which are nearly alike in appearance; but these are entirely unrelated ferns, as illustrated by such species as Adiantum reniforme, Pterozonium reniforme, Cardiomanes reniforme, and Lindsaea reniformis. The sorus, in spite of the fact that much of our classification has been founded on it, is itself equally liable to parallel or convergent evolutionary changes. Indeed, at one time practically all ferns which possess round sori but no indusia were placed in Polypodium; and ferns in which the sporangia covered the dorsal pinna surfaces were put in an all-inclusive genu,, Acrostichum, but now it is evident in case after case that this was an artificial and unnatural disposition. Even so recently as 1928, Bower considered the familiar genus Athyrium as being closely related to Asplenium because these two genera possess approximately the same soral type. However, information available now on such varied characters as scale structure, stipe anatomy, sporangial features, and cytology in the respective genera, rather firmly establishes the fact that the gross soral resemblance is probably only a coincidence, an example of convergent evolution. The purpose of this paper is to discuss and on