Comments: The traditionally recognized Sceloporus undulatus is morphologically highly variable (e.g., see Stebbins 1985, Conant and Collins 1991, Hammerson 1999). Recent genetic studies indicate that the species comprises multiple species that do not conform with traditionally recognized subspecies.
Leaché and Reeder (2002) examined range-wide mtDNA variation and identified at least four apparently monophyletic (but morphologically highly variable) groups, which they proposed as species under the evolutionary species concept (Eastern group: east of Mobile Bay; Central group: east of the Rockies and west of Mobile Bay; Western group: southern Wyoming to central Arizona and northern New Mexico; Southwestern group: eastern Arizona and central New Mexico to northern Mexico and western Texas). All of the groups are discordant with recognized subspecies circumscriptions. For example, the Central group encompasses six nominal subspecies ranging from the foothills of the Rocky Mountains in Colorado to the Gulf Coast of southern Mississippi. Populations of the morphologically distinctive subspecies erythrocheilus in central Colorado grouped with subspecies garmani (Central group) rather than with populations of erythrocheilus in south-central Colorado (Western group). Leaché and Reeder (2002) tentatively proposed the following names: eastern group, S. undulatus; central group, S. consobrinus; western group, S. tristichus; southwestern group, S. cowlesi. However, Leaché and Reeder identified no diagnostic characters for any of the proposed species, and the distributions of proposed species were only coarsely mapped and do not correspond closely with the distributions of previously recognized subspecies, leaving in doubt the specific identities of many Sceloporus populations.
Miles et al. (2002) examined allozyme variation (24 loci from 12 populations, 6 of the 11 recognized subspecies, plus 3 additional species and an outgroup species). Phylogenetic trees were inconsistent with current subspecific designations. Additionally, S. occidentalis, S. virgatus, and S. woodi arose within S. undulatus. The subspecies S. u. hyacithinus and S. u. undulatus are polyphyletic, and S. u. garmani and S. u. tristichus are paraphyletic. Two major lineages were identified: (1) a midwestern grasslands group (includes a population from St. Louis County, Missouri) and (2) various populations from eastern woodlands and western canyonlands.
Further integrated study of genetic variation, using mitochindrial and nuclear DNA, and more detailed genetic examination of various geographic areas (Niewiarowski et al. 2004; Leaché and Cole 2007; Leaché 2009) has helped clarify relationships among "S. undulatus" populations. Recognition of the four species proposed by Leaché and Reeder (2002) seems to be a justifiable change in the treatment of this complex, but the precise distributions of the taxa near some clade boundaries remain problematic. Leaché and Cole (2007) acknowledged the challenges imposed by apparent decoupling of morphological, karyotypic, and mtDNA divergence that may occur among populations in this complex and noted that conclusions about the number of species in the S. undulatus complex are directly linked to the particular "threshold' one imposes to define species status.
Specific distinctness of S. occidentalis and S. undulatus is confirmed by their sympatric reproductive isolation in southwestern Utah (Cole 1984, Smith and Chiszar 1989).