The Western Rattlesnake (Crotalus viridis), the most widespread and phenotypically variable rattlesnake in North America, has been partitioned into nine subspecies based on size, color, pattern, scutellation, and geographic distribution. Several subspecies, especially those occurring in the western portion of the range, have proved difficult to identify due to similarities in scutellation, color and pattern ontogeny, and because specimens fade in preservative. Over a half-century has elapsed since Laurence Klauber’s extensive work on the C. viridis group. Since that time, researchers have used a range of phenotypic characters to further elucidate relationships among members of this group, but a comprehensive phylogenetic analysis incorporating a wide range of morphological characters has not been accomplished. Analyses of snake lineages using mitochondrial (mt) DNA genes as molecular markers appear promising in reconstructing relationships, substantiating
previously identified clades, and revealing new ones. Since 1987, three studies using mtDNA gene sequences have partitioned
the C. viridis complex into eastern and western groups, but researchers have been unable to resolve relationships in the latter.
The goal of this study was to evaluate earlier results and to resolve relationships in the western clade by utilizing two rapidly evolving mtDNA genes (ATPase 8 and 6). We sampled from a wide range of populations, emphasizing those from the
Colorado Plateau because it is an area pivotal to understanding the evolution of C. viridis. In this region, six taxa (C. v. abyssus, C. v. cerberus, C. v. concolor, C. v. lutosus, C. v. nuntius, and C. v. viridis) potentially contact one another in the area of the Grand Canyon. The remaining three taxa occur west of the Colorado Plateau. Crotalus v. helleri and C. v. oreganus are found along the Pacific Coast and inland, and C. v. caliginis is restricted to Isla Sur of the Islas de los Coronados off the coast of Baja California Norte.
Our study includes all nine subspecies of the C. viridis complex plus two outgroups (153 individuals, 111 from the Colorado Plateau), and results were derived from 669 base pairs of sequence data. Net percent sequence divergence ranged from 0.4 ± 0.2 (viridis and nuntius) and 1.0 ± 0.3 (abyssus and lutosus), to 7.3 ± 0.9 (lutosus and nuntius). Weighted and unweighted maximum parsimony (MP), maximum likelihood (ML), and distance analyses rooted with Agkistrodon contortrix and Crotalus scutulatus all supported the same clades. Unweighted MP produced 99 equally most-parsimonious trees. A strict consensus, resampled 1,000 times, supported (94%) the monophyly of the C. viridis species group. The eastern clade (C. v. viridis and C. v. nuntius)
was 100% diagnosable, and in this clade C. v. nuntius was resolved at only 52%. Eastern and western clades differed at 6.1 ± 0.9 percent sequence divergence, and thus are on separate evolutionary trajectories. The eastern clade is most appropriately viewed as a distinct species, with C. v. nuntius placed in synonomy with C. viridis. The western clade is also well-defined (87%) and contains five Colorado Plateau lineages: (1) C. v. cerberus (72%), (2) C. v. concolor (92%), (3) a C. v. lutosus and C. v. abyssus clade (92%) that contains abyssus (88%) and a paraphyletic lutosus, (4) a paraphyletic C. v. oreganus, and (5) a C. v. helleri clade (89%) within which C. v. caliginis is nested. Crotalus v. cerberus is the basal-most taxon in our western clade and distinct from the other western clades. Relationships among the other western clades are less robust, suggesting a more-recent evolutionary history. Nonetheless, because these lineages are also well-defined and on separate evolutionary trajectories, we propose the elevation of C. v. abyssus, C. v. cerberus, C. v. concolor, C. v. helleri, C. v. lutosus, and C. v. oreganus to specific status. Furthermore, because C. v. caliginis is nested in C. v. helleri, we propose that this taxon should be placed in synonomy with C. helleri. The two undescribed clades within our western group (lutosus-like, L3; oreganus-like, O1) will require additional sampling and molecular, morphological, and natural historical analyses to clarify their taxonomy (i.e., potential new species)