Taxonomy of Hill’s Oak (Quercus ellipsoidalis: Fagaceae): Evidence from AFLP Data

Quercus ellipsoidalis (Hill’s oak), an endemic of east-central North America, is morphologically similar to Q. coccinea (scarlet oak) and is subsumed into that species in several floristic treatments. This study uses data from more than 250 amplified fragment length polymorphism (AFLP) markers to investigate whether Q. coccinea and Q. ellipsoidalis are genetically distinct from one another. Whereas Q. coccinea and Q. ellipsoidalis separate from one another in all analyses, Q. velutina (black oak) populations collected from the geographic range of both Q. coccinea and Q. ellipsoidalis do not separate out by geographic region. This, combined with the strong differentiation between Q. coccinea and Q. velutina but weak differentiation between Q. ellipsoidalis and Q. velutina, supports the view that Q. coccinea and Q. ellipsoidalis are not simply regional variants of a single taxon. Moreover, while there is no evidence from the molecular data we collected of hybridization between Q. coccinea and Q. ellipsoidalis, the data suggest that there may be gene flow between Q. ellipsoidalis and Q. velutina. A clearer understanding of the relationships among these taxa is essential to understanding the taxonomy of Quercus section Lobate in eastern North America. Keywords—Amplified fragment length polymorphisms (AFLP), hybridization, oak taxonomy, Quercus section Lobatae. Quercus ellipsoidalis E. J. Hill (Hill’s oak) is one of the most problematic members of Quercus L. section Lobatae Loudon, the black oak section, in east-central North America. In a genus renowned as a “worst case scenario for the biological species concept” (Coyne and Orr 2004: 43), Q. ellipsoidalis is distinguished by the number of workers who have puzzled over its taxonomic status and proper identification (Trelease 1919; Jensen 1977, 1979; Overlease 1977; Maycock et al. 1980; Jensen et al. 1984; Hokanson et al. 1993; Shepard 1993). Quercus ellipsoidalis, when it was first encountered, was initially identified as Q. coccinea Münchh. (scarlet oak; Trelease 1919). Subsequent to the description of Q. ellipsoidalis based on specimens from the Chicago region in northeastern Illinois, U.S.A. (Hill 1899), many botanists accepted that this species was found in Iowa, Minnesota, Wisconsin, Michigan, and the northern counties of Illinois, Indiana, and Ohio, to the exclusion of Q. coccinea. Quercus ellipsoidalis is characterized by deeply lobed leaves with C-shaped sinuses; ellipsoid, often longitudinally striped acorns, the caps glabrous or very sparsely pubescent on the inner surface with tightly imbricated scales; and relatively small terminal buds that are glabrous or sparsely silky-pubescent (very occasionally densely pubescent) at least on the distal one-half to two-thirds (Fig. 1). Specimens housed in the major herbaria documenting the flora of northeastern Illinois and adjacent counties (WIS, MOR, and F) show geographic overlap between Q. coccinea and Q. ellipsoidalis in northwestern Indiana, with a few collections of Q. coccinea scattered in northeastern Illinois and southern Michigan (Fig. 2). The region around the type locality for Hill’s oak is the focus of substantial taxonomic disagreement. Some recognize Quercus ellipsoidalis as distinct from Q. coccinea and view both as present in northeastern Illinois (Trelease 1919; Jensen 1977, 1979; Jensen et al. 1984), while others hold that the two are not reliably distinguishable and best treated as a single species (Overlease 1977; Voss 1985; Shepard 1993; Swink and Wilhelm 1994). Because of difficulty in interpreting the morphological characters and apparent morphological intergradation between the two species, molecular data should be useful in evaluating whether they are best recognized as distinct at the subspecific or specific level, or whether they should be considered variants of a single, wide-ranging and