The Taxonomic Status of Saxifraga palmeri

Abstract

An examination of the type collection of Saxifraga virginiensi& var. subintegra Goodman (Rhodora 52: 138. 1950) leaves no doubt that it is conspecific with S. palmeri Bush (Am. Midl. Nat. 11: 220, 221. 1928). The type of S. virginiensis var. subintegra was based upon Goodman & Waterfall 4748 from MeSpadden Falls, Cherokee Co., Oklahoma, whereas Bush based his description of S. palmeri on a syntype, citing various collections from Van Buren, Arkansas, collected by E. J. Palmer (20731, April 10, 1922; 29710, April 19, 1926; 21020, April 30, 1922). Bush, in addition, cited various other collections from Arkansas made by Palmer. The problem is whether to accept this taxon as a variety of S. virginiensis or as a distinct species. This has involved a study of available material of S. virginiensis, S. texana, and S. palmeri. I wish to thank the curators for the loan of their material from the following herbaria: Chicago Natural History Museum (F); Gray Herbarium, Harvard University (GH); Missouri Botanical Garden (MO); University of Oklahoma (OKL); Oklahoma Agricultural and Mechanical College (OKLA); New York Botanical Garden (NY); Southern Methodist University (SMU); University of Arkansas (UARK); and United States National Museum (US). In his effort to elucidate more fully the detailed ranges of certain Saxifraga species, especially in the southern and midwestern United States, as treated by Johnson (Minn. Stud. Biol. Sci. 4: 1-109. 1923), Bush attempted to distinguish between a northern and eastern S. virginiensis Michx. and a southern and midwestern S. pilosa Haworth (Am. Midl. Nat. 11: 213-220. 1928). In this study he did not take into consideration the closely related Pacific Coast and northwestern states complex confused with S. virginiensis.1 Bush differentiated S. pilosa mainly on its very obtusely dentate, i.e. crenate, leaf blades, racernosely arranged infloreseence, and solitary short scapes, as opposed to the sharply serrate leaf blades, eymosely arranged infloreseence, and several or many scapes of S. virginiensis. While, admittedly S. virginiensis is variable in several characters (color, size, and number of petals, number of stamens, pubeseence and glandularity of scapes and pedicels, length of pedicels, size and shape of leaf blades, relative length of petioles, and number and length of scapes), there does not appear to be any definite correlation in the concomitant variations, nor any justification for maintaining segregated a southern and midwestern S. pilosa from a more northern and eastern S. virginiensis, as held by Bush. Some of the variations in length of petioles, size and shape of leaf blades, number and length of scapes, and length of pedicels, appear to be correlated with relative growth and maturity of the plant (i.e., plants in early spring generally